CELEBRATING 100 YEARS OF MIT AERONAUTICS AND ASTRONAUTICS.
ANNUAL 2013-14 • DEPARTMENT OF AERONAUTICS AND ASTRONAUTICS • MASSACHUSETTS INSTITUTE OF TECHNOLOGY
Department Head Jaime Peraire peraire@mit.edu
Associate Head Editor & Director of Communications Eytan Modiano William T.G. Litant modiano@mit.edu wlitant@mit.edu
AeroAstro is published annually by the Massachusetts Institute of Technology Department of Aeronautics and Astronautics, 33-240, 77 Massachusetts Avenue, Cambridge, Massachusetts 02139, USA. http://aeroastro.mit.edu AeroAstro No. 11, October 2014 Š2014 The Massachusetts Institute of Technology. All rights reserved. DESIGN Opus Design www.opusdesign.us Cover: (Foreground) Professor Mark Drela and research engineer Dr. Alejandra Uranga prepare a model of the AeroAstro-designed D-8 commercial aircraft for testing in NASA’s Langley wind tunnel. A 737-size D-8 could use 70% less fuel than current planes while reducing noise and emissions. (Background) A World War I-era MIT aeronautics class learns to rig a biplane. (NASA and MIT Museum photographs)
Cert no. XXX-XXX-000
A GRAND 100TH YEAR
“MIT should lead in the study of aerial navigation.” – MIT President Richard Cockburn Maclaurin, Technology Review, July 1909 With these words, President Maclaurin set the stage for the introduction of aeronautics studies at the Institute in 1914. In that year, MIT offered its first undergraduate aeronautics course, and introduced the nation’s first graduate program in aeronautical engineering. To learn more about the early days of MIT aeronautics, we invite you to read MIT Museum Director of Collections and Curator, Science and Technology Deborah Douglas’ article in this issue of the AeroAstro annual. For even more history, including an in-depth look at 80 years of innovation in the department’s Instrumentation Laboratory (first at MIT and then at Draper Laboratory), we encourage you to read Kathleen Granchelli’s piece on the Draper Lab and Bill Litant’s historical photo timeline in the pages that follow. To commemorate the 100th anniversary of the Institute’s first aeronautics course and graduate degree program, the Department of Aeronautics and Astronautics is inviting its students, alumni, and friends to celebrate with us. In the spring semester, we kicked off the department’s Centennial Seminar Series with a talk by General Janet Wolfenbarger, USAF. General Wolfenbarger, a Course 16 alumna and Commander of the Air Force Materiel Command, is the first female four-star general in the Air Force, and one of only three woman four-stars in U.S. military history. To learn more about General Wolfenbarger, please see the Q&A with her in this issue of AeroAstro. At this writing, other Centennial Seminars planned for the fall semester include those by Jean Botti, Chief Technical Officer of Airbus Group, and Mark Lewis, Director of the Institute for Defense Analyses Science and Technology Policy Institute.
It’s been a grand 100th year for MIT Aeronautics and Astronautics, say AeroAstro’s Associate Department Head Eytan Modiano (left) and Department Head Jaime Peraire. (William Litant/MIT photo)
In May, we had a different kind of celebration, gathering more than 90 MIT alumni and friends at the Museum of Flight in Seattle. Boeing, which provided sponsorship for the event, arranged for special guest speaker Joe Sutter, former Boeing engineer and the man known as “the father of the 747.” We plan on additional alumni events across the country in the coming year. Without a doubt, the most anticipated of our centennial events was the October 22-24 Centennial Symposium. Publication deadlines necessitate our writing this introduction just prior to the event: anticipated participants were to include such luminaries in the aerospace field ¬ as Wes Bush, Norm Augustine, Kent Kresa, and Elon Musk. Other distinguished guests were many of the Apollo Astronauts, as well as Shuttle and ISS astronauts, industry leaders, academics, and policy makers. A memorable event, and we expect to make much of it available via video web postings. It’s been a good year for the department. We witnessed yet another increase in our total undergraduate enrollment, up to 168 from 154 the previous year. Our graduate program remains highly competitive, with
Lab report
i
a total graduate student body of just under 250. We are fortunate to attract increasing numbers of qualified women and underrepresented minorities to our programs. In fact, diversity was explicitly cited as one of the strengths of our undergraduate program in the recent ABET accreditation exercise, in which both our 16 (Aerospace Engineering) and 16-ENG (‘Flexible’ Engineering) degrees were accredited. This past year, AeroAstro had its debut in the MIT-Harvard massive open online course platform edX, with two online offerings on aerodynamics, detailed in this issue’s article by Dave Darmofal. The department’s research funding remains strong at more than $32 million per year, despite the uncertainty in federal budgets and sequestration. Always proud of its relaWe witnessed yet tionship with Boeing, MIT was honored another increase in our as a 2013 Boeing Company Supplier of the total undergraduate Year in the Innovation category.
enrollment, up to 168 from 154 the previous year.
On the faculty side, two highly esteemed colleagues, Manuel Martinez-Sanchez and Larry Young, retired in January, following long and successful careers in the department. We look forward to continuing interactions with both Manuel and Larry as they transition to this new stage in their lives. We’re most pleased to have hired two new faculty members — Leia Stirling and Woody Hoburg. Leia, who joined the department in July 2013 as an assistant professor affiliated with the Man Vehicle Lab, has research interests in computational dynamics, system automation, human factors, experimental biomechanics, and human-machine interaction for aerospace and medical applications. Woody, who joined us in August 2014, is an assistant professor affiliated with the Aerospace Computational Design Lab. Having spent the last year at Boeing, Woody’s area of interest is aircraft design and manufacturing, and multidisciplinary optimization. To learn more about their exciting research, please take a look at the articles from Leia and Woody in this issue of AeroAstro. Also Dave Miller, Director of the Space Systems Lab, was named NASA Chief Technologist, and will be on a two-year leave of absence from his faculty position in the department. During Dave’s absence, Alvar Saenz-Otero, an SSL senior research scientist is heading the lab.
In November 2013, the department hosted a one-day symposium celebrating the 20th anniversary of STS-61, the first Hubble space telescope-servicing mission. In addition to our own Jeff Hoffman, four of the six other mission astronauts — Story Musgrave, Dick Covey, Ken Bowersox, and Tom Akers — appeared on stage. It was a tremendous event in which we relived the experiences of the crew and those “behind the scene” who made the seemingly impossible possible. As mentioned in previous issues of AeroAstro, the refurbishment and improvement of our spaces and facilities is one of the department’s top priorities. We’re pleased to report that we have renovated 3,000 sq. ft. of new student space in the Ronald McNair Building (Building 37) and new labs in Building 35 devoted to space propulsion, nano-fabrication, and nano-engineered composites to support the research of Paulo Lozano and Brian Wardle. We have also been working with architects and planners to develop a grand plan for the renovation of Building 31, long-time home of the Gas Turbine Laboratory. When completed, the renovated building will support our activities in air transportation and autonomous systems, as well as provide additional teaching space for our growing student population. All in all, it’s been a great year for AeroAstro, and over the coming year we’re sure to have many more stories to tell. Stay in touch and, as always, we invite you to share your thoughts, ideas, and comments.
JAIME PERAIRE Department Head
EYTAN MODIANO Associate Department Head
1
Technology’s new wearable world
35 100 years of MIT Aeronautics and
Astronautics
By Leia Stirling
7
Myriad initial design considerations have far-reaching consequences
37 Images from 100 years of MIT
aeronautics and astronautics By William Litant
By Warren Hoburg
13 Undergrads design, build, and test
45 The most remarkable banquet in the
world
an artificial gravity spacecraft in a microgravity environment by Meera Chander
By Deborah Douglas
53 80 years of Draper Lab innovation
began with MIT’s Instrumentation Lab
23 AeroAstro launches Intro to
by Kathleen Granchelli
Aerodynamics on the web by David L. Darmofal
60 Lab Report A 2013-2014 review of Aeronautics and Astronautics Department Laboratories
31 ALUMNA INTERVIEW
General Janet Wolfenbarger rates four stars from the Air Force and AeroAstro
1
7
13
23
30
37
47
55
62
AeroAstro Man Vehicle Lab undergrad researchers Sean Kropp (left) and Aaron Okello face off with two different methods of motion capture; Sean’s requires special lighting and cameras in a lab environment, while Aaron’s transmits data and can be worn in “natural” environments. In the background, Professor Leia Stirling and graduate student Niek Bekers study this real-time monitoring. With the near-term possibility of flying older and potentially less healthy people into space, real-time monitoring during training activities and missions will allow faster intervention should a medical emergency take place in this higher-risk patient population. (William Litant/MIT photograph)
iv
AEROASTRO 2013-2014
Technology’s new wearable world By Leia Stirling
The presence of wearable technology is expanding at an incredible rate as adoption of these technologies becomes more common. Devices and applications designed for our omnipresent smart phones are being extended to clothing, watches, glasses, jewelry, and even fake eyelashes.
The capabilities of these systems are increasing, with the ability to make phone calls, take pictures, provide web access, and monitor physical performance. However, current consumer physiological monitoring systems are limited to heart rates, step counts, and GPS tracking. These limitations arise from how performance metrics are confounded in a natural environment, including a person’s natural variation in motion patterns, the changing environments in which we operate, our choice of clothing, and how we fatigue.
REAL-TIME BIOFEEDBACK Consider a runner training for the Boston Marathon. When training with a running coach — the expert in this scenario — the runner would be provided real-time feedback information on his or her biomechanics. This could include information on what part of the runner’s foot makes initial contact with the ground, orientation of the ankle when contact is made, and how to align upperbody posture. Without the coach, the runner is limited to knowing his or her pace and heart rate. The expert in this situation has the ability to make informed decisions based on extensive experience and direct observations of the runner and the environment. Current physiological monitoring systems have the potential to provide this information, but non-experts in biomechanics and sensor technology, like our runner, would face a wealth of data from which he or she would be unable to draw the proper conclusions. What makes an “expert” is the extensive training and experience that guides the individual to recognize the relevant information, project the potential issues, and thus make the appropriate recommendations. While this example is focused on running, the concept extends to decision making in many different domains that rely on physiological signals.
Technology’s new wearable world
1
The system architecture for wearable systems can be augmented to provide robust information that is interpretable by a non-expert. If we consider our runner, the device could be embedded within smart clothes and shoes that communicate to a smartphone interface. Sensors within the clothes and shoes would collect information pertaining to location of initial foot contact, ankle orientation, and body posture. However, the data streams would not be directly presented to the user. Instead, these With the creation of data could be presented through visualizations, audio instruccommercial spaceflight, tions, or haptic (e.g., vibration) stimuli to direct changes in the lay people with chronic runner’s biomechanics. As mentioned previously, the reason medical conditions may these devices are not currently available is that there is natural fly into space. variability in the individual and environment that is not well understood. For example, if the sensors are embedded in a shirt, motion of the shirt over the body may affect the interpretation of the data. One of my research group’s goals is to better understand these natural variabilities such that robust personalized physiological monitoring and decision-making are possible. In particular, we want the system to allow variability in sensor placement as a user may not know the optimal place to put the sensor, but the resulting action selected by the system wouldn’t change due to the selector placement. A key element in the system architecture is thus not only how the performance metrics are robustly calculated, but also how the data are presented to the desired user. Non-expert users could be an astronaut, soldier, nurse, or commercial user.
SPACE MEDICINE APPLICATION Historically, astronauts who have flown as part of NASA’s spaceflight program have been in excellent physical condition, with very few, if any, chronic medical conditions. This approach was important in a time when it was unclear what acute effects of exposure to the space environment would have on the human body. The field of space medicine has matured and we are now aware of the acute and chronic risks that are posed to these healthy individuals. However, with the creation of the commercial spaceflight industry, laypeople will have the opportunity to fly into space and will span a much larger age range and health profile than is typical within the astronaut corps. Some of these individuals may have chronic medical conditions that could be adversely affected by extreme environments.
2
AEROASTRO 2013-2014
With the near-term possibility of flying older and potentially less healthy individuals, it is imperative that the health of this population is systematically monitored in order to better understand the effect of the spaceflight environment on various medical conditions. Real-time monitoring of these individuals during their training activities and missions will allow faster intervention should a medical emergency take place in this higher-risk patient population. These systems are also relevant as we consider the longer-duration missions required for future exploration and bases on the Moon and Mars. Communication with experts on Earth will incur time delays as we move further away. A one-way transmission delay to the moon is around 1–2 seconds, which can cause confusion in conversations. The one-way transmission delay to Mars can range from 3–21 minutes, making real-time decision-making by an expert on earth infeasible in the event of a medical emergency. Thus, improving the robustness of physiological monitoring and aiding medical decision-making are required for missions of increased duration, which have greater medical risks.
TELEMEDICINE APPLICATION There are many examples where monitoring a patient in a home environment could provide improved decision-making opportunities such as daily monitoring of rehabilitation progress after an injury or due to a chronic condition, monitoring of electrical activity of the heart for those at higher risk for heart attack, or monitoring of chronic obstructive pulmonary disease for improved disease management. Each of these scenarios The one-way transmission delay to would yield a relevant data stream that would provide a clinician with valuable feedback on the patient condition. Mars can range from 3–21 minutes, However, large streams of data are difficult to directly making real-time decision-making interpret. One challenge is to develop indirect performance by an expert on earth infeasible in metrics that inform decision making. This requires the the event of a medical emergency. development of algorithms that take diverse types of data from different sources and combine them in a manner that allows the appropriate interpretation. The development of these algorithms to establish indirect performance metrics for functionally or clinically relevant tasks will permit the clinician to quickly interpret the data. Even more exciting is the potential to allow the patients themselves
Technology’s new wearable world
3
to be their own advocate in monitoring performance, with dynamic threshold alarms providing guidance for the user on when to contact a clinician. While, this data could also be provided to a robotic system designed to aid human performance. Advances in robotics, including miniaturization and new actuator technologies, have led to exciting advances in wearable robotics. Thus, in addition to sensing information about a person, the system could aid the user’s performance. Let’s specifically consider providing a device that can monitor and aid performance of object grasp and manipulation. There are many activities of daily living that involve precision grasping and manipulation, such as putting toothpaste on a toothbrush or feeding oneself. However, those with traumatic brain injuries or musculoskeletal injuries may lose the ability to actively and accurately control grasping. According to a 2012 Professor Leia Stirling wears a device that she’s using to determine how medical rehabilitation devices might someday be employed in the home report issued by the US Census Bureau, approximately 19.9 million for functional tasks. The device was developed by Stirling and a team people 15 years and older have difficulty with physical tasks from Wyss Institute in collaboration with Boston Children’s Hospital. (William Litant/MIT photograph) related to lifting and grasping, with 6.7 million people reporting difficulty grasping objects like a glass or pencil. Of those interviewed, only 41.1% with a disability were employed, while 79.1% without a disability were employed. By providing added functionality to people with disabilities, we may be able to increase their ability to perform tasks of daily living and increase their self-independence.
OUTSIDE THE LAB While the field of rehabilitation robotics and personal augmentation has seen significant growth, few studies have investigated these devices outside the research lab. The transition from research robotics that require an expert to aid in donning the device and adjusting the parameters for providing assistance, to at-home or in-the-field robotics implemented by the non-expert user requires systems that can be used during functional tasks, are easy to set up and configure, and have control systems that are synergistic with the user. The control system must allow the
4
AEROASTRO 2013-2014
natural variability in the motor patterns. This means that classic controllers that follow trajectories are not typically appropriate, as the trajectory to be followed is typically not known in a real environment. In an unknown environment with potential obstacles, the ability to estimate a user’s orientation alone does not provide enough information to determine when a wearable robot should assist. My research group is working on using surface electromyography (a method for recording the electrical potential generated by muscle motor units) to interpret a user’s intent as he or she interacts with the environment. Previous studies involving healthy young people and amputees provide promise for these techniques; however, muscle function degrades with age as well as with pathology. Thus, it is important that these methods are developed and evaluated with the relevant populations with realistic functional tasks such that they can truly be synergistic with the user in a natural environment. This includes the adaptation of learning algorithms that will permit personalized strategies, while remaining robust to the variability in placement of the system and of the environment. It’s possible to approach each of these problems independently. We are looking at different types of data, used for very different purposes. However, these problems are all linked in that they involve an understanding of the human in a natural environment, use measured data to infer an aspect of the human system, and then must be presented to a decision maker (human or robot system). The design and validation of an algorithm are dependent on the data collected during the development phase. Knowing the data must be presented to a non-expert may also affect how the algorithm is designed. Research from the human-centered viewpoint for decision making involves the understanding of the human in his or her actual environment, including the quantification of task variability, external confounding stimuli, the formalization of relevant expert knowledge, and presentation of the data to the decision maker. Through this unified path, we can address monitoring and aiding of human performance in a wide spectrum of applications from earth to space. LEIA STIRLING is an assistant professor and the Charles Stark Draper Professor of Aeronautics and Astronautics at MIT. Her research interests are in computational dynamics, human-machine interaction, system automation, human factors, and experimental biomechanics. She applies these interests to the development of tightly coupled human-machine systems, including wearable technology. Leia Stirling may be reached at leia@mit.edu.
Technology’s new wearable world
5
Warren “Woody� Hoburg spent 2013-2014 as an advanced technologist with Boeing Commercial Airplanes Product Development. In the fall of 2014 he joined the AeroAstro Department as an assistant professor and the Boeing Professor of Aeronautics and Astronautics. (Boeing photograph)
6
AEROASTRO 2013-2014
MAKING A MILLION DECISIONS
Myriad initial design considerations have far-reaching consequences By Warren Hoburg
Today’s aircraft are some of the most complex engineering systems ever conceived and built.
For example, a Boeing 777-300ER comprises approximately three million parts, provided by 500 suppliers worldwide. Designing, testing, certifying, and producing such a system is a monumental undertaking representing millions of decisions and years of effort. For aircraft manufacturers, airlines, government operators, regulatory agencies, or investors, the stakes are high — decisions made early in the design process lock in operational and manufacturing costs, marketability, and mission constraints for years to come. While design programs once focused heavily on vehicle performance (“further, faster, higher”), today’s design considerations stretch far beyond vehicle systems to production systems, supply chains, manufacturing processes, and operations. This complex web of interactions and tradeoffs drives hundreds of billions of dollars in annual economic activity. There is substantial industrial and academic interest in optimization models that can improve solutions, and enable more informed decisions earlier in the design process. Just as computation and simulation have pervaded all aspects of engineering, numerical optimization is becoming a widely used tool, now embedded in many analysis routines and software packages. This brings great opportunity, but also introduces unique challenges. Currently, starting with a clean sheet of paper, it takes approximately one decade to design a new commercial aircraft. To put this in perspective, Donald Douglas’ (an MIT alumnus and aeronautics assistant professor) 1933 DC-1 was designed and placed in service in fewer than two years. For today’s more complex systems, non-recurring design costs constitute a significant portion of most program budgets. Production tooling requires long lead times, forcing designs to be fixed
Myriad initial design considerations have far-reaching consequences
7
early on, yet unexpected design changes and complications require the ability to rework designs quickly. For all these reasons, there is value in shortening design iterations and empowering decision makers to make optimal decisions in real time. Despite remarkable progress on many fronts, reliable and efficient optimization remains difficult for aerospace systems. Many questions remain. Should problems be split up and tackled separately, or combined and solved centrally? How should we model and handle uncertainty? What roles can or should optimization play at the beginning of a new program, when an objective function may not even be available, or at more mature stages when missions may change in unexpected ways?
CONSIDER THE “TRAVELING SALESMAN” In my research, I spend a lot of time studying the mathematical problem structure underlying decision processes in aerospace engineering. Here’s why. In optimization, innocent-looking problem statements can conceal difficult or impossible computational challenges. Consider for example the “Traveling Salesman Problem,” one of the most famous optimization problems in computer science. Given a list of N “cities” (or soldering points on a circuit board, or fastener installation sites, or CNC goto points, etc.), and the distances between each pair of cities, the Despite remarkable progress objective is finding the shortest path through all the cities. on many fronts, reliable and Intuitively, this involves deciding which of the N cities to visit first, which of the N-1 cities to visit second, and so on, leading efficient optimization remains to N factorial possible routes. For 15 cities, there are 1.3 billion difficult for aerospace systems. possible routes, and for 60 cities, the number of possible routes is roughly the number of atoms in the observable universe. Enumerating all possibilities is out of question. This problem is popular and relevant, years of research have been dedicated to it. The best solution methods today include tuned heuristics that aggressively prune off provably-suboptimal states. In other words, they avoid evaluating routes that are clearly inferior. Using these algorithms, instances with tens of thousands of cities have recently been solved on supercomputers, generally requiring tens to hundreds of CPU-years of
8
AEROASTRO 2013-2014
MIT alumnus and former aeronautics assistant professor Donald Douglas’ 1933 DC-1 was designed and placed in service in fewer than two years. It can take 10 years to do the same with today’s complex aircraft.
effort. However, these approaches, are sensitive to the specifics of the problem, and there exist 100-city instances that will likely never be solved to optimality. In fact, the type of computational complexity inherent in the Traveling Salesman Problem underlies many modern approaches for cryptography and encryption. Contrast the Traveling Salesman situation with another famous problem, the linear program. A linear program is a special type of optimization problem where the objective and constraints are all linear (affine, technically) functions of the decision variables. Solving linear programs reliably has been possible for decades; it is a widely-used technology in operations research. For problems with hundreds of thousands of variables and constraints, solution times are reliably on the order of just a few seconds. However, for many engineering applications, the restriction to linear functions is too severe, and linear programming models are a poor fit for the actual functions of interest. Linear programs, however, are not the only type of optimization problem that is easy to solve reliably and efficiently. One of the quiet breakthroughs of the 21st century has been the maturation and commercialization of algorithms for solving a much broader class of convex optimization problems. Using algorithms developed in the past two decades, many general classes of convex optimization problems can be solved to a global optimum on a desktop computer with performance approaching that of solving linear programs.
Myriad initial design considerations have far-reaching consequences
9
In contrast with more general types of nonlinear optimization, convex optimization problems admit fast, robust solutions that do not require initial guesses, parameter tuning, or other inputs. Techniques from convex optimization underlie recent success stories from a range of fields, including machine Five hundred suppliers from around the world supply the learning, statistics, and controls. And, a special type of convex nearly three million parts it takes to make a Boeing optimization problem, the geometric program, appears espe777-300ER. Designing, testing, certifying, and producing such a system involves millions of decisions and years of effort. cially well-suited to problems in engineering, including (Boeing photograph) high-level tradeoffs in aircraft design. In initial experiments, this approach has enabled us to solve conceptual-stage aircraft design problems with hundreds of decision variables and constraints in a mere hundredth of a second on a dated laptop. My research reinforces the idea that computational tractability (and convexity in particular) can serve as a blueprint for structuring and decomposing computations. Basic, high-level relationships in aircraft design can be modeled or approximated within the convex optimization framework (and therefore solved easily). But for other aspects of aircraft design, I don’t expect to be so lucky. In systems integration and routing, many problems contain subproblems with discrete state spaces similar to that of the Traveling Salesman problem. The same is true of many problems in composite manufacturing. Put simply, this means that aerospace decisions commonly involve subproblems thought of as some of the hardest problems in computer science. Quantitative models from these domains are unlikely to fall to any standard nonlinear (or convex) optimization methods. Instead, we need the right tools for the job(s) and carefully chosen sacrifices.
BREAKING DOWN THE PROBLEM One common sacrifice is accepting a suboptimal solution in exchange for reasonable solution times. But, for how long should we search? And how can we best enable multiple optimization models to interact and exploit mutual benefits? One approach that can be applied involves decomposing the problems into (very difficult) combinatorial decisions, followed by an (easy) convex optimization step. Using this modeling breakdown, which has the added benefit of seam-
10
AEROASTRO 2013-2014
lessly combining interacting problems, we can efficiently focus available computing resources on the discrete search portion of the problems, and thus evaluate more possibilities in an allotted time. Another sacrifice is using simpler low-fidelity surrogate models in place of more complicated high-fidelity models. For example, well-established techniques in model order reduction can dramatically reduce the number of variables involved in a given model, while retaining as much predictive capability as possible. These techniques give us hope for more capable optimization models, but we Optimization is about must also be careful. As the differences between traveling understanding the boundary salesman problems and linear programs illustrate, the between what is and is number of variables involved in a model can be a very poor gauge of computational complexity. More work is needed on not possible. understanding how to combine these models and propagate uncertainties in the resulting optimization problem. Optimization is more than improving some status-quo or baseline solution. It is about understanding the boundary between what is and is not possible. It should help us gain intuition for impacts across systems, help quantify tradeoffs among competing objectives, and improve communication of constraints among teams or business units. In my new MIT research group, we will tackle applied optimization problems using powerful tools from convex optimization and computer science. We see the beginnings of a revolution in this area, in which designers, managers, and policy makers use optimization models in their daily work to improve their decisions and communicate with one another. Our goal is to understand and exploit problem structure in engineering problems, so that we can improve the solutions we find, how fast we find them, and our confidence in them. WARREN HOBURG obtained his B.S. in Aeronautics and Astronautics at MIT and his Ph.D. in Electrical Engineering and Computer Science at the University of California, Berkeley. He spent 2013-2014 as an advanced technologist with Boeing Commercial Airplanes Product Development. In the fall of 2014 he joined the AeroAstro Department as an assistant professor and the Boeing Professor of Aeronautics and Astronautics. Warren Hoburg may be reached at whoburg@mit.edu.
Myriad initial design considerations have far-reaching consequences
11
With NASA mentor Jennifer Rochlis’ (AeroAstro SM ’98, PhD ’02) assistance, senior Meera Chander (left) and recent MIT grad Joshua Oreman test their artificial gravity spacecraft design aboard a NASA reduced gravity flight. Developed by an AeroAstro student team, the test vehicle uses flywheels to induce vehicular motion, producing artificial gravity as well as steering/ navigation without a propulsion system. (NASA photograph)
12
AEROASTRO 2013-2014
FLOATING A UNIQUE IDEA
Undergrads design, build, and test an artificial gravity spacecraft in a microgravity environment by Meera Chander
If we desire to send astronauts on longer-duration space missions, we must alleviate the health risks associated with extended exposure to microgravity. Our experiment shows that a spacecraft with built-in artificial gravity could be a solution.
A round-trip Mars mission could last about two-and-a-half years, but astronauts who have completed missions on the International Space Station just a few months long have experienced significant adverse effects of microgravity, such as decreased bone density, muscle alteration and atrophy, and visual impairment. Rather than trying to repair the damage post-flight, an alternate approach would be to mitigate or totally prevent the cause of these issues in the first place. Our research team of four undergrads, faculty advisor AeroAstro Professor Sheila Widnall, and two mentors from NASA Johnson Space Center tackled this problem through NASA’s Systems Engineering Educational Discovery (SEED) program. Our idea: provide artificial gravity by spacecraft rotation for astronauts during long-duration space missions. Not only did our team design and build a model of our solution, an artificial gravity test vehicle; we also evaluated it as a proofof-concept through hands-on testing in weightlessness!
DISCOVERING THE OPPORTUNITY SEED is one of the many programs under NASA’s Reduced Gravity Education Flight Programs, where student researchers can conduct experiments aboard a microgravity aircraft. During the summer of 2012, I learned of this must-take opportunity from a colleague at my NASA internship. I told fellow AeroAstro student Henna Jethani, and soon we found two other undergrads, Libby Jones of AeroAstro, and Electrical Engineering and Computer Science student Josh Oreman, with
Undergrads design, build, and test an artificial gravity spacecraft in a microgravity environment
13
Main motor and flywheels Electrical components
(ESCs, relay, fuse, kill switch)
Electrical components (ESCs, relay, fuse, kill switch)
Inertial sensors
complementary skillsets to form a cohesive team. For the SEED program, NASA employees advertise research that they wish to conduct, and students submit proposals of interest to work under the mentorship of these employees. Since we were interested in conducting a dynamics-focused project, the team requested Professor Widnall, our 16.07 Dynamics instructor, to be our advisor.
Battery
Electronics boards
Astronaut habitat
The MIT students’ AGTV test model design.
In the fall of 2012, we submitted a proposal to SEED showcasing that we were a strong systems engineering team with a variety of strengths. We submitted a group profile, the projects in which we were interested, our own definition essay of systems engineering, our available budget, and outreach plans. In December 2012, the NASA Reduced Gravity Office selected us from 30 proposals as one of six teams to participate.
Over MIT’s January 2013 Independent Activities Period and into the spring, we worked with our mentors (a long distance relationship between MIT and NASA Johnson Space Center in Houston) to understand the Artificial Gravity Test Vehicle (AGTV) concept; design, build, and test prototypes; and gradually evolve the design. At the end of the semester, we packed our experiment, and in August 2013, we traveled to Ellington Field, near Johnson Space Center, to conduct research on Zero G Corporation’s microgravity testing aircraft.
THE MICROGRAVITY RESEARCH CHALLENGE Complexity, feasibility, time, and cost factors make it difficult to test microgravity experiments in the ideal environment on the International Space Station or on vehicles in space. There are ways, though, to achieve reduced gravity on Earth for enough time to conduct preliminary
14
AEROASTRO 2013-2014
experiments before advancing to the real microgravity environment, one of which is an aircraft that flies approximately parabolic maneuvers. However, this isn’t your normal commercial flight. This modified 727-200 doesn’t have windows and has only a few seats where researchers sit during takeoff and landing. The rest of the plane is padded and open for researchers and experiments to float and experience microgravity. In flight, the aircraft first enters a 45-degree climb, during which the passengers experience about 15-20 seconds of hypergravity (about 1.8-2g). Then, as the pilots bring the nose of the aircraft down and reduce thrust, everything and everyone aboard the aircraft enters free fall and experiences weightlessness for about 20 seconds. The aircraft then gains speed and pitches nose-down before the maneuver starts again for another parabola. In our AGTV, a scaled-down habitat only big enough for a Lego astronaut, is connected to a truss about two feet long, Successful slewing would on the other end of which is a set of spinning wheels. The ultimately allow for steering wheels’ rotation on one end causes the entire spacecraft to and navigation of the entire rotate in the opposite direction without the use of propelvehicle without the addition lant or similar resources. This is due to the principle of of a propulsion system. conservation of angular momentum. Stable rotation of the spacecraft induces a gravity vector along the direction of the truss and within the habitat. In our experiment, a mini-astronaut would only experience 0.1-0.4g of acceleration. We also investigated the use of a secondary spinning set of flywheels on the middle of the truss to slew or tilt, the rotation plane of the assembly in the remaining two axes. Successful slewing would ultimately allow for steering and navigation of the entire vehicle without the addition of a propulsion system. The successful completion of this project would be a step forward for a better understanding of the complicated dynamics behind the AGTV concept. Its performance could be applied to additional testing, and scaled to a larger model that would accommodate full-sized astronauts and provide 1g in their habitat.
Undergrads design, build, and test an artificial gravity spacecraft in a microgravity environment
15
BUILDING THE EXPERIMENT Our research began with identifying and understanding the physical concepts at play. The most important concepts applied were gyroscopic motion, moments of inertia and their associated rotation rates and induced accelerations, angular motion, and dynamic stability of the model about all three axes. With these concepts in mind, we were able to refine the configuration of our AGTV to lessen complexity, place the center of gravity in an optimal location, and add greater chance of rotational stability. We developed a mathematical model to determine the size of the flywheels and size the entire system such that the resulting rotation rate desirable for flight testing (about one revolution every four seconds). This rotation rate would be sufficient to produce measurable accelerations in the astronaut habitat, and was not fast enough to appear dangerous as a free-floating experiment on the reduced gravity aircraft. To make all of these hardware components run and collect data, After ensuring that our we implemented a laptop control and data collection system, sensors and laptop control and a sensor system consisting of two accelerometers and one system worked and produced microelectromechanical gyroscopic sensor.
reasonable and valid outputs, we had to cross our fingers and hope that our system worked.
After the theory work, we performed a number of 1g tests during the project’s design/build phase to refine the design and become familiar with what behavior we might expect in flight. We first conducted a spin-up test of the main motor with two flywheels mounted to one end of a bare truss. We suspended the system from a string attached to the truss at the system’s horizontal center of gravity, and when we rotated the main flywheels via radio control, the entire system rotated in the opposite direction, as we anticipated it would. The first test proved that angular momentum is, in fact, conserved. As we added more components to the truss, such as the slewing flywheel and other electronics, we realized that analyzing our system’s behavior on the ground would be more difficult than we thought. It was impossible to attach the string to the exact center of gravity such that our system was neutrally stable around it; our string suspension setup imparted forces and torques that the systems otherwise would not feel in microgravity. After ensuring that our sensors and laptop control system worked and produced reasonable and valid outputs, we had to cross our fingers and hope that our system worked off of our theoretical knowledge, since our ground test capabilities were limited.
16
AEROASTRO 2013-2014
The AGTV, as built by the students in AeroAstro’s Gelb Student Laboratory. Vehicle components are surrounded by safety shields and padding, protecting the transporting aircraft, researchers, experiments, and the vehicle itself.
CLEARED FOR TAKEOFF In addition to constructing our experiment, we, like all good engineers, had to extensively document our experiment background/description, specifications of our equipment and system, structural verification, electrical component analysis/verification, flight test plan, and hazard analysis. We submitted all of this to NASA as a Test Equipment Data Package (TEDP). Our TEDP had to be approved long before we landed in Houston. When we arrived at Ellington Field and assembled our experiment, our team had to present how our project works and submit our flight test plans to a panel of about 30 electricians, safety engineers, flight test directors, and others in a Test Readiness Review. We thought we would be bombarded with questions and concerns, since our experiment was the only one in this testing group that would float freely, and all the other researchers’ experiments were to be fastened to the aircraft deck. However, the panel readily endorsed our explanations, and we were cleared to conduct our experiment aboard the microgravity aircraft!
Undergrads design, build, and test an artificial gravity spacecraft in a microgravity environment
17
The official MIT Microgravity Research Banner is signed by MIT researchers who fly aboard the zero-G aircraft.
In day-to-day life and in normal scientific research, it’s easy to take gravity for granted. When researching on the microgravity aircraft, we had to worry about things like tying our pens to our clipboards so that they didn’t float away, Velcro-ing our laptop and clipboards onto the aircraft floor, and precisely placing straps that we could hold onto so that we wouldn’t float around uncontrollably along with our experiment. Before our test flights, we had some time to leisurely explore the aircraft and assess our testing environment. We made sure everything was in place — we even marked an “X” on the floor above which we wanted to release our experiment. We anticipated that our actual research time would go by in a flash, and we wanted to make sure everything that we could think of and control was taken care of so that surprises would be minimized on test day.
18
AEROASTRO 2013-2014
TESTING IN MICROGRAVITY August 1, 2013: the day finally arrived for our reduced gravity test flight! The first flight was a checkout that explored the many test variables; the second flight, which took place the following day, expanded on the interesting situations or unclear data points found during the first flight. Both flights were on a 727-200 and involved 30 microgravity parabolas, one lunar-g (.18G) parabola (not used for data collection, just doing moon jumps!), and one Martian-g (.38G) parabola (also not used for data collection). You can see our aircraft’s flight paths by visiting http://flightaware.com/live/flight/AJT578 and http://flightaware.com/live/flight/AJT579. There were two students and one NASA mentor per flight. One student was responsible for releasing the experiment for a trial, the other operated the experiment by sending wireless commands from a laptop, and the NASA mentor retrieved the experiment at the end of the parabola. We anticipated that our
actual research time would go by in a flash — we wanted to make sure … surprises would be minimized.
Our experiment investigated two major concepts: demonstrating a stable overall apparatus spin, induced through the spinning of the main motor, and thus creating artificial gravity in the astronaut habitat (measured through an accelerometer); and demonstrating the slewing maneuver, where the rotational plane of the overall system tilted using a slewing motor. We planned out exactly what we wanted to test during each parabola to get as much data as possible and characterize our AGTV.
EXPERIENCING THE “WEIGHTLESS WONDER” Nothing yet has topped our experience on the “Weightless Wonder.” Before our flights, all of us received a briefing with tips on the do’s and don’ts of microgravity research. For example, we were advised to not move our heads around too much, and to take some motion sickness medicine. When taking off from Ellington Field, all of us were sitting in normal airplane seats, but once we reached the Gulf of Mexico, we were allowed to assume our positions near our experiments. The flight test directors in blue flight suits periodically updated us during each phase of flight. During the first parabola, our mouths were wide open in disbelief as we experienced
Undergrads design, build, and test an artificial gravity spacecraft in a microgravity environment
19
The MIT Microgravity Research team: (front) Meera Chander (AeroAstro, SB ‘14), team lead Henna Jethani (AeroAstro, SB ’14); (center) Libby Jones (AeroAstro, SB ’14), NASA Mentor Thomas Sullivan; (rear) Josh Oreman (Electrical Engineering and Computer Science, SB, ’13), and NASA Mentor Jennifer Rochlis (AeroAstro SM ’98, PhD ’02). Not pictured: team advisor Professor Sheila Widnall. (NASA photograph)
completely foreign sensations — our eyes were seeing one thing, but our bodies were feeling another. Our brains were totally confused. This is why we didn’t plan any research for the first few parabolas, devoting this time solely to getting our bodies accustomed to microgravity. The final few parabolas of each flight also didn’t included planned research — we were allowed this time to explore the unique experience of microgravity, swimming through our research area, dingo flips, and having fun! There were times when some of us felt a little sick, but after just a few moments of lying strapped down, we felt fine.
20
AEROASTRO 2013-2014
AN EXPERIENCE AS VALUABLE AS THE RESEARCH With future long-duration space missions in mind, the MIT SEED Team successfully designed, built, and tested a proof-of-concept spacecraft model that would allow for astronauts to tolerate an artificial gravity environment, alleviating health issues experienced by astronauts. Two main goals of this model were demonstrated: stable rotation of the overall system due to rotation of the system’s reaction flywheels, resulting in artificial gravity production in levels that we expected; and slewing of the overall system to ultimately allow for navigation and steering. Our experiment was successful, but we also grew profoundly as engineers in the process. We learned lessons in teamwork, leadership, and project management while having to communicate in different locations and different time zones. Each team member offered different skillsets and expertise, and we had to trust and rely on each other to complete all tasks, all the while balancing school and other We gained valuable experience commitments. We gained valuable experience in high-stakes in high-stakes testing and testing and at-the-moment decision making. Finally, we were able to work with a real-world application of concepts learned at-the-moment decision making. in class, and also practice compliance with real NASA procedures and regulations. Most of all, though, we are proud to contribute to MIT tradition of microgravity testers, joining the other MIT research groups that have previously flown experiments on microgravity aircraft and signed the official MIT Microgravity Research flag — and are proud to take this experience with us as we progress with our aerospace careers beyond MIT. MEERA CHANDER (AeroAstro SB ’14) is a flight test engineer in Boeing’s Propulsion Analysis Department, where she tests engines on various aircraft. She enjoys flying as a private pilot and stargazing/astronomy. Chander may be reached at mchander@mit.edu.
Undergrads design, build, and test an artificial gravity spacecraft in a microgravity environment
21
Professor Dave Darmofal was instrumental in introducing AeroAstro’s first Massive Online Open Course, 16.101x Introduction to Aerodynamics. His next challenge is to employ online learning tools in on-campus teaching. (William Litant/MIT photo)
22
AEROASTRO 2013-2014
16.101X: AN AERODYNAMIC MOOC
AeroAstro launches Intro to Aerodynamics on the web by David L. Darmofal
MIT’s move into online courses through the creation of MITx and edX has led to a wave of excitement in education on campus.
Not only are we asking questions about how to effectively educate students around the world in Massive Online Open Courses (MOOCs), but also we wonder how MOOCs might improve our on-campus education. In AeroAstro, we introduced our department’s first MOOC 16. 101x Introduction to Aerodynamics, in the fall of 2013. In fact, we believe that this was the first time that aerodynamics was taught in a MOOC setting.
OUTCOMES-BASED DESIGN The main motivation for creating 16.101x was to eventually use the content to improve our on-campus undergraduate subjects; in particular, Unified Engineering and 16.100 Aerodynamics (a junior/senior level technical elective). As a result, the learning objectives for 16.101x were derived from the on-campus subjects. A best practice in educational design is to state the objectives of a subject as desired measurable outcomes in terms of student abilities, and then use the outcomes in the design of all aspects (lectures, sample problems, homework, projects, exams, etc.) of the subject. For example, Measurable Outcome 5.2 of 16.101x is, “A student successfully completing 16.101x will be able to derive the Bernoulli equation from the incompressible momentum equations, describe the assumptions required to apply the Bernoulli equation, and apply the Bernoulli equation to solve fluid dynamic problems.” The 16.101x structure was organized into a set of 11 modules; eight were required parts of the course, while the other three were provided for background, prerequisite knowledge. The different modules overlapped with our on-campus subjects. The required portions of 16.101x
AeroAstro launches Intro to Aerodynamics on the web
23
PRE-REQUISITE, NOT IN GRADE
Aircraft Performance
Incompressible Laminar Boundary Layers 16.100
UE, 16.100
Differential Form of Governing Equations
Control Volume, Cons. Of Mass & Momentum UE
UE, 16.100
Shock-expansion Theory
Cons. Of Energy UE
UE
Incompressible Potential Flow Fundamentals UE, 16.100
Boundary Layer Transition & Turbulence 16.100
2D Potential Flow Aerodynamic Models UE, 16.100
2D Inviscid Compressible Aerodynamics 16.100
Modular structure for 16.101x including pre-requisite modules (shown in light blue). Also shown is the overlap of the modules with the on-campus undergraduate subjects of Unified Engineering (UE) and 16.100 Aerodynamics.
have the most overlap with 16.100, while the basic gas dynamics and control volume analysis in Unified Engineering was a prerequisite to 16.101x. Each module contained: »» Typeset notes for reading with a few short tablet-based videos highlighting specific parts of the content. »» Embedded questions in the notes to help students assess their understanding of the reading. A solution video was provided for each question. »» Sample problems demonstrating the module’s measurable outcomes. A solution video was provided for each problem. »» Homework problems to assess students’ performance on the modules’ measurable outcomes. A solution video was provided for each problem, though not released to students until after the due date for the module. »» Discussion forum where students could communicate with each other and the instructional staff about the material.
24
AEROASTRO 2013-2014
3D Potential Flow Aerodynamic Models UE, 16.100
This portion of the notes includes the derivation of the Bernoulli equation and is labeled with the corresponding Measurable Outcome (MO 5.2). This MO box can be clicked to reveal the Measurable Outcome Index for that outcome.
This embedded question is placed in the notes and helps students to determine if they have understood how to apply the Bernoulli equation (MO 5.2) and its relationship to the pressure coefficient (MO 5.3). The solution video for this embedded question can be found at http://bit.ly/1jVgBlA
The videos were, on average, about 10 minutes in length and almost entirely tablet-based in which the writing and voice of the instructor could be observed. We believe that these tabletbased videos are a more personable format compared to recordings of classroom lectures. In many ways, it is like having the instructor sitting next to you and solving a problem. In total, 174 tablet videos were created for approximately 30 hours of content.
AeroAstro launches Intro to Aerodynamics on the web
25
“I am extremely grateful I got the opportunity to take this course, as I believe its a small but significant step in my dream of becoming an aerospace engineer.” Online comment, 16.101x student, India, age 23
Beyond using the outcomes in the design of 16.101x, every part of the 16.101x content (notes, embedded questions, sample problems, and homework problems) was labeled with a clickable button for each relevant outcome. These buttons when clicked would lead to a measurable outcome index revealing all of the content related to that specific outcome.
The Measurable Outcome (MO) Index allowed students to access the material through the outcomes. The index entry for MO 5.2 on the derivation and application of Bernoulli’s equation provides links to the three sections of the 16.101x content where students can go to learn about the outcome, and the five problems where students can be assessed on the outcome.
UNDERSTAND THE IMPACT 16.101x was available on edX for registration on May 22, 2013. By the first day of class on September 26, the enrollment had climbed to 23,912 students. While this number was huge, we knew that
26
AEROASTRO 2013-2014
typically only about 5% of enrollees end up being certified as successfully completing a course. In our case, we observed the following participation: »» 23,912: enrollment at launch »» 5,858: attempted first problem on first module »» 1,510: attempted first problem on second module »» 623: attempted first problem on midterm exam »» 411: attempted first problem on final exam »» 288: certificates awarded To receive a certificate for 16.101x, students had to correctly complete 70% of course content, which was weighted as 15% for embedded questions, 15% for homework problems, and 35% for each exam. Thus, based on the number of students attempting the final exam, we had about 400 students participate throughout the entire subject, and of those about 70% were awarded a certificate for passing 16.101x. For on-campus comparison, about 40 students take 16.100 each year. To help understand the effectiveness of the 16.101x MOOC, a subject evaluation survey was administered with 272 or more students responded to each question. As shown in the figures, at least 44% of students gave the highest effectiveness (a rating of 7) to the notes, embedded questions, sample problems, and videos. The discussion forum was only given the highest rating by 24% of respondents, but overall was still effective with 66% of all respondents giving a rating above neutral. Beyond questions about specific aspects of 16.101x, we also asked how likely would they recommend 16.101x to a friend on a scale of 1 (lowest) to -10 (highest). Again, we observe the largest response rate (42%) for the highest rating with 85% being above neutral. In addition to the effectiveness of 16.101x, the average time spent per week was reported to be 8.7 hours.
“A really well thought-out and balanced course: easily the best I’ve come across. You’ve set quite a paradigm for theeDx concept.” Online comment, 16.101X student, UK, age 67
AeroAstro launches Intro to Aerodynamics on the web
27
The notes contributed to my ability to achieve the measurable outcomes. 50%
The questions embedded in the notes contributed to my ability to achieve the measurable outcomes. 44%
40%
40%
28%
30%
20%
8%
10%
0%
1%
1 (Strongly disagree)
2
4 (Neutral)
5
6
7 (Strongly agree)
The sample problems contributed to my ability to achieve the measurable outcomes. 50%
0%
0%
0%
1%
1 (Strongly disagree)
2
3
4 (Neutral)
44%
25%
20%
24% 20%
10%
11% 1%
1 (Strongly disagree)
1% 2
5%
5% 3
4 (Neutral)
5
6
7 (Strongly agree)
21%
21%
5
6
15%
10%
10%
6
25% 20%
30%
5
The discussion forum contributed to my ability to achieve the measurable outcomes.
40%
0%
13% 7%
10%
3% 3
30%
30%
16%
20%
0%
48%
50%
7 (Strongly agree)
0%
7% 4%
3%
0 2 (Strongly disagree)
3
4 (Neutral)
7 (Strongly agree)
The course videos contributed to my ability to achieve the measurable outcomes.
On a scale of 1-10, how likely are you to recommend 16101x to a friend?
50%
50%
50%
42%
40%
40%
30%
30%
20%
20% 10%
2% 0%
20%
15%
1 (Strongly disagree)
1% 2
4% 3
8%
10%
15% 7%
2% 1%
4 (Neutral)
5
6
7 (Strongly agree)
0%
0 1 (Not at all likely)
1%
2%
2%
2
3
4
5 (Neutral)
17%
10% 3% 6
7
8
9 10 (Extremely likely)
Near the end of the course, students were asked about the effectiveness of 16.101x and if they would recommend 16.101x to their friends. More than 270 students responded to each question.
28
AEROASTRO 2013-2014
EMPLOYING THE MOOC ON CAMPUS A major focus of our ongoing work is determining how to leverage this material for our on-campus subjects. This past spring, we employed the 16.101x content in the aerodynamics portion of Unified Engineering in combination with employing a flipped classroom model. A flipped classroom is one in which students are required to read and do some homework on material prior to discussing it in class. We have had previous success with the flipped classroom model in both 16.100 Aerodynamics and 16.90 Computational Methods in Aerospace Engineering. However, while the flipped classroom as implemented this spring in Unified received mixed reviews (both strong positives and “The last 5 months were strong negatives) from students, the online material was challenging to me, but the way favorably reviewed. It’s still early in our consideration of using MITx resources for on-campus classes and we have you presented your knowledge generally found that educational innovation often takes a to me, including the embedded few attempts to achieve a high level of effectiveness. Beyond the on-campus efforts, we will likely offer 16.101x again in the next year or so. Also, 16.110x, a graduate version of aerodynamics, has already been offered by AeroAstro this spring. Further, we expect other new online subjects to be offered by the Department in the future with the goals of not only improving our on-campus education, but also having an impact on the world outside the Dome.
questions, sample problems, homework, videos and the exams, helped me a lot.”
Online comment, 16.101x student, Germany, age 24
DAVID DARMOFAL is the Bisplinghoff Fellow and Professor of Aeronautics and Astronautics at MIT. His interests include engineering education and computational fluid dynamics for internal and external flows. Professor Darmofal notes that the development and offering of 16.101x was a team effort that included the work of Dr. Chad Lieberman, co-instructor and creator of the LaTeX-based development process for 16.101x; Dr. Alejandra Uranga, co-instructor; Savithru Jayasinghe, 16.101x graduate teaching assistant for 16.101x; and a team of undergraduate forum moderators/beta-testers, most notably Alexander Feldstein, Elizabeth Qian, and Jacobi Vaughn. He also gratefully acknowledges the support of MITx. AeroAstro launches Intro to Aerodynamics on the web
29
Janet C. Wolfenbarger (AeroAstro SM ’85) is pinned with her fourth star by daughter Callie; husband, retired Air Force Col. Craig Wolfenbarger; and parents Eldon and Shirley Libby during a promotion ceremony June 5, 2012, at the National Museum of the United States Air Force. (US Air Force photograph)
30
AEROASTRO 2013-2014
AEROASTRO ALUMNA INTERVIEW:
General Janet Wolfenbarger rates four stars from the Air Force and AeroAstro Janet Wolfenbarger was commissioned at Eglin Air Force Base in Florida upon her 1980 graduation from the United States Air Force Academy with a BSc in engineering sciences. She furthered her education with MSc degrees in aeronautics and astronautics from MIT in 1985 and national resource strategy from Industrial
Academy because I couldn’t decide which engineering discipline I liked best. After serving in the Air Force for a few years and being afforded the opportunity to attend MIT to earn my master’s degree, I knew I wanted to specialize in aeronautics and astronautics engineering to best align with my career in the Air Force.
College of the Armed Forces in 1994. She is also a graduate of the Air Command and Staff College. Her awards and decorations include the Legion of Merit and the Meritorious Service Medal. On February 6, 2012, President Obama nominated Wolfenbarger to become the Air Force’s first woman four-star general. She assumed her new rank and became Commander, Air Force Materiel Command on June 5, 2012.
AeroAstro: When did you first know you were interested in (aerospace) engineering? Wolfenbarger: I discovered early in my academic career that I especially liked math. I thought being an engineer would be a good way to use my favorite subject on the job. I earned a Bachelor of Science degree in engineering sciences at the U.S. Air Force
AeroAstro: What things particularly stand out about your time at MIT and AeroAstro? Wolfenbarger: One of the things I am most thankful for from my time at MIT was experiencing an entirely different approach to teaching and learning than I had experienced previously. That learning technique is something that has stayed with me throughout my career. I’ve been able to take the lessons from my two engineering degrees and apply them in every job I’ve had as an Air Force officer. I did so either directly when I had the opportunity to work on technical projects or indirectly when I applied the rigor of the engineering approach to non-technical efforts.
Alumna interview: General Janet Wolfenbarger rates four stars from the Air Force and AeroAstro
31
AeroAstro: What have you done since leaving MIT? Wolfenbarger: I have served in the acquisition business for most of my almost 34 years in the Air Force. I spent time on the leading edge fighter, bomber, and transport aircraft programs in the Air Force. This included eight-plus years on the F/A-22 program, five-plus years on the B-2 Bomber, including time as the Director of the B-2 program, and two-and-a-half years as the Director of the C-17 program. I served in three assignments at the Pentagon and at the Major Command Staff level two times. In June 2012, I was promoted to the rank of General Officer, the first woman in the Air Force to attain this rank, and given the chance to command Air Force Materiel Command — a job my Air Force career and my education has prepared me for over the course of the last three decades.
AeroAstro: Tell us more about your current job. Wolfenbarger: As the commander of Air Force Materiel Command, the Air Force is counting on me, and more importantly the almost 80,000 men and women of my command, to equip our Air Force for world-dominant airpower. AFMC has a very important mission: supporting the warfighter. We do this through our four mission areas to include science and technology, life cycle management, test and evaluation, and sustainment. I like to summarize what we do by saying, “When the warfighter calls for a new capability, we think it, we build it, we break it, we make it better, we deliver it to the fight, and we keep it there as long as it’s needed.” By far, the biggest challenge AFMC faces today is providing required support to the warfighter in an environment where money is tight. I contend that while the budget environment is
32
AEROASTRO 2013-2014
challenging, it also can and should be embraced as an opportunity to figure out ways to accomplish our missions more efficiently and more effectively. I have never seen the aperture more wide open to have good ideas get a fair hearing. AFMC continues to efficiently and effectively accomplish its mission, while also working to preserve the welfare of the command’s people. Our mission is as serious today as it ever has been, and we’re committed to doing everything we can to make every defense dollar count and directly support the warfighter. As one of only 11 active-duty four-star generals, I also work handin-hand with the other senior leaders in our Air Force to resolve the toughest issues facing our military and our nation. As the first female four-star, I also understand that others look to me as a role model, and I have a responsibility to demonstrate — via speeches and other engagement opportunities — that our Air Force has embraced the value of diversity in our institution.
AeroAstro: What advice would you offer to high school students about considering engineering careers? Wolfenbarger: Certainly, the Air Force has embraced the importance of STEM, or Science, Technology, Engineering and Mathematics. STEM applies directly to virtually every career field. From auto mechanics to aircraft engineers, rocket scientists to firefighters, everyone in the workforce is touched somehow by STEM. Participating in STEM coursework throughout high school and college ensures students are well-prepared to enter the workforce in the future. I will also say that these fields are absolutely critical to the execution of the Air Force mission. We require great depth and great
Gen. Wolfenbarger chats with students following an AeroAstro-sponsored talk she gave in April 2014. (William Litant photograph)
skill sets across the engineering and scientific career fields. The success of the Air Force depends on this continued innovation and technical excellence.
AeroAstro: What advice would you offer to current AeroAstro students to best position themselves for their careers? Wolfenbarger: My advice to anyone — men, women, Airmen, or college students preparing for the work force — is to do the
very best you can at whatever task is given to you, and always do it with a positive attitude. My recipe for success has been that simple. I have spent my career working hard and doing the very best I could in every job that was given to me, and I can give no better advice than that. It has served me well. The lessons learned at MIT, along with hard work, rigor, and curiosity, will guide your way as they have helped me. When you are attending MIT, you are standing on the shoulders of giants. And you can achieve anything — it’s up to you.
Alumna interview: General Janet Wolfenbarger rates four stars from the Air Force and AeroAstro
33
546th Propulsion Maintenance Squadron supervisor Debbie Patterson explains the F108 engine line’s kitting cart system to Gen. Wolfenbarger at Tinker Air Force Base, Okla. (U.S. Air Force photograph)
AeroAstro: Can you share some thoughts on being the Air Force’s first female four-star and major command commander? Wolfenbarger: When I’m asked this question, I always respond with the same answer — I am humbled, I am honored, and I am excited to lead such a great command in the world’s premier Air Force. I spent the majority of my career in AFMC, and I can’t tell you how special it is to have the opportunity to command the major command I grew up in. I never anticipated that my career
34
AEROASTRO 2013-2014
would include a promotion to brigadier general, much less this opportunity to serve at the highest rank in the Air Force. It’s our great Air Force that gave me this opportunity, because as a service it has embraced a culture of diversity. It’s a culture that has been cultivated throughout many years, driven by leaders at every level who acknowledge and appreciate the value of contributions from every Airman. In the last three-plus decades, we’ve made great progress but certainly our work doesn’t stop there. Diversity has made our nation and our Air Force stronger, and it will continue to do so.
CENTENNIAL CELEBRATION 100 years of MIT Aeronautics and Astronautics
Myriad initial design considerations have far-reaching consequences
35
Listing for MIT’s first aeronautics class, from the 1913-1914 Course Catalogue. The class was under Course 13, the Department of Naval Architecture. Course 16, the Department of Aeronautics, was created in 1926. (MIT Archives)
36
AEROASTRO 2013-2014
A CENTENNIAL ALBUM
Images from 100 years of MIT aeronautics and astronautics By William Litant
“President Maclaurin stated that he believes (MIT) should lead in the study of aërial navigation in the United States.” Technology Review, July 1909
With that statement, MIT President Richard Maclaurin set the stage for what, in 1914, would become Class 13.72 Aeronautics. That same year, an aeronautics master’s degree course was approved. In 1926, Aeronautics became a course (Course 16) in the Mechanical Engineering Department. In 1939, Course 16 became a department unto itself. Over the ensuing years, Aeronautics and Astronautics alumni have led the engineering programs of most U.S. aerospace corporations and served as astronauts, presidential advisors, NASA administrators, and US Air Force Secretaries and chief scientists. Department milestones range from foundational research unpinning the current air transportation system to creating the technology that made lunar exploration possible. We are a vibrant department focused on aerospace vehicle and information engineering, and the engineering of large-scale complex aerospace systems. Through our educational programs and research we are advancing the stateof-the-art in transportation, exploration, communication, national security, and energy and the environment. Throughout 2014 and into 2015, the MIT Department of Aeronautics is celebrating 100 years of educational achievement, technological accomplishments, and remarkable students, faculty, and staff. Following are a few highlights from our first 100 years. For more on the Department’s history, visit http://aeroastro.mit.edu/about-aeroastro/brief-history.
Images from 100 years of MIT aeronautics and astronautics
37
1914 MIT establishes the nation’s first aeronautical engineering course: Aeronautics for Naval Constructors. Subjects include theoretical hydrodynamics, applied hydrodynamics, wind tunnel laboratory, airship theory, aerial propellers, and theory and practice of airplane design. The course was co-designed and is directed by Jerome Hunsaker (SM ’12, SCD ’23). The same year, Hunsaker, assisted by Donald Douglas (Course 2, SB ’14), builds the first structure on MIT’s new Cambridge campus — a wind tunnel, located on Vassar Street.
Military aviators learn the basics of rigging (adjusting wing and tail surfaces) during a World War One MIT class. (MIT Museum)
1918 America is involved in the World War. Aeronautical Engineering is transferred from Naval Architecture to the Department of Physics. Army and Navy officers are sent to MIT for advanced aeronautical training and special ground school classes are created for military aviators.
1932 Isabel Ebel, the only woman studying aero engineering among MIT’s student body of 30 women and 3,000 men, becomes the first woman to receive a degree from Course 16. She is unable to find a job until 1939 when Grumman hires her.
1938 The Wright Brothers Wind Tunnel is dedicated. It features a 13’ diameter variable-pitch fan driven by a 2,000 horsepower induction motor, can be pressurized, and has a top design speed of 400 mph. Its first dozen years include extensive World War Two design development testing by companies like Sikorsky, Grumman, Republic, Consolidated Vultee and Chance Vought. Over time, it will be used to test architectural designs, motorcycles, ships’ sails, wind turbines, and, of course, a myriad of aircraft designs.
1953 A U.S. Air Force B-29 flies from Bedford, Massachusetts to Los Angeles guided by Charles Stark “Doc” Draper’s Space Inertial Reference Equipment, the forerunner of today’s autopilot systems. SPIRE did its job with no information from the outside world other than the initial coordinates at the Bedford airstrip. Only once during the 12.5-hour 2,250-mile flight did the pilot have to touch the controls. This was the first long-distance inertially navigated aircraft flight. 1961 NASA selects the Instrumentation Lab to develop Apollo’s guidance, control, and computer systems. Alumn and professor Bob Seamans is NASA deputy administrator. Eight years later, alumn Buzz Aldrin is the second man to walk on the moon.
38
AEROASTRO 2013-2014
MIT’s 1914 wind tunnel was a four feet square open-circuit design with a flow of 30 miles per hour. The tunnel’s balance (the device that measures aerodynamic loads during testing) is in the MIT Museum collection. (MIT Museum)
The iconic Wright Brothers Wind Tunnel. Now in its fourth quarter-century, researchers are using it to design aircraft for 2030 and beyond. (William Litant/MIT)
1963-1965 MIT aerospace research accelerates with the creation of the Space Propulsion Lab, the Man Vehicle Lab, the Center for Space Research, and the Flight Transportation Lab, a predecessor of today’s International Center for Air Transportation.
1970s-1980s Two aerospace firsts in the late 1970s and early 1980s involve MIT alumni. A. Thomas Young (ScD ’72) directs NASA’s Viking I and II 1976 missions to Mars. Orbital flights of the Space Shuttle begin in 1981 led by James A. Abrahamson (SB ’55).
An MIT Instrumentation Lab engineer checks Apollo Command Module onboard guidance computer programs in a special simulator that can run complete missions to test program accuracy. (Charles Stark Draper Historical Collection, MIT Museum)
Doc Draper shows the SPIRE autopilot to CBS newsman Eric Sevareid. (MIT Museum)
Images from 100 years of MIT aeronautics and astronautics
39
Space Shuttle Columbia lifts off April 12, 1981 under the direction of NASA associate administrator and AeroAstro alumn James Abrahamson (SB ‘55). (NASA)
The Man Vehicle Lab, founded in 1962, prepares Spacelab experiments to measure changes in astronaut balance. Crew member and MIT/MVL alumnus Byron Lichenberg (SM ’75, ScD ’79) is shown during on the lab’s “vestibular sled.” Instruments inside the helmet measure eye movements and movement perception. (MVL)
Daedalus undergoing tests at NASA’s Dryden Flight Research Center prior to its record-setting flight.
40
AEROASTRO 2013-2014
Alumna-astronaut Janice Voss (PhD ‘87). Between 1993 and 2000 she flew as a mission specialist on five Shuttle missions. (NASA)
The Beaver Works team with two of its FAST aircraft. To the left is FAST A, offering dash to objective, long flight endurance, and a single large payload capacity. To the Right is FAST B, which is backpackable, hand-launched, and carries a single small payload. The aircraft were built using common molding and tooling.
The 2001 renovation of Building 33 included construction of the 6,000 s.f. Gerhard Neumann Hangar as a space for AeroAstro to work on large projects. It also houses two small wind tunnels. (Cambridge Seven Associates) AeroAstro alumni astronauts (from left) Greg Chamitoff (PhD ‘92) and Mike Fincke (SB ‘89), and space tourist Richard Garriott, aboard the International Space Station in 2008 with three SPHERES microsatellites satellites. (NASA)
1988 Lead by John Langford (AeroAstro SB ’79, SM ’85, PhD ’87) a team of 40 MIT students, faculty, and alumni has designed and built, and now sets a world record for human-powered flight with the aircraft Daedalus 88. Piloted by cycling champion Kanellos Kanellopoulos, Daedalus flies 71.5 miles (115.11 km) in 3 hours, 54 minutes from Crete to Santorini. 1990s Virtually every major aerospace achievement in the 1990s has an MIT connection. Kent Kresa (AeroAstro SB ’59, SM ’61, EAA ’66), Northrop Grumman president and CEO, leads B-2 stealth bomber development. Janice Voss, (EECS SM ’77, AeroAstro PhD ’87) flies on the shuttle rendezvousing with Mir, the first joint US-Russian mission in 20 years. Alan Mulally, (Sloan SM ‘68) leads Boeing’s 777 development team and Tom Imrich (AeroAstro SB ’69, SM ’71) is Boeing’s chief research pilot. MIT engineers design the Mars Pathfinder laser altimeter.
2001 The historic 1928 Daniel Guggenheim Aeronautical Laboratory, MIT Building 33, is renovated, creating 50,000 s.f. of flexible and open space that allows students, faculty, and staff to work in modern team environments on projects of varying size and complexity. The program incorporates renovation and new space for the Arthur and Linda Gelb Laboratory, the Robert C. Seamans Jr. Laboratory, and the Gerhard Neuman Hangar. The spaces are designed to facilitate the department’s landmark new ConceiveDesign-Implement-Operate approach to engineering education.
2006 SPHERES microsatellites, conceived and constructed by undergraduates in AeroAstro’s capstone class, are delivered to the International Space Station where they will serve as test beds for spacecraft autonomous rendezvous and docking maneuvers.
2014 With MIT and Lincoln Laboratory’s Beaver Works collaboration, AeroAstro undergrads design and build projects for real-world customers. Their most recent project is FAST, the Flexible System Aircraft Testbed, funded by the Defense Advanced Research Projects Agency (DARPA), which offers the ability to produce diverse unmanned aerial vehicles from common molding and tooling architecture.
Images from 100 years of MIT aeronautics and astronautics
41
The June 15, 1916 banquet at Symphony Hall celebrating the opening of MIT’s new Cambridge campus. Seated at one table was a remarkable assemblage of aviation pioneers including Alexander Graham Bell, Orville Wright, Glenn H. Curtiss, Glenn L. Martin, Donald Douglas, and others. (MIT Museum)
42
AEROASTRO 2013-2014
The most remarkable banquet in the world By Deborah Douglas
A unique assemblage of aero luminaries sets the stage for a century of MIT achievement
Remarkable, indeed, for the 1,500 MIT alumni and guests (the Post claimed 3,000) that had squished into Symphony Hall the previous evening, while thousands of others listened across the United States by telephone. The newspapers noted the presence at MIT’s now famous “Telephone Banquet” of prominent figures in Massachusetts politics and education, accompanied by luminaries and aviation giants Alexander Graham Bell and Orville Wright. It was the capstone event of a multi-day extravaganza marking the Institute’s triumphant move from Boston to Cambridge. For most banquet attendees, dining elbow-to-elbow on medallion of Penobscot salmon, larded filet of beef, sweetbreads glace, and roast jumbo squab, the past two days had marked their first glimpse of the Institute’s spectacular new campus, and the buzz was all about the “The most remarkable banquet in the world” is the caption for a sketch that appeared on the front page announcement of the nearly $2.7 million dollars that had been raised of June 16, 1916 Boston Post. just that very day (about $58 million in 2014 dollars). The boisterous cheers and songs of those present were punctuated with greetings from alumni clubs in 34 cities, linked to Symphony Hall by a unique intercity telephone network created specifically for the event. Professor Robert Rogers could hardly contain himself when writing a few weeks later for Technology Review. It was, he said, a “magnificent and orgulous celebration.”
The most remarkable banquet in the world
43
A FASCINATING DISCOVERY
The first public display of the Wright Brothers’ 1903 flyer was at MIT’s new Cambridge campus as a highlight of an open house event held during the June 1916 dedication festivities. Location was Building 3. The open space was later floored over to create the Bioinstrumentation Lab (first floor,) the Hatsopoulos Microfluidics Lab (second floor,) and the Experimental Nonlinear Dynamics Lab (third floor) (MIT Museum)
Reading accounts of this dinner closely, AeroAstro alumnus John Tylko (SB ’79) recently made a fascinating discovery. In the papers of Jerome Hunsaker, architect of MIT’s first aeronautics course, Tylko discovered that the dinner table where Alexander Graham Bell and Orville Wright were seated, also included Glenn H. Curtiss, Glenn L. Martin, Admiral Washington L. Capps, Lester D. Gardner, Godfrey L. Cabot, Paul Litchfield, James Means, Donald Douglas, and MIT professors Cecil Hobart Peabody, William Hovgaard, Joseph C. Riley, and Edwin Bidwell Wilson. You would be hard-pressed in 1916 to propose a more impressive and influential dinner party in American aviation. Can you imagine the conversations at dinner that night? The atmosphere must have been crackling (and chaotic!) and filled with optimism. This may have the first such gathering of aero luminaries at MIT, but it would hardly be the last.
The formal course in aeronautical engineering, “Course for Naval Constructors, Aeronautics,” (what would later become Course 16) was barely into its second year, but already its influence was being felt. Richard C. Maclaurin, who became MIT president in 1909 with a mandate to build a new campus and transform MIT into a modern research university, was especially enthusiastic about this fledgling academic program. On the subject of aeronautics, Maclaurin had turned to his old friend and colleague Richard Glazebrook, director of the National Physical Laboratory, for advice. Glazebrook was very encouraging, and, back in Cambridge, Maclaurin began talks with Department of Naval Architecture head Professor Cecil Peabody (also enthusiastic) about starting a program under the aegis of his department, Course 13.
44
AEROASTRO 2013-2014
MIT students had already undertaken their own rudimentary experiments. Beginning in 1895, and continuing through 1911, several engineering undergraduates made use of an ingenious “wind tunnel” to conduct their senior thesis research. A duct was built off an Institute building’s forcedair ventilation system, and several years of dedicated efforts enabled the students to replicate Samuel P. Langley’s data and theories for low velocity flows. In the summer of 1909, MIT students had worked with Orville Wright on the 1909 flyer, and, in Nova Scotia, they worked under the tutelage of Alexander Graham Bell at his Aëronautical Experimental Station. That fall, they founded the Tech Aero Club and immediately began building a glider, and participating in major aerial expositions. Also in 1909, a young Navy midshipman, Jerome Hunsaker, arrived at MIT with a focused enthusiasm on maritime vessel Map depicting the elaborate telephone circuit allowing thousands of MIT alumni to participate in the closing event of design. Until, that is, Hunsaker witnessed the first Bostonthe 1916 Dedication Ceremony. (Institute Archives and Special Harvard Aero Meet in September 1910. A sense of wonder about Collections, MIT) how flying machines worked replaced earlier dismissals. At dinner with MIT President Maclaurin, Hunsaker became a real convert as Maclaurin was genuinely convinced this was an important new technology. Hunsaker soon began to study the technical literature, especially several untranslated works by French engineers, including Gustav Eiffel’s pioneering work on aerodynamics, La Résistance de L’Air, which he translated and arranged to be published by Houghton Mifflin.
THE NEW FIELD OF AERONAUTICS Hunsaker took some of the first classes offered by MIT in the new field of aeronautics and quickly emerged as one of the brightest aeronautical technical rising stars. So bright and so able that Maclaurin decided to groom Hunsaker for bigger things. In the summer of 1913, Maclaurin, who was a member of National Aerodynamical Laboratory Commission, arranged for Hunsaker to
The most remarkable banquet in the world
45
accompany Catholic University Professor Albert Zahm who was being sent to survey the top European research laboratories and universities with the objective of advising the U.S. government on how best to support the technical development of aviation. Hunsaker spent time in London, Paris and Germany meeting the world’s leading figures and sizing up the latest technologies. He worked for a few weeks in France under Leonard Bairstow of the aerodynamics group at the National Physical Laboratory as well as with Eiffel, in both instances carefully studying their equipment and research techniques. That fall, he began work at MIT under special assignment from the Navy. His goal: create one of the world’s best laboratories for aerodynamics research, and start the nation’s first program offering an aeronautical engineering degree. Hunsaker enlisted the aid of an exceptional MIT undergraduate, Donald Douglas. Douglas had gone to the U.S. Naval Academy in 1909. He witnessed the Wright Brothers famous demonstration to the US Army at Fort Meyer that fall, and became passionate about airplanes. Douglas transferred to MIT in 1912 when he realized that naval aviation was a thing of the future. Though Douglas received his SB degree in mechanical engineering (Course 2) he was quickly in the thick of aeronautics research as, in the spring of 1914, Hunsaker teamed him up with MIT undergraduate (and future aeronautics pioneer) Edward P. Warner and tasked them with helping him
46
AEROASTRO 2013-2014
Interior (far left) and exterior (left) of the 1914 wind tunnel. Erected on Vassar Street, it was MIT’s first facility in Cambridge, predating dedication of the new campus by two years. The tunnel’s balance (measuring device) is in the MIT Museum collection. (MIT Museum)
build MIT’s first “real” wind tunnel. The trio had a deadline: classes for the new aeronautical engineering curriculum would begin that fall and the new MIT Aerodynamics Laboratory had to be ready.
“WHY MIT?” Hunsaker’s tunnel and the promise of “systematic instruction in the theory and design of aeroplanes” attracted several star students. In December 1914, Technology Review reported that Captain Virginius E. Clark, a rising figure in military aviation had matriculated. “There is no other college or institution in the country that is fitted for scientific investigation of practical matters in aerodynamics,” replied Clark when asked, “Why MIT?” With the start of war in Europe that fall, there was a major new external imperative driving interest in aeronautical engineering. Though the United States would not enter the conflict for another three years, actions by the federal government created a market demand for professionally educated engineers. First, the military had decided to use aircraft in significant numbers as part of any future war effort. Second, the establishment of the Army’s McCook Field research facility and the National Advisory Committee for Aeronautics (NACA) decision to build its own research laboratory created employment opportunities. “There is no other college or instituNot only did these labs hire college-educated engineers, they tion in the country that is fitted for issued contracts to universities to conduct research.
scientific investigation of practical
All of this must have been the talk of the table at Symphony matters in aerodynamics.” Hall that evening in June 1916. It is unlikely that anyone Captain Virginius E. Clark present that evening could have predicted the astonishing transformation that would occur nine months later. Just after the banquet, the Navy ordered Hunsaker back to Washington where, under Rear Admiral David Taylor, he would establish the Bureau of Aeronautics. Alexander Klemin, a newly minted graduate, was asked to step in, but when the war started, he was recruited to become the “officerin-charge” of the Research Department at McCook Field. (He would then move to head the new aeronautical engineering program at New York University, write a key early textbook and become
The most remarkable banquet in the world
47
(Left) Edward P. Warner (SB ‘17, SM ‘19) was likely in line to become head of MIT aeronautics but, like others in fledgling MIT aeronautics, he was called to serve Army aviation, and then was named the NACA’s chief physicist in 1919. He returned as a professor of aeronautics in 1920, teaching until his 1926 appointment as Assistant Secretary of the Navy for Aeronautics. (MIT Museum) (Right) Donald Douglas, (Course 2 SB ’14) was recruited by Jerome Hunsaker to teach aeronautics. The two, joined by Edward Warner, built MIT’s first true wind tunnel in 1914. (MIT Museum)
technical editor of Aviation magazine.) Edward Warner, a 1917 graduate, would have replaced Klemin but he too was called to serve Army aviation, and then was named the NACA’s chief physicist in 1919. (Warner returned as a professor of aeronautics in 1920, teaching until his 1926 appointment as assistant secretary of the Navy for aeronautics.) Between 1917 and 1919 this pattern replicated itself again and again as federal appropriations grew from hundreds of thousands to hundreds of millions of dollars. MIT graduates were in great demand, in part because the needs were extensive, but also because of the emerging network of graduates and faculty. The connections among MIT, the Navy, and the NACA were especially strong, not surprisingly because of Hunsaker. MIT graduates in aeronautical engineering (and in all of the other disciplines!) turned to each other for contacts and jobs. They especially trusted each other on technical matters. And, they were an ebullient, extroverted lot manifesting an extraordinary esprit des corps. They wrote articles and textbooks, edited journals, taught classes and founded departments. They designed airplanes and engines, and spent countless hours in wind tunnels and laboratories. They took part in epic flights, and served with distinction in every branch of government. They
48
AEROASTRO 2013-2014
founded companies and the premier engineering society, the Institute of Aeronautical Sciences (today’s American Institute of Aeronautics and Astronautics). They won Collier Trophies and Guggenheim medals, and received countless citations. When things did not go well, they stood by each other. The collective achievements of this first generation are staggering. Collective achievement is more than a descriptive phrase. It is a hallmark of the Department to today. No one attending the 1916 banquet would have used phrases like “intellectual capital” or “social capital,” and concepts like “branding” or “identity” would have elicited quizzical looks. Yet, that group clearly understood that to make MIT a leading center or education and research required a profound and sustained commitment to a common cause. Absent that esprit, it could have easily fallen apart. History has many examples of this and there are many great aeronautical engineering departments founded at this time that no longer exist. The terrible war in Europe made this group more resolute than they might have been otherwise. But the bonds that had already formed among this group, amplified by the conviviality of the evening and the inspiration of the Dedication ceremonies, were also crucial. That combination of circumstance and disposition would provide unprecedented opportunities — opportunities that none hesitated to take advantage of — and a legacy that would inspire a century of achievement. DEBORAH DOUGLAS is the MIT Museum’s Director of Collections; and Curator, Science Technology. She may be reached at ddouglas@mit.edu. This short essay has drawn extensively the excellent research of John Tylko, William Trimble, Earl Thornton, and Tom Crouch. The author has drawn on the archival collections of the MIT Museum and the Institute Archives as well as her previously published essay, “The End of ‘Try-and-Fly’: The Origins and Evolution of American Aeronautical Engineering Education through World War II” in Engineering in a Land-Grant Context: The Past, Present, and Future of an Idea (Purdue, 2005).
The most remarkable banquet in the world
49
Charles “Doc” Stark Draper aboard the Instrumentation Lab-designed Command Module guidance and navigation simulator on which the Apollo astronauts trained. On August 1, 1961, the lab was notified it had been selected to develop Apollo’s guidance and navigation system — the Apollo Program’s first contract award. In 1973, the Instrumentation Lab was spun off as the independent Charles Stark Draper Laboratory (MIT Museum, Charles Stark Draper collection)
50
AEROASTRO 2013-2014
DOC’S LABORATORY
80 years of Draper Lab innovation began with MIT’s Instrumentation Lab by Kathleen Granchelli
As the MIT Aeronautics and Astronautics Department marks its 100th anniversary, Draper Laboratory joins in the celebration of the department’s many achievements over the years, reflects on our common heritage, and looks forward to our continuing partnerships in educating students while supporting faculty in joint research projects.
The MIT Aeronautics and Astronautics Department’s 2014 centennial immediately follows a year in which Draper Laboratory celebrated its founding 80 years ago by Dr. Charles Stark Draper as the MIT Instrumentation Laboratory. Following 40 years as part of AeroAstro, Draper was spun off in 1973 as an independent, not-for-profit organization chartered to do research and development in the national interest and to support advanced technical education. The chronicle of eight decades of engineering achievements begins with Doc’s early days at MIT, where his interests in improving aircraft instrumentation led to the development of applications for inertial sensors. Among his pioneering achievements were the development of the Mark 14 gyro-stabilized gunsight for shipboard anti-aircraft guns used in the Pacific Theater during World War II and the first transcontinental aircraft flight navigated solely with an inertial navigation system.
Doc’s monikers — “the father of inertial navigation” and “Mr. Gyro” — were as famous as his MIT Instrumentation Laboratory, which continued to develop technical capabilities for new national defense needs and human space exploration following the early work on the Mark 14. Beginning in the late 1950s, the Lab began a relationship with the U.S. Navy to design and develop the strategic guidance system for the Fleet Ballistic Missile Program that has lasted nearly 60
80 years of Draper Lab innovation began with MIT’s Instrumentation Lab
51
years, and has plans that reach out to 2080. Draper has designed and developed the boost guidance system for every submarine-launched ballistic missile deployed by the U.S., and has had an intellectual leadership role with the guidance systems for intercontinental ballistic missiles. The Lab was tapped in the same era by the new national space agency, NASA, to develop the guidance system for the Apollo Moon Program. Draper provided essential guidance and control capability for every U.S. human space flight program since Apollo. For Apollo, not only did the Lab develop the inertial guidance system, but it also designed the flight computer and all the software for the orbiter and lander, which required creating one of the first higher-order computer languages and the concept of interrupt-driver, real-time software. Since divesture from MIT, the independent Laboratory’s technical expertise has expanded in step with the needs of the nation while we maintained an active collaboration with AeroAstro. Based on its success with digital controls on the Apollo guidance computer, the Laboratory collaborated with NASA on the Based on its success with digital first digital fly-by-wire aircraft control systems.
controls on the Apollo guidance computer, the Laboratory collaborated with NASA on the first digital fly-by-wire aircraft control systems.
Beginning 30 years ago, Draper transitioned micro-electromechanical systems (MEMS) from the Lab to practical applications, first with solid-state gyroscopes and accelerometers. Later, the Lab used MEMS to develop close-in sensors for intelligence collection, and implantable sensors and tissue engineering for biomedical applications. From these experiences, Draper developed a significant competence in vanishingly small systems (VSS), including electronics packaging, very low-power system design, and low-rate production. Future development of VSS technology will include a heavy emphasis on attaining smaller (nanoscale) features in existing sensors and processes, as well as developing novel, specialized polymeric and piezoelectric materials. MEMS and VSS will continue to be foundational technologies for the Laboratory, supporting continued advances in new sensor and system applications. Draper is laying the groundwork for practical guidance, navigation, and control systems that will integrate multi-axis, cold atom accelerometers, gyroscopes, and clocks—all using a single technology platform.
52
AEROASTRO 2013-2014
Aboard a U.S. Air Force B-29, Doc (right) and CBS newsman Eric Sevareid examine Draper’s Space Inertial Reference Equipment, the forerunner of today’s autopilot systems. In February 1953, the aircraft flew from Bedford, Massachusetts to Los Angeles guided by SPIRE with no information from the outside world other than the initial coordinates input at the Bedford airstrip. Only once during the 12.5-hour 2,250-mile flight did the pilot have to touch the controls. This was the first long-distance inertially navigated aircraft flight. (MIT Museum, Charles Stark Draper collection)
Additional pioneering developments over the years include the software for autonomous undersea systems and highly reliable, fault-tolerant computer architectures. With the dawn of the 21st century, Draper has strengthened its software capabilities, building from our established embedded systems skills to address issues associated with data fusion and human systems collaboration. Today, the Laboratory is working to develop new approaches including human-guided algorithms and neurocognitive monitoring and applying these techniques to problems such as real-time decision-making and to the design and testing of advanced autonomous systems.
80 years of Draper Lab innovation began with MIT’s Instrumentation Lab
53
Doc Draper demonstrates a 40-mm gun director on MIT Building 33’s front lawn. In the early days of World War II, Sperry Gyroscope contracted with the Instrumentation Laboratory to help develop the specialized instrument that would allow a Navy anti-aircraft gunner to keep up with the new fast-flying airplanes. The U.S. Navy ordered 85,000 Mark 14 Gunsights, which were later credited with altering the balance of power in favor of the United States in the Pacific conflict. (MIT Museum, Charles Stark Draper collection)
While the technology and problems have evolved over time, Draper has had a constant commitment to working on critical national issues. Drawing on its important legacy at MIT, the Laboratory attracts the best scientists and engineers who are motivated by the thrill of “making things that work,” which Doc Draper always said was the key to his lab’s success. In the early days, outstanding people were attracted by Doc’s personality and his vision for the applications of inertial sensing and instrumentation. Today, the same quality of staff come to Draper because of the challenging problems we take on, the appeal of our mission of national service, the opportunity to work with a superb set of colleagues, and the reward of helping to train the next generation of engineers.
54
AEROASTRO 2013-2014
COMMITMENT TO EDUCATION Technical education, an important element of the Laboratory’s mission, has its genesis in Doc’s motivation to offer real-world engineering opportunities for his students. The Laboratory’s unique nature is due in part to its continuing dedication to the process of education. The belief that real problems provide a valid teaching mechanism has infused the culture of our organization over its history so that education is achieved not only by students in the formal sense, but by all staff at the Laboratory in the conduct of their work. We believe that the best way to learn technology is for people to be engaged in solving real-world problems. At its 1973 spinoff, the Laboratory and MIT formally agreed to continue the education of selected graduate students seeking advanced degrees. These outstanding graduate students Former Draper Laboratory Fellow Akil Middleton, now a technical staff member, are designated Draper Laboratory Fellows and works on the Guided Airdrop program aerial guidance unit used for hardwarein-the-loop simulations. More than 1,000 graduate students, including several work on Draper projects that include chalU.S. astronauts, military flag officers, and three Draper vice presidents, have lenging research problems. Each is an active completed their degrees through the DLF program, with the majority capping their careers with leadership positions in government, DoD, industry, and member of a research team at one of Draper’s academia. (Draper Laboratory photo) facilities, conducting thesis research while supervised by a Draper technical staff member. Draper provides full tuition coverage and a monthly stipend for the DLF for the duration of his/her degree program. Typically, 30 percent of DLFs are active duty military students. The majority of DLFs have been in degree programs at MIT, with the highest percentage from AeroAstro.
80 years of Draper Lab innovation began with MIT’s Instrumentation Lab
55
Draper mission design leader Bobby Cohanim (right), and MIT student Eph Lanford work on TALARIS, a prototype planetary hopping vehicle. TALARIS was jointly developed by Draper and AeroAstro students guided by Professor Jeffrey Hoffman. (Draper Laboratory photo)
More than 1,000 graduate students have completed their degrees through the DLF program, with many of them capping their careers with leadership positions in government, Department of Defense, industry, and academia, including U.S. astronauts, military flag officers, and two Draper vice presidents. Their collective research has contributed significantly to the world’s body of knowledge while feeding a steady stream of well-educated engineers and scientists into the national reservoir of talent.
56
AEROASTRO 2013-2014
AEROASTRO COLLABORATION Draper values its ongoing collaborations with AeroAstro faculty and students, which as been supported with funding through the Laboratory’s University Research and Development Program, Internal Research and Development Program, Draper Laboratory Fellow Program, and other student internships. A multi-year collaboration between Professor Jon How’s Aerospace Control Laboratory and the Draper’s Guided Airdrop program has resulted in new situation awareness and planning technologies that will be used to support future flight tests of the Guided Airdrop program. Multiple AeroAstro graduate students, including several DLFs, have worked on this effort. With Professor Jeffrey Hoffman, Draper has worked on TALARIS, a prototype planetary hopping vehicle, which was designed, built, and tested on site at Draper. More than 100 undergraduate students and 10 DLFs contributed to this project, which resulted in a partnership on a Google Lunar X Prize team. In addition, Draper, Professor Hoffman, and students have teamed for work on the NextGenEVA program to develop an advanced human mobility space suit.
More than 100 undergraduate students and ten Draper Laboratory Fellows contributed to TALARIS, which resulted in a partnership on a Google Lunar X Prize team.
A Draper team led by Space Systems Director Seamus Tuohy has collaborated with MIT Professor Sara Seager on the development of ExoPlanetSat, a cubesat just 10 cm tall, 10 cm wide, and 30 cm long that could look for planets that orbit stars other than the sun. Draper contributed expertise to the project as well as funding of work with MIT students to design the spacecraft’s platform. Draper looks forward to our future collaborations with MIT AeroAstro in the development of technology solutions for national initiatives and to continue to educate selected students by offering real-world research opportunities on Laboratory programs. KATHLEEN GRANCHELLI is the Draper Laboratory media and community relations director. She may be reached at kgranchelli@draper.com.
80 years of Draper Lab innovation began with MIT’s Instrumentation Lab
57
LAB REPORT: A 2013-2014 review of Aeronautics and Astronautics Department Laboratories Reports provided by the research laboratories and centers
AEROSPACE COMPUTATIONAL DESIGN LABORATORY. . . . . . . . . . . . . . . . . 59 AEROSPACE CONTROLS LABORATORY.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 AEROSPACE ROBOTICS AND EMBEDDED SYSTEMS GROUP.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61 THE AUTONOMOUS SYSTEMS LABORATORY. . . . . . . . . . . . . . . . . . . . . . . . . . . . 62 COMMUNICATIONS AND NETWORKING RESEARCH GROUP. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 GAS TURBINE LABORATORY.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64 HUMANS AND AUTOMATION LABORATORY. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65 INTERNATIONAL CENTER FOR AIR TRANSPORTATION. . . . . . . . . . . . . . . . . 65
In AeroAstro’s Arthur and Linda Gelb Student Laboratory, undergraduate Libby Jones shows visitors aircraft her team has designed and built for the American Institute of Aeronautics and Astronautics’ annual Design-Build-Fly competition. The occasion for the display was the department’s April 23, 2014 open house, staged in recognition of MIT aeronautics’ centennial. (William Litant/MIT photo)
LABORATORY FOR AVIATION AND THE ENVIRONMENT. . . . . . . . . . . . . . . 66 LABORATORY FOR INFORMATION AND DECISION SYSTEMS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66 THE LEARNING LABORATORY. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 MAN VEHICLE LABORATORY. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68 NECSTLAB. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68 SPACE PROPULSION LABORATORY. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69 SPACE SYSTEMS LABORATORY. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70 SYSTEM ENGINEERING RESEARCH LAB. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72 TECHNOLOGY LABORATORY FOR ADVANCED MATERIALS AND STRUCTURES. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73 WIRELESS COMMUNICATION AND NETWORK SCIENCES GROUP. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74 WRIGHT BROTHERS WIND TUNNEL. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
58
AEROASTRO 2013-2014
AEROSPACE COMPUTATIONAL DESIGN LABORATORY The Aerospace Computational Design Laboratory’s mission is the advancement and application of computational engineering for the design, optimization, and control of aerospace and other complex systems. ACDL research addresses a comprehensive range of topics including advanced computational fluid dynamics and mechanics; uncertainty quantification; data assimilation and statistical inference; surrogate and reduced modeling; and simulation-based design techniques. Advanced simulation methods developed by ACDL researchers facilitate the understanding and prediction of physical phenomena in aerospace systems and other applications. The lab has a long-standing interest in advancement of computational fluid dynamics for complex three-dimensional flows, enabling significant reductions in time from geometry to solution. Specific research interests include aerodynamics, aeroacoustics, flow control, fluid structure interactions, hypersonic flows, high-order methods, multi-level solution techniques, large eddy simulation, and scientific visualization. Research interests also extend to chemical kinetics, transport-chemistry interactions, and other reacting flow phenomena important for energy conversion and propulsion. ACDL’s efforts in uncertainty quantification aim to endow computational predictions with quantitative measures of confidence and reliability, while addressing broad underlying challenges of model validation. Complementary efforts in statistical inference and data assimilation are aimed at estimating and improving physical models and predictions by conditioning on observational data. Current research has developed effective methods for error estimation, solution adaptivity, sensitivity analysis, uncertainty propagation and the solution of stochastic differential equations, the solution of large-scale statistical inverse problems, nonlinear filtering in partial differential equations, and optimal experi-
mental design. Applications range from aerospace vehicle design to large-scale geophysical problems and subsurface modeling. ACDL research in simulation-based design and control is aimed at developing methods to support better decision-making in aerospace and other complex systems, with application to conceptual, preliminary, and detailed design. Recent efforts yielded effective approaches to PDE-constrained optimization, real time simulation and optimization of systems governed by PDEs, multiscale and multi-fidelity optimization, model order reduction, geometry management, and fidelity management. ACDL applies these methodologies to aircraft design and to the development of tools for assessing aviation environmental impact. ACDL faculty are Youssef Marzouk (director), David Darmofal, Mark Drela, Jaime Peraire, Qiqi Wang, and Karen Willcox. Research staff include Steven Allmaras, Robert Haimes, and Cuong Nguyen. Visit the Aerospace Computational Design Laboratory at http://acdl.mit.edu
AEROSPACE CONTROLS LABORATORY The Aerospace Controls Laboratory researches autonomous systems and control design for aircraft, spacecraft, and ground vehicles. Theoretical research is pursued in such areas as decision making under uncertainty; path planning, activity, and task assignment; mission planning for unmanned aerial vehicles, sensor network design; and robust, adaptive, and nonlinear control. A key aspect of ACL is RAVEN (Real-time indoor Autonomous Vehicle test ENvironment), a unique experimental facility that uses (i) a motion capture system to enable rapid prototyping of aerobatic flight controllers for helicopters and aircraft, and robust coordination algorithms for multiple vehicles; and (ii) a ground projection system that enables real time animation of the planning environment, beliefs, uncertainties, intentions of the vehicles, predicted
Lab report
59
behaviors (e.g., trajectories), and confidence intervals of the learning algorithms. Recent research includes the following: Robust Planning in Uncertain Environments: ACL developed consensusbased bundle algorithm (CBBA) as a distributed task-planning algorithm that provides provably good, conflict-free, approximate solutions for heterogeneous multi-agent missions. Aside from extensions to task time-windows, coupled agent constraints, asynchronous communications, and limited network, CBBA has been validated in real-time flight test experiments. ACL has also extended its development of chance-constrained rapidlyexploring random trees (CC-RRT), a robust planning algorithm to identify probabilistically feasible trajectories, to new aerospace domains. For instance, ACL recently developed CC-RRT* to solve robust pursuit-evasion problems. ACL is also involved in a multiyear Draper URAD on precision landing of guided parafoils, with robustness to large and dynamic wind environments. Finally, ACL is participating in a multi-year MURI focused on enabling decentralized planning algorithms under uncertainty. Ongoing ACL research has demonstrated that the use of flexible nonparametric Bayesian models for learning models of uncertain environment can greatly improve planning performance. UAV Mission Technologies: ACL has recently demonstrated autonomous, closed-loop UAV flight in MIT’s Wright Brothers’ Wind Tunnel. This novel capability allows the ACL to test flight controllers for windy environments in a controlled and systematic manner. ACL has also developed a novel hovering vehicle concept capable of agile, acrobatic maneuvers in cluttered indoor spaces. The vehicle is a quadrotor whose rotor tilt angles can be actuated, enabling upside-down hovering flight with appropriate control algorithms. Additionally, as part of research on long-duration UAV mission planning, ACL has constructed an autonomous recharge platform, capable of autonomous battery replacement and recharging for small UAVs. This capability allows ACL to demonstrate complex, multi-agent missions lasting for several hours.
60
AEROASTRO 2013-2014
Information-Gathering Networks: Recent ACL research has addressed maximizing information gathering in complex dynamic environments, through quantifying the value of information and the use of mobile sensing agents. The primary challenge in such planning is the computational complexity, due to both the large size of the information space and the cost of propagating sensing data into the future. ACL researchers created adaptive efficient distributed sensing in which each sensor propagates only high value information, reducing the network load and improving scalability. Recently-developed algorithms embed information planning within RRTs to quickly identify safe information-gathering trajectories for teams of sensing agents, subject to arbitrary constraints and sensor models. Task Identification and Decision-Making: Markov Decision Processes (MDP) and Partially Observable MDPs (POMDP) are natural frameworks for formulating many decision-making problems of interest. ACL has identified approximate solution techniques which can utilize this framework while lessening the curse of dimensionality and the curse of history typically encountered for exact solutions. ACL has also developed a Bayesian Nonparametric Inverse Reinforcement Learning algorithm for identifying tasks from traces of user behavior. This technique allows a user to “teach” a task to a learning agent through natural demonstrations. ACL has also enabled fast, real-time learning in combination with cooperative planning in uncertain and risky environments, while maintaining probabilistic safety guarantees for the overall system behavior. Finally, by efficiently using potentially inaccurate models of physical systems, ACL has developed a method that minimizes samples needed in real-world learning domains such as a car learning to race around a track. Robust State Estimation: Many navigation and robotic mapping systems are subject to sensor failures and sensor noise that do not match the assumed system models. In many cases, this model mismatch can cause divergence of the state estimates and poor
navigation system performance. ACL has developed several robust state estimation algorithms that address these issues by learning a model for the sensor noise while simultaneously generating the navigation solution. These algorithms apply hierarchical and nonparametric Bayesian models along with inference techniques such as Expectation-Maximization and variational inference to learn the noise models. In practice, the robust algorithms provide significantly more accurate solutions while requiring little additional computation relative to non-robust state estimation techniques. ACL has also applied this Bayesian framework to the Simultaneous Localization and Mapping (SLAM) problem to develop algorithms for vision-based SLAM that are robust to landmark misidentifications that cause non-robust SLAM algorithms to fail catastrophically. ACL faculty are Jonathan How and Steven Hall. Visit the Aerospace Controls Laboratory at http://acl.mit.edu
AEROSPACE ROBOTICS AND EMBEDDED SYSTEMS GROUP The Aerospace Robotics and Embedded Systems group’s mission is the development of theoretical foundations and practical algorithms for real-time control of large-scale systems of vehicles and mobile robots. Application examples range from UAVs and autonomous cars, to air traffic control and urban mobility. The group researches advanced algorithmic approaches to control high-dimensional, fast, and uncertain dynamical systems subject to stringent safety requirements in a rapidly changing environment. An emphasis is placed on the development of rigorous analysis, synthesis, and verification tools to ensure the correctness of the design. The research approach combines expertise in control theory, robotics, optimization, queuing theory and stochastic systems, with randomized and distributed algorithms, formal languages, machine learning, and game theory.
Current research areas include the following: »» Autonomy and future urban mobility: Autonomous, self-driving cars are no longer science fiction, but will be ready for commercial deployment soon. The group’s work on selfdriving vehicles is broad, spanning the whole spectrum from technology development to the analysis of socio-economic impact of such technology. Recent work includes: »» Affordable autonomy: can we design safe and reliable selfdriving vehicles at a cost that make them affordable for the general public? Our demo vehicles at the Singapore-MIT Alliance for Research and Technology were developed with less than $30,000 worth of computers and sensors. »» Provable safety: how do we make sure that the vehicle will behave safely, and respect all the rules of the road? We developed algorithms that provably satisfy all “hard” rules, while minimizing violations of “soft” rules or recommendations. »» Autonomy for mobility on demand: How would self-driving vehicles impact urban mobility in the future? We envision fleets of shared self-driving vehicles, develop algorithms for their sizing and operations, and analyze their effects using real data from several cities worldwide. Real-time motion planning and control: The group is developing stateof-the art algorithms for real-time control of highly maneuverable aircraft, spacecraft, and ground vehicles. Focus areas include optimality and robustness, as well as provable safety and correctness with respect to temporal-logic specifications (e.g., rules of the road, rules of engagement). Current projects include high-speed flight in cluttered environments and high-speed offroad driving. Multi-agent systems: Large, heterogeneous groups of mobile vehicles, such as UAVs and UGVs, are increasingly used to address complex missions for many applications, ranging from national
Lab report
61
security to environmental monitoring. An additional emphasis in this work is scalability; our objective is not only the design of distributed algorithms to ensure provably efficient and safe execution of the assigned tasks, but also to understand exactly how the collective performance and implementation complexity scale as the group’s size and composition change. Transportation networks: Traffic congestion, and extreme sensitivity to, for example, environmental disruptions, is a well-known effect of increasing access to transportation. As infrastructure development saturates, new approaches are necessary to increase the safety, efficiency, and environmental sustainability of transportation networks. The group’s research in this area concentrates on the exploitation of real-time information availability through wireless communications among vehicles, and with existing infrastructure, to achieve this goal. Emilio Frazzoli directs the Aerospace Robotics and Embedded Systems group. Visit the Aerospace Robotics and Embedded Systems group at http://ares.lids.mit.edu
THE AUTONOMOUS SYSTEMS LABORATORY The Autonomous Systems Laboratory is a virtual lab led by Professors Brian Williams and Nicholas Roy. Williams’ group, the Model-based Embedded and Robotics (MERS) group, and Roy’s Robust Robotics Group are part of the Computer Science and Artificial Intelligence Lab. ALS work is focused on developing autonomous aerospace vehicles and robotic systems. ASL-developed systems are commanded at a high-level in terms of mission goals. The systems execute these missions robustly by constantly estimating their state relative to the world, and by continuously adapting their plan of action, based on engineering and world models.
62
AEROASTRO 2013-2014
Below are several recent demonstrations. »» The lab has demonstrated autonomous vehicles to maximize utility in an uncertainty environment, while operating within acceptable levels of risk. Autonomous underwater vehicles enable scientists to explore previously uncharted portions of the ocean, by autonomously performing science missions of up to 20 hours long without the need for human intervention. Performing these extended missions can be a risky endeavor. Researchers have developed robust, chanceconstraint planning algorithms that automatically navigate vehicles to achieve user specified science goals, while operating within risk levels specified by the users. (Video at http://www.csail.mit.edu/videoarchive/research/robo/ auv-planning) »» Another demonstration involves human-robot interaction between a robotic air taxi and a passenger. The task is for the autonomous vehicle to help the passenger rethink goals when they no longer can be met. Companies like the MIT spinoff Terrafugia offer vehicles that can fly between local airports and can travel on local roads. To operate these innovative vehicles, one must be trained as a certified pilot, thus limiting the population that can benefit from this innovative concept. In collaboration with Boeing, MERS has demonstrated in simulation the concept of an autonomous personal air vehicle in which the passenger interacts with the vehicle in the same manner that they interact today with a taxi driver. (Video at http://www.csail.mit.edu/videoarchive/ research/robo/personal-aerial-transportation.) »» A third demonstration involves human-robot interaction between an astronaut and the Athlete Lunar Rover. MERS has developed methods for controlling walking machines,
guided by qualitative “snapshots” of walking gait patterns. These control systems achieve robust walking over difficult terrain by embodying many aspects of a human’s ability to restore balance after stumbling, such as adjusting ankle support, moving free limbs, and adjusting foot placement. Members of the MERS group applied generalizations of these control concepts to control the JPL Athlete robot, a six-legged/wheeled lunar rover that performs heavy lifting and manipulation tasks by using its legs as arms. (Video at http://www.csail.mit.edu/videoarchive/research/robo/ athlete-mers.) ASL faculty are Brian Williams and Nicholas Roy. Visit the Model-based Embedded Systems at http://mers.csail.mit.edu and the Robust Robotics Group at http://groups.csail.mit.edu/rrg
COMMUNICATIONS AND NETWORKING RESEARCH GROUP The Communications and Networking Research Group’s primary goal is design of network architectures that are cost effective, scalable, and meet emerging needs for high data-rate and reliable communications. To meet emerging critical needs for military communications, space exploration, and internet access for remote and mobile users, future aerospace networks will depend upon satellite, wireless, and optical components. Satellite networks are essential for providing access to remote locations lacking in communications infrastructure, wireless networks are needed for communication between untethered nodes, such as autonomous air vehicles, and optical networks are critical to the network backbone and in high performance local area networks. The group is working on a wide range of projects in the area of communication networks and systems, with application to satellite, wireless, and optical systems. The group has been developing
efficient network control algorithms for heterogeneous wireless networks. Existing wireless networks are almost exclusively confined to single hop access, as provided by cellular telephony or wireless LANs. While multi-hop wireless networks can be deployed, current protocols typically result in extremely poor performance for even moderate sized networks. Wireless Mesh Networks have emerged as a solution for providing last-mile Internet access. However, hindering their success is our relative lack of understanding of how to effectively control wireless networks, especially in the context of advanced physical layer models, realistic models for channel interference, distributed operations, and interface with the wired infrastructure (e.g., the internet). CNRG has been developing effective and practical network control algorithms that make efficient use of wireless resources through the joint design of topology adaptation, network layer routing, link layer scheduling, and physical layer power, channel, and rate control. Robust network design is another exciting area of recent pioneering research by the group. In particular, the group has been developing a paradigm for the design of highly robust networks that can survive a massive disruption that may result from natural disasters or intentional attack. The work examines the impact of large-scale failures on network survivability and design, with a focus on interdependencies between different networked infrastructures, such as telecommunication networks, social networks, and the power grid. The group’s research crosses disciplinary boundaries by combining techniques from network optimization, queueing theory, graph theory, network protocols and algorithms, hardware design, and physical layer communications. Eytan Modiano directs the Communications and Networking Research Group. Visit the Communications and Networking Research Group at http://cnrg.mit.edu
Lab report
63
GAS TURBINE LABORATORY The Gas Turbine Laboratory’s mission is to advance the state-ofthe-art in fluid machinery for power and propulsion. The research is focused on advanced propulsion systems, energy conversion and power, with activities in computational, theoretical, and experimental study of loss mechanisms and unsteady flows in fluid machinery, dynamic behavior and stability of compression systems, instrumentation and diagnostics, advanced centrifugal compressors and pumps for energy conversion, gas turbine engine and fluid machinery noise reduction and aero-acoustics, novel aircraft and propulsion system concepts for reduced environmental impact. Current research projects include: »» a unified approach for vaned diffuser design in advanced centrifugal compressors »» improved performance return channel design for multistage centrifugal compressors »» investigation of real gas effects in supercritical CO2 compression systems
»» a two-engine integrated propulsion system »» propulsor design for exploitation of boundary layer ingestion »» an acoustic shielding method for assessment of turbomachinery noise in advanced aircraft configurations »» aerodynamics and heat transfer in gas turbine tip shroud cavity flows »» secondary air interactions with main flow in axial turbines »» compressor aerodynamics in large industrial gas turbines for power generation »» turbine tip clearance loss mechanisms »» flow and heat transfer in modern turbine rim seal cavities »» modeling cavitation instabilities in rocket engine turbopumps
»» modeling instabilities in high-pressure pumping systems,
»» diagnostics and prognostics for gas turbine engine system stability characterization
»» aeromechanic response in a high performance centrifugal compressor stage ported shroud operation in turbochargers
»» investigation of the origins of short-wavelength instability inception in axial compressors
»» manifestation of forced response in a high performance centrifugal compressor stage for aerospace applications
»» assessment of thermal effects on compressor transients.
»» return channel design optimization using adjoint method for multistage centrifugal compressors »» multiparameter control for centrifugal compressor performance optimization
64
»» performance improvement of a turbocharger twin scroll type turbine stage
AEROASTRO 2013-2014
Faculty and research staff include Daniel Cuppoletti, David Darmofal, Fredric Ehrich, Alan Epstein (emeritus), Edward Greitzer, Arthur Huang, Claudio Lettieri, Zoltan Spakovszky (director), Choon Tan, Neil Titchener, and Alejandra Uranga. Visit the Gas Turbine Lab at http://mit.edu/aeroastro/labs/gtl
HUMANS AND AUTOMATION LABORATORY Research in the Humans and Automation Laboratory focuses on the multifaceted interactions of human and computer decision-making in complex socio-technical systems. With the explosion of automated technology, the need for humans as supervisors of complex automatic control systems has replaced the need for humans in direct manual control. A consequence of complex, highly-automated domains in which the human decision-maker is more on-the-loop than in-the-loop is that the level of required cognition has moved from that of well-rehearsed skill execution and rule following to higher, more abstract levels of knowledge synthesis, judgment, and reasoning. Employing human-centered design principles to human supervisory control problems, and identifying ways in which humans and computers can leverage the strengths of the other to achieve superior decisions together is HAL’s central focus. Current research projects include investigation of human understanding of complex optimization algorithms and visualization of cost functions, human performance modeling with hidden Markov models, collaborative human-computer decision making in time-pressured scenarios (for both individuals and teams), human supervisory control of multiple unmanned vehicles, and designing displays that reduce training time. Lab equipment includes an experimental testbed for future command and control decision support systems, intended to aid in the development of humancomputer interface design recommendations for future unmanned vehicle systems. In addition, the lab hosts a state-of-the-art multiworkstation collaborative teaming operations center, as well as a mobile command and control experimental test bed mounted in a Dodge Sprint van awarded through the Office of Naval Research. Current research sponsors include the Office of Naval Research, the U.S. Army, Lincoln Laboratory, Boeing, the Air Force Research Laboratory, the Air Force Office of Scientific Research, Alstom, and the Nuclear Regulatory Commission.
HAL faculty include Mary L. Cummings, director; Nicholas Roy; and Thomas Sheridan. Visit the Humans and Automation Laboratory at http://mit.edu/aeroastro/labs/halab
INTERNATIONAL CENTER FOR AIR TRANSPORTATION The International Center for Air Transportation undertakes research and educational programs that discover and disseminate the knowledge and tools underlying a global air transportation industry driven by technologies. Global information systems are central to the future operation of international air transportation. Modern information technology systems of interest to ICAT include global communication and positioning, international air traffic management, scheduling, dispatch, and maintenance support, vehicle management, passenger information and communication, and real-time vehicle diagnostics. Airline operations are also undergoing major transformations. Airline management, airport security, air transportation economics, fleet scheduling, traffic flow management, and airport facilities development, represent areas of great interest to the MIT faculty and are of vital importance to international air transportation. ICAT is a physical and intellectual home for these activities. ICAT, and its predecessors, the Aeronautical Systems Laboratory and Flight Transportation Laboratory, pioneered concepts in air traffic management and flight deck automation and displays that are now in common use. ICAT faculty include R. John Hansman (director), Cynthia Barnhart, Peter Belobaba, Hamsa Balakrishnan, and Amedeo Odoni. Visit the International Center for Air Transportation at http://mit.edu/ aeroastro/ labs/ICAT/
Lab report
65
LABORATORY FOR AVIATION AND THE ENVIRONMENT The Laboratory for Aviation and the Environment was founded in the 1990s as the Aero-Environmental Research Laboratory by Ian A. Waitz, now dean of the MIT School of the Engineering. One of the aviation industry’s defining challenges is addressing aviation’s environmental impact in terms of noise, air quality, and climate change. LAE’s goal is to align the trajectory of aerospace technology and policy development with the need to mitigate these impacts. It does so by increasing the understanding the environmental effects of aviation, by developing and assessing fuel-based, operational and technological mitigation approaches, and by disseminating knowledge and tools. LAE also contributes to cognate areas of inquiry in aerospace, energy and the environment LAE researchers are analyzing environmental impacts and developing research tools that provide rigorous guidance to policy-makers who must decide among alternatives when addressing aviation’s environmental impact. The MIT researchers collaborate with international teams in developing aircraft-level and aviation system-level tools to assess the costs and benefits of different policies and mitigation options. A current LAE focus is on studying the environmental sustainability of alternative aviation fuels from biomass or natural gas. This research includes both drop-in fuel options, which can be used with existing aircraft engines and fuel infrastructure, as well as non-drop-in options such as liquefied natural gas, which would require modifications to aircrafts and infrastructure. Environmental metrics considered include lifecycle greenhouse gas emissions, land requirements and water consumption. LAE researchers are also estimating tradeoffs among different metrics and usages to better understand the full consequences of introducing a certain alternative fuel into the aviation system.
66
AEROASTRO 2013-2014
LAE has developed and publicly released a code that allows for modeling and evaluation emissions and their impacts throughout the troposphere and stratosphere in a unified fashion. LAE has also recently released a new global emissions dataset for civil aviation emissions, which represents the most current estimate of emissions publicly available. It is widely used by researchers worldwide, in areas including atmospheric modeling and aviation and the environment. Other recent work quantifies air pollution and associated health effects attributable to the different economic sectors in the United States, and the environmental and economic impacts of higher octane gasoline usage for road transportation. LAE faculty includes Steven Barrett, director; Robert Malina, associate director; Hamsa Balakrishnan; John Hansman; Ian Waitz; and Karen Willcox. Also associated with LAE are Ray Speth, research scientist; and Brian Yutko, postdoctoral associate. Visit LAE at http://lae.mit.edu
LABORATORY FOR INFORMATION AND DECISION SYSTEMS The Laboratory for Information and Decision Systems is an interdepartmental research center committed to advancing research and education in the analytical information and decision sciences: systems and control, communications and networks, and inference and statistical data processing. Dating to 1939, LIDS has been at the forefront of major methodological developments, relevant to diverse areas of national and worldwide importance, such as telecommunications, information technology, the automotive industry, energy, defense, and human health. Building on past innovation and bolstered by a collaborative atmosphere, LIDS members continue to make breakthroughs that cut across traditional boundaries.
Members of the LIDS community share a common approach to solving problems and recognize the fundamental role that mathematics, physics, and computation play in their research. Their pursuits are strengthened by the laboratory’s affiliations with colleagues across MIT and throughout the world, as well as with leading industrial and government organizations. LIDS is based in MIT’s Stata Center, a dynamic space that promotes a high level of interaction within the lab and with the larger MIT community. AeroAstro faculty affiliated with the laboratory are Emilio Frazzoli, Jonathan How, Eytan Modiano, and Moe Win. Visit LIDS at http://lids.mit.edu/
THE LEARNING LABORATORY The AeroAstro Learning Laboratory, located in Building 33, promotes student learning by providing an environment for hands-on activities that span our conceive-design-implementoperate educational paradigm. The Learning Lab comprises four main areas: Robert C. Seamans Jr. Laboratory. The Seamans Laboratory occupies the first floor. It includes: »» The Concept Forum — a multipurpose room for meetings, presentations, lectures, videoconferences and collaboration, distance learning, and informal social functions. In the Forum, students work together to develop multidisciplinary concepts, and learn about program reviews and management. »» Al Shaw Student Lounge — a large, open space for social interaction and operations.
Instrumentation Laboratory, Mechanical Projects Area, Projects Space, and the Composite Fabrication-Design Shop. The Gelb Laboratory provides facilities for students to conduct handson experiential learning through diverse engineering projects starting as first-year students and continuing through the last year. The Gelb facilities are designed to foster teamwork with a variety of resources to meet the needs of curricular and extracurricular projects. Gerhard Neumann Hangar. The Gerhard Neumann Hangar is a high bay space with an arching roof. This space lets students work on large-scale projects that take considerable floor and table space. Typical of these projects are planetary rovers, autonomous vehicles, and re-entry impact experiments. The structure also houses low-speed and supersonic wind tunnels. A balcony-like mezzanine level is used for multi-semester engineering projects, such as the experimental three-term senior capstone class. Digital Design Studio. The Digital Design Studio, located on the second floor, is a large room with multiple computer stations arranged around reconfigurable conference tables. Here, students conduct engineering evaluations and design work, and exchange computerized databases as system and subsystem trades are conducted during the development cycle. The room is equipped with information technologies that facilitate teaching and learning in a team-based environment. Adjacent and networked to the main Design Studio are two smaller design rooms: the AA Department Design Room, and the Arthur W. Vogeley Design Room. These rooms are reserved for the use of individual design teams and for record storage. The department’s IT systems administrator is positioned for convenient assistance in an office adjacent to the Design Center.
Arthur and Linda Gelb Laboratory. Located in the building’s lower level, the Gelb Laboratory includes the Gelb Machine Shop,
Lab report
67
MAN VEHICLE LABORATORY The Man Vehicle Laboratory improves the understanding of human physiological and cognitive capabilities as applied to human-vehicle and human-robotic system safety and efficacy, as well as decision-making augmented by technological aids. MVL develops countermeasures and display designs to aid pilots, astronauts, clinicians, patients, soldiers, and others. Research is interdisciplinary, and uses techniques from manual and supervisory control, signal processing, estimation, robotics, sensory-motor physiology, sensory and cognitive psychology, biomechanics, human factors engineering, artificial intelligence, and biostatistics. MVL has flown experiments on the Space Shuttle, the Mir Space Station, and on many parabolic flights, and developed experiments for the International Space Station. Space applications include advanced space suit design and dynamics of astronaut motion, adaptation to rotating artificial gravity, mathematical models for human spatial disorientation, accident analysis, artificial intelligence, and space telerobotics training. Ongoing work includes the development of countermeasures using a short radius centrifuge and development of a g-loading suit to maintain muscle and bone strength. New major projects include a collaborative study of adaptation in roll tilt perception and manual control to altered G environments using a centrifuge at the Massachusetts Eye and Ear Infirmary, and a study with UC Davis on customized and just in time space telerobotics refresher training. Non-aerospace projects have included locomotive cab automation and displays, advanced helmet designs for brain protection in sports and against explosive blasts, the development of wearable sensor systems and data visualizations for augmenting clinical decision-making, and data fusion for improving situation awareness for dismounted soldiers. A new initiative with Russian colleagues emphasizes innovative solutions to the protection and performance enhancement of astronauts during space exploration.
68
AEROASTRO 2013-2014
Research sponsors include NASA, the National Space Biomedical Research Institute, the Office of Naval Research, the Department of Transportation’s FAA and FRA, C.S. Draper Laboratory, the Center for Integration of Medicine and Innovative Technology, the Deshpande Center, the MIT/Skoltech program and the MIT Portugal Program. The laboratory also collaborates with the Volpe Transportation Research Center, and the Jenks Vestibular Physiology Laboratory of the Massachusetts Eye and Ear Infirmary. MVL faculty include Charles Oman, Jeffrey Hoffman and Dava Newman (co-directors), Laurence Young, Julie Shah, and Leia Stirling. They teach subjects in human factors engineering, space systems engineering, real-time systems and software, space policy, flight simulation, space physiology, aerospace biomedical engineering, the physiology of human spatial orientation, statistical methods in experimental design, and leadership. The MVL also serves as the office of the Director for the NSBRI-sponsored HST Graduate Program in Bioastronautics (Young), the Massachusetts Space Grant Consortium (Hoffman), and the MIT Portugal Program’s Bioengineering Systems focus area (Newman). Visit the Man Vehicle Laboratory at http://mvl.mit.edu
NECSTLAB The necstlab (pronounced “next lab”) research group explores new concepts in engineered materials and structures, with a focus on nanostructured materials. The group’s mission is to lead the advancement and application of new knowledge at the forefront of materials and structures understanding, with research contributions in both science and engineering. Applications of interest include enhanced aerospace advanced composites, multifunctional attributes of structures such as damage sensing, and microfabricated (MEMS) topics. A significant effort over the past decade has been to use nanoscale materials to enhance
performance of advanced aerospace materials and their structures through the industry supported NECST Consortium.
»» electroactive nanoengineered actuator/sensor architectures focusing on ion transport
The necstlab group has interests that span fundamental materials synthesis questions through to structural applications of both hybrid and traditional materials. Much of the group’s work supports the efforts of the NECST Consortium, an aerospace industry-supported research initiative that seeks to develop the underlying understanding to create enhanced-performance advanced composites using nanotechnology. Beyond the NECST Consortium Members, necstlab research is supported directly or through collaboration by industry, AFOSR, ARO, NASA, NIST, NSF, ONR, and others.
»» nanoengineered (hybrid) composite architectures for laminate-level mechanical performance improvement
The necstlab maintains collaborations around the MIT campus, particularly with faculty in the Mechanical Engineering, Materials Science and Engineering, and Chemical Engineering departments; and MIT labs and centers including the Institute for Soldier Nanotechnologies, Materials Processing Center, Center for Materials Science and Engineering, and the Microsystems Technology Laboratory, as well as Harvard’s Center for Nanoscale Systems. Important to the contributions of the necstlab are collaborations with leading research groups from around the world through formal and informal collaborations. In the fall of 2014, the group moved into a new laboratory in MIT Building 35. Example past and current research projects include: »» BioNEMS materials design and implementation in microfluidics »» buckling mechanics »» carbon nanostructure synthesis from non-traditional catalysts »» continuous growth of aligned carbon nanotubes
»» multifunctional properties including damage sensing and detection »» manufacturing »» polymer nanocomposite mechanics and electrical and thermal transport »» Si MEMS devices including piezoelectric energy harvesters, microfabricated solid oxide fuel cells, stress characterization, and 3D MEMS »» VACNT characterization and physical properties necstlab faculty include Brian L. Wardle, director, and John Dugundji, emeritus. Visit necstlab at http://mit.edu/aeroastro/labs/necstlab
SPACE PROPULSION LABORATORY The Space Propulsion Laboratory studies and develops systems for increasing performance and reducing costs of space propulsion and related technologies. A major area of interest to the lab is electric propulsion in which electrical, rather than chemical, energy propels spacecraft. The benefits are numerous, hence the reason electric propulsion systems are increasingly applied to communication satellites and scientific space missions. These efficient engines allow exploration in more detail of the structure of the universe, increase the lifetime of commercial payloads, and look for signs of life in far away places. Areas of research include plasma engines and plumes, and their interaction with spacecraft and thruster materials, and numerical and experimental models
Lab report
69
of magnetic cusped thrusters. SPL students also work on ultrafast (nanosecond) high voltage discharges to trigger combustion reactions and eventually reduce aircraft engine pollution. SPL also has a significant role in designing and building micropropulsion electrospray thrusters, including their integration into space missions. In addition to providing efficient propulsion for very small satellites in the 1 kg category (like CubeSats), these engines will enable distributed propulsion for the control of large space structures, such as deformable mirrors and apertures. The science behind electrosprays is explored as well, mainly on the ionic regime where molecular species are directly evaporated from ionic liquid surfaces. Also, applications beyond propulsion are investigated; for example, the use of highly monoenergetic molecular ion beams in focusing columns for materials structuring and characterization at the nano-scale. SPL facilities include a supercomputer cluster where plasma and molecular dynamics codes are routinely executed and a state-ofthe-art laboratory including five vacuum chambers, clean room environment, electron microscopy, materials synthesis capabilities, nanosatellite qualification equipment (vibration/thermal and in-vacuum magnetically-levitated CubeSat simulator) and plasma/ion beam diagnostic tools to support ongoing research efforts. SPL faculty are professors Paulo Lozano, director, and Manuel Martinez-Sanchez, emeritus. Visit the Space Propulsion Lab at http://mit.edu/aeroastro/labs/spl
70
AEROASTRO 2013-2014
SPACE SYSTEMS LABORATORY The Space Systems Laboratory research contributes to the exploration and development of space. SSL’s mission is to explore innovative space systems concepts while training researchers to be conversant in this field. The major programs include systems analysis studies and tool development, AeroAstro student-built instruments and small satellites for exploration and remote sensing, precision optical systems for space telescopes, and microgravity experiments operated aboard the International Space Station and the NASA Reduced Gravity aircraft. Research topics focus on space systems and include dynamics, guidance, and control; active structural control; space power and propulsion; modular space systems design; micro-satellite design; real-time embedded systems; software development; and systems architecting. SSL has a unique facility for space systems research, the Synchronized Position Hold Engage and Reorient Experimental Satellites (SPHERES). The SPHERES facility is used to develop proximity satellite operations such as inspection, cluster aggregation, collision avoidance and docking, as well as formation flight. The SPHERES facility consists of three satellites 20 centimeters in diameter that have been inside the International Space Station since May 2006. In 2009, SSL expanded the uses of SPHERES to include STEM outreach through an exciting program called Zero Robotics (http://zerorobotics.mit.edu), which engages high school and middle school students in a competition aboard the ISS using SPHERES. A pilot program began in 2009 with two teams from Idaho, and has expanded to approximately 100 U.S. and 50 European teams annually. In 2013, finalists were joined by astronauts Barbara Morgan, Gregory Johnson, and Gregory Chamitoff, and by “Ender’s Game” director Gavin Hood and special effects supervisor Mattew Buttler as they watched in real time via video downlink their code run on the satellites aboard the ISS.
There have been several recent and exciting hardware augmentations to SPHERES. In October 2012, the SPHERES facility on the ISS was expanded to include vision-based navigation through the Visual Estimation for Relative Tracking and Inspection of Generic Objects (VERTIGO) program. VERTIGO uses vision-based navigation and mapping algorithms through a stereo camera system and an upgraded processor. In 2013, the University of Maryland Space Power and Propulsion Laboratory, Aurora Flight Sciences, and SSL upgraded the SPHERES facility to include the Resonant Inductive Near-Field Generation System, which has been used to test electromagnetic formation flight and wireless power transfer through a pair of tuned resonant coils that generate a time varying magnetic field. The SPHERES-Slosh module was also launched in 2013, in collaboration with the Florida Institute of Technology, and has enabled surveys of fluid slosh behavior in zero gravity. Recently, SSL partnered with the Naval Research Laboratory and Aurora Flight Sciences for work on the Defense Advanced Research Projects Agency Phoenix program for satellite servicing and assembly missions. To this end, a cadre of Universal Docking Ports and Halos will be launched and operated on the SPHERES ISS facility in the fall of 2014. The Universal Docking Ports enable active docking and undocking of the satellites creating a rigid assembly; they add fiducial-based vision navigation. The Halo structure enables attachment of up to six electro-mechanical devices around a single SPHERES satellites, allowing researchers to study complex geometrical system reconfiguration. Durging the fall 2013 and spring 2014 semesters, the MIT 16.83x senior space systems design capstone class developed a prototype for the INSPECT (Integrated Navigation Sensor Platform for EVA Control and Testing) program, which added a thermal imager, optical range finder, and control moment gyros to the ground SPHERES + Halo system.
SSL is also active in the area of nanosatellites, particularly CubeSats. In March 2014, the Microsized Microwave Atmospheric Satellite (MicroMAS), a joint effort between MIT SSL students, led by Professor Kerri Cahoy, and MIT Lincoln Laboratory staff, led by Dr. Bill Blackwell, was delivered to NanoRacks for launch to the ISS, from which MicroMAS will be deployed in summer 2014. MicroMAS is a dual-spinning three-unit CubeSat hosting a microwave radiometer payload that captures temperature map images of Earth and is important for characterizing hurricanes and tropical storms. SSL will develop a follow-on to MicroMAS called the Microwave Radiometer Technology Acceleration mission (MiRaTA) which pairs an advanced miniature microwave radiometer with a GPS radio occultation receiver to help improve radiometer calibration. MiRaTA is sponsored by the NASA Earth Science Technology Office and scheduled for a 2016 launch. SSL students are working with engineers at Aurora Flight Sciences and the AeroAstro Space Propulsion Laboratory on a cluster formation-flight nanosatellite project. SSL students are also engaged with Professor Sara Seager and students in the MIT Earth and Planetary Sciences department, engineers at the Charles Stark Draper Laboratory, and NASA Jet Propulsion Laboratory on the novel ExoplanetSat nanosatellite which uses a two-stage control system (reaction wheels plus piezo stage) to maintain precise pointing at a target star to obtain exoplanet transit measurements using advanced photometry. The Wavefront Control Laboratory, led by Cahoy, is a part of SSL that focuses on precision active optical systems for space applications. WCL students are working on three projects. One is the Deformable Mirror Demonstration Mission (DeMi), which will validate and demonstrate the capabilities of high actuator count MEMS deformable mirrors for high contrast astronomical imaging. The DeMi optical payload will characterize MEMS
Lab report
71
72
deformable mirror operation using both a Shack Hartmann wavefront sensor as well as sensorless wavefront control. The second project is a CubeSat Free Space Optical communications downlink that uses a staged control system with MEMS fast steering mirrors. The third project is a bistatic laser system for active characterization of the bidirectional reflectance distribution function of space materials. WCL students also are investigating whether or not we can use standard communications satellite components as space weather sensors, and developing algorithms that can predict the onset of space-weather related anomalies. (http://web.mit.edu/cahoylab).
SSL is directed by Dr. Alvar Saenz Otero who took the reins this year from Professor David W. Miller, who is on leave from MIT to NASA its chief technologist. Other SSL personnel include professors Kerri Cahoy, Jeffrey Hoffman, Olivier de Weck, and Richard Binzel; Dr. Rebecca Masterson, Dr. Danilo Roascio, research specialist Paul Bauer, fiscal officers Suxin Hu and Ngan Kim Le, and administrative assistant Marilyn E. Good. Collaborators include AeroAstro professors Manuel Martinez-Sanchez and Paulo Lozano, and EAPS professor Sara Seager.
SSL is also developing and building the REXIS (REgolith X-ray Imaging Spectrometer) student collaboration instrument, which will be aboard NASA’s next New Frontiers mission: OSIRIS-REx (Origins, Spectral Interpretation, Resource Identification, Security Regolith Explorer). The mission is an asteroid sample return mission, which will launch in 2016 to visit the Near-Earth Asteroid Bennu. REXIS is one of five instruments onboard and uses a 2x2 array of Lincoln Laboratory designed charged-coupled devices to measure the X-ray fluorescence from Bennu to allow characterization of the surface of the asteroid among the major meteorite groups as well as a coded aperture mask to map the spatial distribution of element concentrations in the regolith. Professor Richard Binzel, who maintains a joint MIT EAPS–AeroAstro appointment and research engineer Dr. Rebecca Masterson are leading the project in collaboration with EAPS, Kavli Institute and Harvard College Observatory. REXIS has included the work of more than 50 undergraduate and graduate students and has successfully completed its Critical Design Review. The team is wrapping up the engineering model testing and is on the path to the flight hardware build. The REXIS flight instrument will be delivered to Lockheed Martin for integration onto the OSIRISREx spacecraft in the summer of 2015.
SYSTEM ENGINEERING RESEARCH LAB
AEROASTRO 2013-2014
Visit the Space Systems Laboratory at http://ssl.mit.edu
The increasingly complex systems we are building today enable us to accomplish tasks that were previously difficult or impossible. At the same time, they have changed the nature of accidents and increased the potential to harm not only life today but also future generations. Traditional system safety engineering approaches, which started in the missile defense systems of the 1950s, are being challenged by the introduction of new technology and the increasing complexity of the systems we are attempting to build. Software is changing the causes of accidents and the humans operating these systems have a much more difficult job than simply following pre-defined procedures. We can no longer effectively separate engineering design from human factors and from the social and organizational system in which our systems are designed and operated. The System Engineering Research Lab’s goal is to create tools and processes that will allow us to engineer a safer world. Engineering safer systems requires multidisciplinary and collaborative research based on sound system engineering principles — it requires a holistic systems approach. LSSR has participants from multiple engineering disciplines and MIT schools as well as col-
laborators at other universities and in other countries. Students are working on safety in aviation (aircraft and air transportation systems), spacecraft, medical devices and healthcare, automobiles, railroads, nuclear power, defense systems, energy, and large manufacturing/process facilities. Cross-discipline topics include: »» hazard analysis »» accident causality analysis and accident investigation »» safety-guided design »» human factors and safety »» integrating safety into the system engineering process »» identifying leading indicators of increasing risk »» certification, regulation, and standards »» the role of culture, social, and legal systems on safety »» managing and operating safety-critical systems The System Engineering Research Lab is directed by Professor Nancy Leveson. Visit the System Engineering Research Lab at http://sunnyday.mit.edu/safety.html
TECHNOLOGY LABORATORY FOR ADVANCED MATERIALS AND STRUCTURES A dedicated and multidisciplinary group of researchers constitute the Technology Laboratory for Advanced Materials and Structures. They work cooperatively to advance the knowledge base and understanding that will help facilitate and accelerate advanced materials systems development and use in various advanced structural applications and devices. TELAMS has broadened its interests from a strong historical background in composite materials, and this is reflected in the name
change from the former Technology Laboratory for Advanced Composites. Thus, the research interests and ongoing work in the laboratory represent a diverse and growing set of areas and associations. Areas of interest include: »» composite tubular structural and laminate failures »» MEMS-scale mechanical energy harvesting modeling, design, and testing »» MEMS device modeling and testing, including bioNEMS/MEMS »» structural health monitoring system development and durability assessment »» thermostructural design, manufacture, and testing of composite thin films and associated fundamental mechanical and microstructural characterization »» continued efforts on addressing the roles of lengthscale in the failure of composite structures »» numerical and analytical solid modeling to inform, and be informed by, experiments »» continued engagement in the overall issues of the design of composite structures with a focus on failure and durability, particularly within the context of safety In supporting this work, TELAMS has complete facilities for the fabrication of structural specimens such as coupons, shells, shafts, stiffened panels, and pressurized cylinders made of composites, active, and other materials. TELAMS testing capabilities include a battery of servohydraulic machines for cyclic and static testing, a unit for the catastrophic burst testing of pressure vessels, and an impact testing facility. TELAMS maintains capabilities for environmental conditioning, testing at low and high temperature, and in hostile and other controlled environments.
Lab report
73
There are facilities for microscopic inspection, nondestructive inspection, high-fidelity characterization of MEMS materials and devices, and a laser vibrometer for dynamic device and structural characterization. This includes ties to ability for computer microtomography. With its linked and coordinated efforts, both internal and external, the laboratory is committed to leadership in the advancement of the knowledge and capabilities of the materials and structures community through education of students, original research, and interactions with the community. There has been a broadening of this commitment consistent with the broadening of the interest areas in the laboratory. In all these efforts, the laboratory and its members continue their extensive collaborations with industry, government organizations, other academic institutions, and other groups and faculty within the MIT community. TELAMS faculty include Paul A. LagacĂŠ and John Dugundji (emeritus). Visit the Technology Laboratory for Advanced Materials and Structures at http://mit.edu/telams
WIRELESS COMMUNICATION AND NETWORK SCIENCES GROUP The Wireless Communication and Network Sciences Group is involved in multidisciplinary research that encompasses developing fundamental theories, designing algorithms, and conducting experiments for a broad range of real-world problems. Its current research topics include location-aware networks, network synchronization, aggregate interference, intrinsically-secure networks, time-varying channels, multiple antenna systems, ultra-wide bandwidth systems, optical transmission systems, and space.
74
AEROASTRO 2013-2014
The group is working on location-aware networks in GPS-denied environments, which provide highly accurate and robust positioning capabilities for military and commercial aerospace networks. It has developed a foundation for the design and analysis of large-scale location-aware networks from the perspective of theory, algorithms, and experimentation. This includes derivation of performance bounds for cooperative localization, development of a geometric interpretation for these bounds, and the design of practical, near-optimal cooperative localization algorithms. It is currently validating the algorithms in a realistic network environment through experimentation in the lab. The lab is engaged in the development of a state-of-the-art apparatus that enables automated channel measurements. The apparatus makes use of a vector network analyzer and two vertically polarized, omni-directional wideband antennas to measure wireless channels over a range of 2–18 GHz. It is unique in that extremely wide bandwidth data, more than twice the bandwidth of conventional ultra-wideband systems, can be captured with high-precision positioning capabilities. Data collected with this apparatus facilitates the efficient and accurate experimental validation of proposed theories and enables the development of realistic wideband channel models. Work is underway to analyze the vast amounts of data collected during an extensive measurement campaign. Lab students are also investigating physical-layer security in large-scale wireless networks. Such security schemes will play increasingly important roles in new paradigms for guidance, navigation, and control of unmanned aerial vehicle networks. The framework they have developed introduces the notion of a secure communications graph, which captures the informationtheoretically secure links that can be established in a wireless network. They have characterized the s-graph in terms of local
and global connectivity, as well as the secrecy capacity of connections. They also proposed various strategies for improving secure connectivity, such as eavesdropper neutralization and sectorized transmission. Lastly, they analyzed the capability for secure communication in the presence of colluding eavesdroppers. To advocate outreach and diversity, the group is committed to attracting undergraduates and underrepresented minorities, giving them exposure to theoretical and experimental research at all levels. For example, the group has a strong track record for hosting students from both the Undergraduate Research Opportunities Program and the MIT Summer Research Program. Professor Moe Win maintains dynamic collaborations and partnerships with academia and industry, including the University of Bologna and Ferrara in Italy, University of Lund in Sweden, University of Oulu in Finland, National University of Singapore, Nanyang Technological University in Singapore, Draper Laboratory, the Jet Propulsion Laboratory, and Mitsubishi Electric Research Laboratories.
for tunnel wall interaction effects. Industrial testing has included auxiliary propulsion burner units, helicopter antenna pods, and in-flight trailing cables, as well as concepts for roofing attachments, a variety of stationary and vehicle mounted ground antenna configurations, the aeroelastic dynamics of airport control tower configurations for the Federal Aviation Authority, and the less anticipated live tests in Olympic ski gear, space suits for tare evaluations related to underwater simulations of weightless space activity, racing bicycles, subway station entrances, and Olympic rowing shells for oarlock system drag comparisons. In its more than 75 years of operations, Wright Brothers Wind Tunnel work has been recorded in hundreds of theses and more than 1,000 technical reports. WBWT faculty and staff include Professor Mark Drela and senior technical instructor Richard Perdichizzi. Visit the Wright Brothers Wind Tunnel at http://aeroastro.mit.edu/wbwt
Moe Win directs the Wireless Communication and Network Sciences Group. Visit the Wireless Communication and Network Sciences Group at http://wgroup.lids.mit.edu
WRIGHT BROTHERS WIND TUNNEL Since its opening in September 1938, the Wright Brothers Wind Tunnel has played a major role in the development of aerospace, civil engineering and architectural systems. In recent years, faculty research interests generated long-range studies of unsteady airfoil flow fields, jet engine inlet-vortex behavior, aeroelastic tests of unducted propeller fans, and panel methods
Lab report
75