Transforming the world of
ANNUAL REPORT 2013
MESSAGE FROM THE DIRECTOR The Robotics & Intelligent Machines Center (RIM) has a strong tradition of exceptional research, a robust interdisciplinary focus, and a solid reputation for innovation and entrepreneurship. After celebrating our seventh anniversary on June 13, we are poised to become a new Georgia Tech Interdisciplinary Research Institute (IRI) for robotics, which will be formed in fall 2013. In our new role, we will continue to design the next generation of robotics technologies and to prepare robotics students to be impactful contributors upon entering the high-tech workforce. As an IRI, we will support and enhance collaborations between robotics experts from Number of Robotics Faculty in June 2006:
across the entire Tech campus and will bridge connections to create new opportunities for our faculty with industry, government, and other research partners.
Number of Robotics Faculty in June 2013:
Henrik I. Christensen Distinguished Professor, KUKA Chair of Robotics & Director, Robotics & Intelligent Machines Center
Transforming the world of
MISSION Through our integrated research and participation in education programs, the Robotics & Intelligent Machines Center develops innovative solutions for manufacturing, service, first responders, and a variety of other areas. Using state-of-the-art facilities and working with strategic partners, we improve society by investigating novel robotics technologies that enhance the lives of everyone.
CONTENTS Research ...................................2 Industry Partnerships .............7 Education & Outreach.......... 10 National Robotics Week .........12 Seminars.................................13 Robotics Roadmap.................14 Robotics VO ...........................15 MOOCs...................................16 Spotlights................................17 Infrastructure ........................ 18 Faculty .................................... 21 Financial Report.................... 24
CONTRIBUTORS Director: Henrik I. Christensen Executive Director: Alex Goldstein Assoc. Director of Education: Frank Dellaert Assoc. Director of Research: James M. Rehg Asst. to the Director: Nina M. White Research Infrastructure: Daniel Walker Financial Administration: Joi Adams Communications: Josie G. Giles Photography: Rob Felt, Rory Fraser, Steve Jurvetson, Melanie Gonick, Gary Meek, Josh Meister, Zach Porter, and Raftermen Photography Additional photos courtesy of: BAE Systems, Cornell University College of Engineering, Ekso Bionics, Haughton Mars Project, Kiva Systems, NASA, Rethink Robotics, and the U.S. Department of Defense Copy Editing: Betsy S. Brown
CONTACT Robotics & Intelligent Machines Center Georgia Institute of Technology 801 Atlantic Drive NW Atlanta, GA 30332-0280 U.S.A. Phone: 404.385.8746 RIM@cc.gatech.edu www.robotics.gatech.edu
2013 RIM A N N U A L
R E P O R T
CHANGING THE WORLD: Strategic Research—Our researchers define the future role of robots in society for both personal and commercial applications.
Unparalleled Education—An international leader in education, Georgia Tech offers the first interdisciplinary PhD degree program in robotics.
Efficient Tech Transfer—We leverage strengths and resources by transcending boundaries to bring innovations to market more quickly.
BIOLOGICALLY INSPIRED DECEPTIVE ROBOTS
specific locations. The animal then patrols the hidden caches, routinely going back and forth to check on them. When another squirrel shows up, hoping to raid the hiding spots, the hoarding squirrel changes its behavior. Instead of checking on the true locations, it visits empty cache sites, trying to deceive the predator. Building upon an earlier study he conducted with Georgia Tech Research Institute Research Engineer Alan Wagner, Arkin implemented the same strategy into a robotic model and demonstration. The deceiving robot lured the
Using deceptive behavioral patterns of squirrels and birds,
“predator” robot to false locations, delaying the discovery of
researchers have developed robots that can deceive each
the protected resources. “This application could be used by
other. Led by Regents’ Professor Ronald Arkin, who suggests
robots guarding ammunition or supplies on the battlefield,”
the applications could be implemented in the future by the
says Arkin. “If an enemy were present, the robot could change
military, the research demonstrates several deceptive behav-
its patrolling strategies to deceive humans or another intelligent
iors, including squirrels gathering acorns and storing them in
machine, buying time until reinforcements arrive.”
Our researchers improve society by investigating novel robotics technologies that enhance the lives of everyone. HUMANIZED INTELLIGENT SYSTEMS The focus of the Human-Automation Systems (HumAnS) Lab is centered on the concept of humanized intelligence, the process of embedding human cognitive capability into the control path of autonomous systems. The group studies how human-inspired techniques, such as soft computing methodologies, sensing, and learning, can be used to enhance the autonomous capabilities of intelligent systems. The lab addresses issues of autonomous control as well as aspects of interaction with humans and the surrounding environment, using basic scientific and engineering methodologies to improve human-agent system performance and to create a synergistic method for decision making and task coordination in real-world scenarios. Led by Motorola Foundation Professor Ayanna Howard, the HumAnS Lab develops new technologies for applications in rehabilitation, therapy, education, and methods for engaging the novice user.
MULTI-RESOLUTION PATH PLANNING FOR AUTONOMOUS AGENTS The current practice in control system design is to develop control algorithms without consideration of the available hardware required for its implementation. Finding suitable control hardware to implement an algorithm is an after-the-fact task. For many autonomous vehicle problems that involve embedded control hardware with limited computational resources (both in terms of available CPU and memory), using this traditional method may not be possible. Research in the Dynamics and Control Systems Laboratory (DCSL) aims at reversing this line of thinking to introduce a new paradigm so the operational limitations of the hardware become part of the control algorithm design specifications. Led by Deanâ€™s Professor Panagiotis Tsiotras, the DCSL conducts research in control of aerospace, mechanical, and autonomous systems, including in specific areas such as energy storage and power generation in orbiting spacecraft, design and construction of unmanned aerial vehicles, and improving safety control systems in automobiles. 3
The Robotics & Intelligent Machines Center at Georgia Tech designs robots that empower people. ROBOTS FOR HUMANITY The Healthcare Robotics Lab leads the Robots for Humanity project to explore ways for Henry Evans, a mute quadriplegic, to use a PR2 robot as his surrogate. After suffering a stroke when he was just 40 years old, Evans went through extensive therapy and has now regained the ability to move his head and use a finger, which allows him to operate computers. Associate Professor Charles Kemp leads a team of researchers actively researching ways for Evans and others to perform simple tasks on a daily basis that they cannot accomplish by themselves. Using a head tracker to operate a variety of experimental user interfaces, Evans directly moves a PR2 robotâ€™s body, which allows him to invoke autonomous actions, such as navigating through a room or intelligently maneuvering within clutter, gently making contact with objects while accomplishing a task.
YOUTUBE STABILIZATION New research in video stabilization helps even the most inexperienced novice automatically create more professional videos for YouTube. Using a novel algorithm that automatically applies constrainable, L1-optimal camera paths, Professor Irfan Essa and his research team have developed a technique to generate stabilized videos with ease. Their method removes undesirable motions beyond the conventional filtering of camera paths, which only suppresses high frequency jitter. By incorporating additional constraints on the path of the camera directly in the algorithm, their algorithm computes camera
the need for user interaction or costly 3D reconstruction of the
paths composed of constant, linear, and parabolic segments,
scene and works as a post-process for videos captured by any
thus mimicking the camera motions employed by professional
camera or from an online source.
cinematographers. The approach achieves stabilization without 4
MICRO AUTONOMOUS SYSTEMS AND TECHNOLOGY military operations. The five-year MAST program, led by BAE Systems, includes collaborators from Georgia Tech and other U.S. universities. Several RIM faculty members participate in the research, including Ronald Arkin, Henrik Christensen, Tom Collins, and Frank Dellaert. One initiative involves a group of robots working together entirely without human input, communicating with one another to coordinate their movements. A video camera on each robot identifies doorways and windows, while a laser scanner measures walls. An onboard inertial Imagine the potential to save human lives if first responders
measurement unit (IMU) keeps track of their movements and
could use autonomous robots to quickly map out unknown
aids stability. A graph-based system called simultaneous
territory and discover potential threats. The U.S. Armyâ€™s Micro
localization and mapping (SLAM) allows each robot to build a
Autonomous Systems and Technology (MAST) Collaborative
local area map using its sensory data, while also keeping track
Technology Alliance Program envisions this scenario by using
of its own location within that map. After completing the
swarms of small rolling, hopping, crawling, or flying robots
evaluation of the landscape, a detailed floor map is then
working together on reconnaissance missions in civilian or
transmitted to a nearby human research team.
HUMAN-ROBOT INTERACTION Using Simon as her student, Associate Professor Andrea Thomaz redefines how robots and humans interact. She envisions a not-so-distant future where anyone can purchase a robot, take it home, and instruct it to do virtually any task. To make them more intuitive and intelligent, robots must first learn to think like their human counterparts. They must also quickly adapt and learn to perform tasks that most benefit the unique environment in which they are placed. To achieve these goals, Simon interacts with just about anyone who wants to know him better and teach him new skills. Lessons include everything from clearing the dinner table to sorting objects by color. Based on the results, Thomaz and her team of researchers tweak Simonâ€™s algorithms to make him a more efficient communicator and learner. With a growing number of robots being used in hospitals, and the promise of them one day providing in-home assistance to the elderly and disabled, potential healthcare applications for robots that can be trained by anyone, in any setting, are virtually limitless. Research in human-robot interaction (HRI) moves the world closer to personal robots becoming as common in U.S. homes, offices, and hospitals as any major appliance. 5
GTSAM GTSAM is a library of C++ classes that implement smoothing and mapping (SAM) in robotics and vision, using factor graphs and Bayes networks as the underlying computing paradigm rather than sparse matrices. Developed by Professor Frank Dellaert, GTSAM uses incremental smoothing and mapping (iSAM), a novel approach to a simultaneous localization and mapping problem based on fast incremental matrix factorization. iSAM provides an efficient and exact solution by updating a QR factorization of the naturally sparse smoothing information matrix, therefore recalculating only the matrix entries that actually change. Additionally, iSAM is efficient even for robot trajectories with many loops as it avoids unnecessary fill-in of the factor matrix by periodic variable reordering. Also, to enable data association in real-time, the approach provides efficient algorithms to access the estimation uncertainties of interest based on the factored information matrix. Using various simulated and real-world datasets for both landmark and pose-only settings, the research allows for a systematic evaluation of the different components of iSAM, as well as of the overall algorithm.
UNDERGROUND ROBOT LOCOMOTION Future teams of subterranean search and rescue robots may owe their success to the fire ant, an insect that builds extensive networks of underground tunnels. By studying these ants in the laboratory using video tracking equipment and X-ray computed tomography, researchers have uncovered fundamental principles of locomotion that robot teams could one day use to travel quickly and easily through underground tunnels. Among the principles is building tunnel environments that assist in moving around by limiting slips and falls, and by reducing the need for
complex neural processing. Led by Associate Professor Daniel
because the problems ants face are the same kinds of prob-
Goldman, a recent study revealed that ants in confined spaces
lems a digging robot working in a confined space would
use their antennae for locomotion as well as for sensing the
potentially faceâ€”the need for rapid movement, stability, and
environment. While the research focused on understanding the
safetyâ€”all with limited sensing and brain power. Incorporating
principles behind how ants move in confined spaces, the
controls that mimic the one ants have been given by nature
results could have implications for future teams of small robots
could enhance the future design of robotic digging machines.
Industry Partnerships Our corporate research sponsors are well positioned to take advantage of new technologies as they are developed in the Robotics & Intelligent Machines Center. They not only provide vital financial support, but they also play an active role in RIM’s strategy to efficiently deliver innovative concepts from the lab to the marketplace. RIM’s unique interdisciplinary research environment provides our industry partners with unprecedented access to some of the world’s best computer scientists, engineers, and other researchers who are committed to collaboration.
THE WORLD NEEDS BETTER ROBOTS: Benefits—Our partnership program allows members to explore opportunities for research collaborations, testing services, consulting relationships, student hiring, and other valuable interactions.
Results—We have engaged in numerous exciting and productive partnerships with industry partners. The following profiles represent a snapshot of some of our recent projects.
Georgia Tech received a 2012 Boeing Performance Excellence Award, and in 2011, Tech was- one of only 16 organizations RIM researchers named a Boeing Supplier of the Year. engaged in
BOEING ing of fluid flow, advanced manufacturing technology, design,
industrial research partnerships in FY2013.
and aircraft technology through basic and applied research. Boeing supports various research activities at Georgia Tech related to manufacturing technologies, such as control systems on cranes, mobile platforms and robotics for moving parts in a factory environment, and active flow control for wing tips. Numerous professors from the College of Engineering, the College of Computing, the Georgia Tech Manufacturing Boeing consistently recognizes Georgia Tech for its exception-
Institute, and the Georgia Tech Research Institute conduct
al performance and contributions to the company’s overall
research that directly impacts Boeing’s work, including Henrik
success. As one of Boeing’s eight strategic universities,
Christensen, Steve Danyluk, Ari Glezer, Dimitri Mavris, Leon
Georgia Tech provides increased knowledge and understand-
McGinnis, and William Singhose.
PSA PEUGEOT CITROËN Known as an OpenLab, Georgia Tech-CNRS (UMI 2958) is one of three academic partners with PSA Peugeot Citroën. Selected for its recognized capabilities in robotics and innovative materials and processes, including nanomaterials, surface treatments, and composites, the Materials & Processes OpenLab is coordinated by UMI GT-CNRS and housed in Georgia Tech-Lorraine facilities in Metz, France. Creation of the OpenLab is part of PSA Peugeot Citroën’s strategic plan to remain on the leading edge of automotive products and services. Research with Georgia Tech and its other partners make it possible for the company to explore major themes concerning the future of the automobile and to closely track the latest scientific discoveries worldwide. 8
GOOGLE Thad Starner, a pioneer of wearable computing and a member of the development team charged with creating Google Glass, has been working on the project since its inception. The glasses-like device includes a tiny computer attached to the right earpiece allowing users to access the Internet, take photos, film videos of the world around them, and then broadcast the content in seconds. Although the product has raised controversy, from privacy concerns to safety issues, Google Glass offers an equal amount of benefits to the ways people may use the device in their everyday lives, including potential applications as assistive devices for those with disabilities such as muscular dystrophy and Parkinson’s disease.
Working with Yaskawa Motoman, robotics researchers aim to improve the utility and performance of flexible motion systems by enabling high-speed precision manipulation through advanced control methods. To mitigate the undesirable effects of vibration, these systems are often operated well below their capable operating speeds, limiting efficiency and hindering functionality. RIM researchers develop new applications for two-armed manipulation systems for manufacturing, logistics, and service industries, especially in industrial efforts using the Robot Operating System (ROS), an
open-source, BSD-licensed set of libraries and tools for robot applications.
BMW BMW’s research partnership with Georgia Tech provides new impetus for the company’s production and assembly locations worldwide. The collaboration enables both BMW and Georgia Tech to exploit their strengths in engineering and production systems to prepare for increasing internationalization, specifically in the fields of new drive technologies, materials and substances, and supply chain management. Robotics research, especially human-robot collaboration in manufacturing, plays an important role in this partnership, as automation is integral to maximizing efficiency and reducing costs. Tech researchers also contribute to further developments of Mobileye, a unique artificial vision technology that helps drivers by acting as a “third eye,” constantly monitoring objects in front of a vehicle. 9
Education & Outreach
INTERDISCIPLINARY PHD PROGRAM IN ROBOTICS
Georgia Tech offers the first interdisciplinary PhD degree program in robotics for students
enrolled in a participating home school in either the College of Computing or
2012-2013 Enrollment: COC
entering the high-tech workforce. RIM serves as the flagship for
ENROLLMENT BY COLLEGE
Techâ€™s robotics efforts; therefore, the research center has an
researchers who are prepared to be impactful contributors upon
E n g i n e e ri
experience, the program educates a new generation of robotics
m p u ti n g
the College of Engineering. A fully integrated and multidisciplinary
New Students (Fall 2013):
integral relationship with the program, and almost all RIM faculty members serve as research advisors to students pursuing the degree. The program supports Techâ€™s mission to provide instruction in disciplines related to science, technology, and interdisciplinary areas. In addition to performing original research culminating in a doctoral thesis, students complete 36 hours of coursework and focus on three of six core robotics areas: Mechanics, Controls, Perception, Human-Robot Interaction (HRI), Autonomy, and Artificial Intelligence. 10
RIM’s impact extends far beyond the borders of the main Atlanta campus to encompass robotics education and research in international locations. BARCELONA SUMMER STUDY ABROAD PROGRAM Sponsored by the College of Computing, the Barcelona Summer Study Abroad Program is held on the Universitat Politècnica de Catalunya (UPC) campus and typically includes courses from the schools of Architecture, Computer Science, Literature, Media & Communication, and Modern Languages. RIM professors, including Irfan Essa and Frank Dellaert, participate in the program by leading robotics-related courses, including CS 4001: Computing & Society, CS 4464: Computational Journalism, and CS 4475: Computational Photography.
GEORGIA TECH–LORRAINE Georgia Tech-Lorraine (GTL), located in Metz, France, offers undergraduate courses and an extensive graduate program, including options leading to the designated master’s degrees in computer science, electrical engineering, and mechanical engineering, the undesignated master’s (MS), and the PhD degree from Tech. In addition to these degrees, cooperative agreements with local partner institutions enable GTL graduate students to pursue dual degrees in engineering and science. RIM professor Cedric Pradalier maintains a lab on the GTL campus, the Data-driven Robotics for Environment Assessment and Monitoring (DREAM) Lab, where he conducts research on developing mobile robotic systems to monitor environments over long periods of time, in particular in environments where change happens over a wide spectrum of different time scales. Pradalier also leads courses in robotics as part of the computer science degree programs, including CS 7630: Autonomous Robots. Additionally, RIM professors Frank Dellaert and Beki Grinter also teach courses on the GTL campus. 11
Since the launch of National Robotics Week in 2010, RIM has participated annually in the event by sponsoring an open house and conducting lab tours and demonstrations for students in middle school and high school. More than 400 students participated in RIMâ€™s 2013 event held on April 11, with one group traveling from Tennessee to attend. Tours were held in multiple locations, offering participants an opportunity to learn more about research projects in 16 different robotics labs. National Robotics Week was established by a U.S. House of Representatives resolution to recognize robotics technology as a pillar of the 21st century and to leverage its ability to inspire technology education. 12
RIM Seminar Series Each year, RIM hosts a seminar series featuring guest speakers from around the world who are leaders in robotics.
FALL 2012 Robert J. Webster Vanderbilt University Jan-Michael Frahm University of North Carolina at Chapel Hill Nathan Michael Carnegie Mellon University J. Sean Humbert University of Maryland-College Park Ashutosh Saxena Cornell University Mark Dredze Johns Hopkins University Alberto Rodriguez Carnegie Mellon’s Robotics Institute
SPRING 2013 Dan Moroni Liquid Robotics Raquel Urtasun Toyota Technological Institute at Chicago Dennis Hong Virginia Tech Kristen Grauman University of Texas at Austin Tim Barfoot University of Toronto Tomás Lozano-Pérez Massachusetts Institute of Technology Jonathan Hurst Oregon State University 13
Robotics Roadmap Building on the highly successful initial “A Roadmap for U.S. Robotics,” published in 2009, a new and updated report for 2013 outlines the progress of robots in multiple industries over the last five years, identifies goals for the coming decade, and emphasizes the importance of the robotics research pipeline for maintaining U.S. innovation. First announced by President Obama in 2011, the National Robotics Initiative was officially established in 2012 by a joint initiative of the National Science Foundation, the National Institutes of Health, the National Aeronautics and Space Administration, and the United States Department of Agriculture.
Highlighting robotics as a key economic enabler, the 2013 roadmap explores the potential of robotics technology to transform U.S. society by developing new markets and industries, creating new jobs, and addressing a number of issues of national importance. RIM Director Henrik Christensen presented the report to the Congressional Robotics Caucus on March 20, 2013.
“Robots have the potential to bring manufacturing jobs back to the U.S., to improve our quality of life, and to make sure our first responders and warfighters stay safe,” said Christensen. “We hope this report will help foster the discussion on how we can build partnerships and allocate resources to move the robotics industry forward.” 14
President Obama launched the National Robotics Initiative as part of a broader effort to promote a renaissance of American manufacturing.
Key Activities & Responsibilities Maintenance and promotion of a roadmap for robotics in the U.S.
Launched in 2012, the Robotics Virtual Organization (VO) is an online consortium that engages the community of researchers, corporations, and program managers who are actively engaged in the National Robotics Initiative to ensure unparalleled robotics research and development in the United States. Originally formed with sponsorship by the National Science Foundation, the creation of the Robotics VO network was spearheaded by RIM Director Henrik Christensen and is now self-sufficient, serving as a catalyst to maximize the impact of robotics-related activities in the U.S. and facilitate the development of next-generation robotics applications. In addition to providing a website that serves as a definitive one-stop source for educational materials, technology RO
roboBOTIC S tics -vo.u VO s
transfer opportunities, and
Consultation to agencies and companies on the utilization of robotics.
Organization of educational initiatives from K-12 to graduate education to secure access to the human capital needed for continued economic growth involving robotics in a broad sense. (This includes educational resources, outrea ch, and topical efforts such as summer schools.) Study mechanisms and bes t practices for transition of technology from basic res earch to commercial applications, including issu es such as standardization and utilization of commo n platforms. Organize and secure mecha nisms for dissemination of information about nation al robotics initiatives to society at large, academia, and industry. Provide resources to organiz e and support special interest groups (SIGs) in term s of studies and coordination of topics. (This primarily focuses on coordinating resources suc h as a web services, mailing lists, and discussion forums.)
relevant news, the consortium will host its first annual meeting in October 2013, where leaders will present on cutting-edge research and participate in workshops to bridge connections in the robotics community. 15
July 17, 2013 marked the one-year anniversary of Georgia Tech’s partnership with Coursera, one of the premier companies offering massive open online
courses (MOOCs). Founded by Stanford University professors Andrew Ng and Daphne Koller, Coursera partners with top universities and organizations around the world to offer online courses at no cost by using technology that enables instructors to teach thousands of students in a single course session. RIM faculty members played a prominent role in Tech’s debut on Coursera by developing three of the Institute’s first MOOCs. Magnus Egerstedt offered Control of Mobile Robots, Irfan Essa taught Computational Photography, and Tucker Balch led Computational Investing, Part 1, which began in August 2013, for the third time. Over the last year, Tech has offered 15 different courses in a variety of disciplines on the Coursera platform with a combined enrollment of more than 400,000 students. Currently, Coursera has almost 4.5 million students, offers more than 400 courses, and partners with 75 global institutions and 10 U.S. university systems, including the University System of Georgia. Learn more about Tech’s Coursera MOOCs: coursera.org/gatech. In May 2013, Tech’s College of Computing announced it had teamed up with Udacity (another premier MOOC provider founded by Sebastian Thrun, David Stavens, and Mike Sokolsky) and AT&T to offer the first accredited Online Master of Science in Computer Science (OMS CS) that students can earn exclusively through the MOOC format and for a fraction of the cost of traditional, on-campus programs. RIM faculty member Charles Isbell Jr. is slated to teach one of the program’s first courses. Learn more about the OMS CS degree: omscs.gatech.edu. 16
Spotlights SURE PROGRAM Motorola Foundation Professor Ayanna Howard received a National Science Foundation (NSF) Research Experiences for Undergraduates (REU) Summer Undergraduate Research in Engineering (SURE) continuing grant to add a robotics component to Georgia Tech’s SURE program. Funded with co-support from the Department of Defense and NSF’s Division of Engineering Education and Centers, the program supports undergraduate students in an immersive, 10-week summer robotics research experience designed to attract qualified minority students into graduate school in the fields of engineering and science. Scheduled to launch in May 2014, eight RIM faculty members will serve as initial program advisors: Ayanna Howard, Charles Isbell Jr., Charles Kemp, Mike Stilman, Andrea Thomaz, Jun Ueda, Patricio Vela, and Fumin Zhang. Learn more about the program: sure.gatech.edu.
TED-ED & TEDx EMORY
Two RIM faculty members gave talks in conjunction with TED, a nonprofit devoted to “ideas worth spreading,” especially those related to education and technology. In November 2012, Ayanna Howard presented “Making Robots Smart(er)” as part of the annual TEDYouth event in New York City. In April 2013, Henrik Christensen presented “Robots for Everyone” at Emory University in Atlanta as part of TEDx Emory, one of the many independently organized TED events held worldwide. Watch the talks online: tinyurl.com/TED-Howard and tinyurl.com/TED-Christensen.
INVENTURE PRIZE ChewBots are toy robots that drive dogs crazy by wiggling around and running away. Invented by Chris Taylor, who recently graduated with a BS in Mechanical Engineering, ChewBots placed first in Georgia Tech’s fourth-annual undergraduate student InVenture Prize competition with Taylor receiving $20,000 for his invention. Taylor enrolled in fall 2012 in Tech’s Startup Semester program for students who have entrepreneurial aspirations. It was then that he built his first prototypes for ChewBots. As Taylor explained to the InVenture judges, the dog toy industry in the U.S. generates about $50 billion in sales every year, yet there has been little innovation in the market. ChewBots will soon change that. Sign up for product updates: chewbots.co. 17
labs conduct roboticsrelated research at Tech.
HOME SERVICE ROBOT & COMPANION Housed in the Socially Intelligent Machines Lab, Simon, a social robot, uses 37 degrees of freedom in total (a 13-dof head, a 2-dof torso, two 7-dof arms, and two 4-dof hands) and communicates with humans using speech, gestures, and eye gaze. Research in this lab envisions enabling robots to function in dynamic human environments by allowing them to flexibly adapt their skill set via learning interactions with end-users. Research with Simon explores the ways in which machinelearning agents can exploit principles of human social learning.
To date, work with socially guided machine learning (SG-ML) in
research thrusts: (1) Interactive Machine Learning and (2)
the Socially Intelligent Machines Lab has focused on two
Natural Interaction Patterns for Human-Robot Interaction (HRI).
We leverage Georgia Techâ€™s strengths and resources by reaching across traditional boundaries to embrace a multidisciplinary approach to robotics research. STING RACER: SELF-DRIVING CAR Built from a Porsche Cayenne, Sting Racer is controlled by eight dual Pentium computers running Ubuntu Linux installed in the trunk space. For navigation, tracking other vehicles, and situation assessment, Sting is equipped with Novatel GPS sensors, an inertial measurement unit (IMU), six cameras, nine laser scanners (Riegl & SICK LMS), and seven 22 GHz radar units (EATON). Originally designed for the 2007 DARPA Urban Design Challenge, Sting continues to provide a platform for ongoing research in autonomous vehicle enhancements, especially advanced sensing mechanisms that react to urban traffic environments and improved artificial intelligence for automatically determining best routes and right-of-way in normal traffic. Ultimately, the development of new autonomous vehicle technologies may lead to a driverless car approved by the Department of Transportation for personal use on U.S. interstates.
MOBILE MANIPULATORS The Humanoid Robotics Lab develops dexterous mobile robots, known as Golems, which resemble human beings. By developing algorithms for planning and control, this research group enables robots to autonomously perform human tasks. Recent significant advances include designing platforms for mobility and manipulation and developing robots that look, think, and act in a way that people perceive as human or intelligent. Golem Hubo will compete in the next DARPA Robotics Challenge, focused on humanitarian, disaster relief, and related operations, while Golem Krang continues to garner attention for its ability to solve complex problems and escape dangerous situations by using everyday objects and materials found at hand. 19
INDUSTRIAL MANIPULATORS The Bio-Robotics and Human Modeling Laboratory focuses on coordinated design of robotic components, such as mechanical and electrical systems. Several recent projects involve developing better industrial manipulators by overcoming some of the inherent limitations that result when humans and machines work together. For example, the group is investigating methods for effectively adjusting the controller gains of an automated lift device based on the muscle stiffness of an operator. The data collected during these experiments will lead to better control of various human-machine interfaces in assembly lines.
AUTONOMOUS FLIGHT PLATFORMS RIM researchers develop advanced methods for vision-aided inertial navigation systems that enable autonomous ďŹ‚ight of aerial vehicles (AVs). Small AVs are often incapable of carrying large computer systems, which presents limitations for mapping and navigation when GPS is unavailable. Research has demonstrated a viable solution by using a combination of a monocular vision sensor and an inertial measurement unit (IMU). Simulations and real-time ďŹ‚ight test results demonstrate an autonomous closed-loop control of an aircraft in a loss-of-GPS scenario, which is an invaluable contribution to the future development of AVs for both military and civilian applications.
HEALTHCARE ROBOTICS The growing need in society to enable older adults to remain in an independent living environment makes the research area of assistive mobile manipulation critically important. A team of researchers in the Healthcare Robotics Lab works on projects in this area, including one devoted to human-robot interaction and another focused on software development. Highly synergistic, both projects focus on the development of assistive capabilities for the PR2 robot, with an emphasis on in-home care for older adults. The human-robot interaction (HRI) component ensures that software development is closely connected to real-world needs and will provide capabilities that both inform and enable cutting-edge studies of human-robot interaction. 20
Faculty INNOVATIVE THINKERS: Multidisciplinary Teamâ€”More
Our faculty members have technically diverse backgrounds and conduct innovative research to advance robotics and promote its increasing importance in daily life. Robotics research at Georgia Tech
than 50 faculty members, hailing from four colleges at Georgia Tech plus the Georgia Tech Research Institute and the Enterprise Innovation Institute (EI2), actively participate in roboticsrelated research and education.
attracted more than $16 million in sponsored research last year. Core
Prolific Researchâ€”RIM faculty
issues surrounding their governance and ethical use. Automation and
members produced more than 100 research papers and conference presentations during the past year.
research areas include mechanisms, control, perception, artificial intelligence, human interaction, and system application technologies. We advance the field of robotics through our research into the ways robots can learn from and interact with humans and by exploring personal robotics are key growth drivers for U.S. and international economies, and our interdisciplinary approach and strength in these areas position Tech as a research leader.
Review more than 1,000 publications by RIM faculty members: smartech.gatech.edu/handle/1853/36697 21
ROBOTICS FACULTY Distribution of Faculty Members by Unit 18
21 College of Engineering
College of Computing
Georgia Tech Research Institute
College of Sciences
PROMOTIONS & APPOINTMENTS
Frank Dellaert: professor, School of Interactive Computing
Stephen P. DeWeerth: associate dean, College of Engineering
Magnus Egerstedt: Schlumberger Professor, School of Electrical & Computer Engineering
Daniel Goldman: associate professor, School of Physics
Ayanna Howard: Motorola Foundation Professor, School of Electrical & Computer Engineering
Charles Isbell Jr.: senior associate dean, College of Computing
Charles Kemp: associate professor, Department of Biomedical Engineering
Thad Starner: professor, School of Interactive Computing
Andrea Thomaz: associate professor, School of Interactive Computing
Anthony J. Yezzi: Ken Byers Professor, School of Electrical & Computer Engineering
Fumin Zhang: associate professor, School of Electrical & Computer Engineering
Tech has more than
robotics professors and researchers.
AWARDS & HONORS ▶
Ronald Arkin: named a 2012 Distinguished Lecturer by the IEEE Society of Social Implications of Technology
Henrik Christensen: elected to the College Industry Council for Material Handling
Magnus Egerstedt and Irfan Essa: named Institute of Electrical and Electronics Engineers (IEEE) Fellows
Daniel Goldman: 2012 DARPA Young Faculty Award
Beki Grinter: inducted into 2013 ACM CHI Academy
Mark Guzdial: 2012 IEEE Computer Society Undergraduate Teaching Award
Charles Kemp, Karen Liu, and Lena Ting: NSF’s Division of Emerging Frontiers in Research and Innovation (EFRI) award to develop a “therapeutic robot” to help rehabilitate people with mobility problems
Karen Liu: 2012 ACM Significant New Research Award
James Rehg: 2012 Method of the Year Award from the Journal of Nature Methods
Thad Starner: “Invention of the Year” by Time magazine for Google Glass; Georgia Tech Class of 1934 Outstanding Innovative Use of Education Technology Award
Mike Stilman: 2013 DARPA Young Faculty Award
Andrea Thomaz: named one of 2012’s “Brilliant 10” by Popular Science magazine
Alan Wagner: Air Force Office of Scientific Research Young Investigator Program (AFOSR YIP) award
Financial Report: FY2013 Highlights In FY2013, RIM received more than $16M in grants and contracts from a wide variety of government agencies and industry partners. About 75% of awarded grants were overseen by tenured or tenure-track faculty members. Funded projects represented all of the five key robotic research areas: manufacturing, healthcare, service, defense, and space.
Distribution Across Agencies/Industry 10% Other
5% U.S. Tech Army
U.S. Air Force
Industry 10% 15% Force
15% 15% U.S. Army
Distribution Across Campus Units
13% Research Institute Georgia Tech
College of Sciences
46% Engineering College of
COE 32% Computing
LEADERSHIP Henrik I. Christensen
RIM Director, Distinguished Professor & KUKA Chair of Robotics, School of Interactive Computing Human-Centered Robotics, Sensory/Data Fusion, Systems Integration
Irfan A. Essa
Randy D. Trumbower
Karen M. Feigh
Craig A. Tovey
Director of Off-Campus Initiatives & Professor, School of Interactive Computing Human-Computer Interaction, Computer Vision, Computational Perception
Assistant Professor, School of Aerospace Engineering Computational Cognitive Modeling
RIM Executive Director, School of Interactive Computing Strategic Planning, Development of Outreach and Research Initiatives, Technology Transfer
Ashok K. Goel
Daniel I. Goldman
RIM Associate Director of Education & Professor, School of Interactive Computing Robot Mapping & 3D Reconstruction, Probabilistic Inference, Computer Vision
James M. Rehg
RIM Associate Director of Research & Professor, School of Interactive Computing Machine Distributed Computing, Computer Vision, Computer Graphics
Professor, School of Interactive Computing Artificial Intelligence, Computational Design & Computational Creativity
Associate Professor, School of Physics Biomechanics of Locomotion of Organisms and Robots
Alexander G. Gray
Associate Professor, School of Computational Science & Engineering Data Analytics, Machine Learning & Visualization, Modeling & Simulation
Rebecca E. Grinter
FACULTY Ronald C. Arkin
Associate Dean & Regentsâ€™ Professor, School of Interactive Computing Multiagent Autonomous Systems, Human-Robot Interaction, Robot Ethics
Tucker R. Balch
Associate Professor, School of Interactive Computing Large-scale Multiagent & Robotic Systems, Distributed Learning & Planning, Cooperation
Professor, School of Interactive Computing Intelligent Systems, Computer Graphics & Visualization
Wayne J. Book
Professor Emeritus, School of Mechanical Engineering Automation & Mechatronics, Acoustics & Dynamics, Modeling, Fluid Power
Principal Research Engineer, Georgia Tech Research Institute Autonomous Operations Processing
Professor, School of Interactive Computing Human-Computer Interaction, Computer Supported Cooperative Work, Human-Robot Interaction
Wiley D. Holcombe
Senior Research Engineer, Georgia Tech Research Institute Automated Systems
Ayanna MacCalla Howard
Motorola Foundation Professor, School of Electrical & Computer Engineering Human-Inspired Control, Space Robotics, Perception & Reasoning
David L. Hu
Assistant Professor, School of Mechanical Engineering Biomimetic Technologies, Locomotion of Animals for Biologically Inspired Robots
Charles L. Isbell Jr.
Senior Associate Dean & Professor, School of Interactive Computing Adaptive Interfaces & Programming, Computational Narrative, Machine Learning, Multiagent Systems, Pattern Analysis
Eric N. Johnson
Principal Research Engineer, Georgia Tech Research Institute Machine Vision Systems
Lockheed Martin Associate Professor of Avionics Integration, School of Aerospace Engineering Digital Avionics Systems, Flight Control, Navigation
Stephen P. DeWeerth
Charles C. Kemp
Associate Dean & Professor, School of Electrical & Computer Engineering Sensor Systems & Motor Learning, Remote Interfacing, Embedded Systems
Schlumberger Professor, School of Electrical & Computer Engineering Networked Control, Hybrid & Embedded Control Systems
Associate Professor, Department of Biomedical Engineering Healthcare Robotics, Autonomous Mobile Manipulation, Human-Robot Interaction
Research Scientist II, Georgia Tech Research Institute Human-Robot Interaction, Machine Learning, Distributed Sensing, Computer Vision
Professor, School of Mechanical Engineering Automation & Mechatronics, System Dynamics, Control, Automation
Associate Professor, School of Mechanical Engineering Automation & Mechatronics, Computer-Aided Engineering & Design, Mechanical Systems
Cheng-Yun Karen Liu
Associate Professor, School of Interactive Computing Computer Graphics & Animation, Optimal Control, Numerical Methods, Computational Biomechanics
Gary V. McMurray
Division Chief, Food Processing Technology, Georgia Tech Research Institute Robotic Systems for Food, Transportation & Biomedical Industries
Adjunct Professor, School of Interactive Computing Mobile Robotic Systems to Monitor Environments
Steven M. Potter
Associate Professor, Department of Biomedical Engineering In Vitro Learning & Memory, Brain-like Computing Systems
Professor, School of Aerospace Engineering Sensing & Active Control of Aerospace Systems
Mark O. Riedl
Assistant Professor, School of Interactive Computing Artificial Intelligence, Interactive Narratives, Autonomous Agents
Engineering Project Manager II, Enterprise Innovation Institute (EI2) Industrial Automation & Machine Building
Thad E. Starner
Professor, School of Interactive Computing Wearable Computing, Perceptually Enabled Intelligent Agents
Assistant Professor, School of Interactive Computing Humanoid Robotics, Autonomous Mobile Manipulation
Andrea L. Thomaz
Associate Professor, School of Interactive Computing Artificial Intelligence, Human-Robot Interaction, Machine Learning
Lena H. Ting
Associate Professor, Department of Biomedical Engineering Biomechanics, Neural Movement Control
Adjunct Assistant Professor, School of Applied Physiology Robotics for Neuromotor Impairments Following Central Nervous System Injury
Professor, School of Industrial & Systems Engineering Algorithms for Robot Motion Planning
Deanâ€™s Professor, School of Aerospace Engineering Autonomous Systems, Nonlinear & Optimal Control
Assistant Professor, School of Mechanical Engineering Automation & Mechatronics, Automatic Control
Emeritus Professor, School of Electrical & Computer Engineering Semiautonomous Systems, Control of Unmanned Aerial Vehicles
Patricio A. Vela
Associate Professor, School of Electrical & Computer Engineering Control Theory, Computer Vision
Alan R. Wagner
Research Engineer II, Georgia Tech Research Institute Social Robotics, Human-Robot Interaction, Robot Deception & Trust
Associate Professor & Director of Center for Music Technology, School of Music Robotic Musicianship
Lora G. Weiss
Chief Scientist & Technical Director, Autonomous Systems, Georgia Tech Research Institute Autonomous & Unmanned Systems
Michael E. West
Principal Research Engineer, Georgia Tech Research Institute Autonomous Systems, Underwater Robotics, Navigation & Control
Anthony Joseph Yezzi
Byers Professor, School of Electrical & Computer Engineering Computer Vision, Object Recognition, Visual Tracking
Associate Professor, School of Electrical & Computer Engineering Mobile Sensor Networks, Autonomous Underwater Vehicles, Motion Planning in Complex Environments
Associate Chair for Undergraduate Studies & Associate Professor, School of Industrial & Systems Engineering Manufacturing Systems, Facility Layout, Automation
Henrik I. Christensen, Director 801 Atlantic Drive NW Atlanta, GA 30332-0280 U.S.A. Phone: 404.385.8746 firstname.lastname@example.org www.robotics.gatech.edu Copyright 2013 • Georgia Institute of Technology • Robotics & Intelligent Machines Center (RIM) • An equal education and employment opportunity institution