Page 1

ISSUE 2020


Imprint INFORM GmbH Pascalstr. 35, 52076 Aachen, Germany info@inform-software.com inform-software.com Picture credits Cover - © Remo Zehnder Page 02 - © Remo Zehnder Page 03 - © Ernst Alexander Page 06 - © macrovector / Adobe Stock Page 10 - © Remo Zehnder Page 13 - © Remo Zehnder Page 14 - © Ernst Alexander Page 18 - © INFORM Page 20 - © INFORM Illustration Page 22 - © MariiaDemchenko / Adobe Stock Page 24 - © Vadarshop / shutterstock.com Page 26 - © PR Image Factory / shuttertock.com Page 28 - © SFIO CRACHO / shutterstock.com Page 29 - © Lonely Crane / Adobe Stock Page 31 - © Henry Olden + Macrovector / shutterstock Page 34 - © WrightStudio / Adobe Stock Page 38 - © Castleski / shutterstock.com Page 40 - © INFORM Illustration Page 41 - © Gorodenkoff / shutterstock.com Page 42 - © Hybrid Images / gettyimages.com Page 43 - © Birdlkportfolio / gettyimages.com Page 44 - © spainter_vfx / shutterstock.com Page 45 - © Claudiad / gettyimages.com

Welcome Artificial intelligence has emerged strongly over the past two years and is seemingly everywhere we look. Across the logistics sector, the conversation has shifted from “if” AI will play a role in your organization to “how.” Machine learning, as a subset of AI, too, is now playing a crucial role. For post and parcel operators, this is all too real with the likes of Amazon and Alibaba having forged a path forward leveraging the technology to revolutionize commerce and e-commerce, and now their sights are firmly set on delivery in many locations the world over. We also see the world pivoting toward Industry 4.0 and automation. What started in manufacturing and moved into parts of the logistics chain, chiefly maritime terminal operations, is now forging a path still further through the logistics industry. This is due in part to automated hardware decreasing in cost, in part to it becoming easier to implement, and in part to skilled labor shortages. Automation across the post and parcel sector stands out as the most significant opportunity for the sector in the decade to come. Combined, the advances in AI and automated hardware will lead to truly innovative operational models. We’ve quietly been delivering automation across logistics for nearly three decades now, and we’ve built that experience into our AI empowered Yard Management Solution. For those willing to think differently about how to solve the challenges our industry faces today, tomorrow is sure to be bright, and we look forward to taking that step together with you.

Dr. Eva Savelsberg Senior Vice President Logistics Division


14 INTERVIEW What do you think Eva Savelsberg?

06 Automated Vehicles

18 Demystifying AI

10 ROI for your YMS

13 SUCCESS STORY Swiss Port – Yard Management System

22 Millennials and Stamps


Our new “Industry Insights” section pulls together learnings from other logistics industries providing a blend of outside the box thinking mixed with real world applications.

34 INDUSTRY INSIGHTS Machine Learning in Terminal Operations

New Section

38 INDUSTRY INSIGHTS Digital Twins Technology


42 INDUSTRY INSIGHTS Shifting the Dynamics of Workforce Management


26 Tech Talk – Emerging Tech

46 INFORM – IT Systems for Intelligent Decisions

30 To Boldly go


AUTOMATED VEHICLES Reshaping the Logistics Industry in 2020 By Dr. Eva Savelsberg, Markus Sekula, and Matthew Wittemeier


Having grown up in a small town in rural America, I had a few friends who were farmers. When harvest time came around each year, I remember hearing about their fancy combines guided by GPS. “I just sit there in the air-conditioned cab and play on my phone while the combine goes back and forth across the field.” This was years ago, and a self-driving combine was a luxury since so much of their day-to-day work outside of this new machinery was arduous and difficult.

The concept of autonomous drones, robots, and self-driving vehicles (SDVs) is now becoming commonplace. We’re now on the cusp of eliminating the cab and allowing the farmer to sit in front of a laptop, tablet, or even use their smartphone to direct the combine. That said, we’re about to see a drastic uptick in the amount of capital expenditure and implementation of drone, robotics, and SDV technology across the logistics industry. Labor shortages, Amazon, luxury convenience, 5G technology, the last mile, and consumer demand for the free delivery of goods within hours are regular discussions across the logistics industry. Underpinning many of these discussions is also the broader Industry 4.0 one: the automation of services by robots and AI technologies. In 2017, INFORM put together a three-part blog series titled Drones, Robots and Selfdriving vehicles: Reshaping the Logistics Industry which was also published in the 2018 book Exploring New Frontiers: Reshaping the Postal Industry. Now, three years on from the first release of that series, we’ve taken the time to dust it off and look at what’s changed around these automated technologies and what the near future might

look like for them. Again, for the sake of this discussion, we’ll continue to focus on three categories of automated vehicles: unmanned aerial vehicles (UAVs, aka drones), land-based small vehicles (robots), and automotive self-driving vehicles (SDV). Drones According to a study from PwC from 2017, the market for drone technology was estimated to reach $127 billion in 2020, not including military uses. Since this report was completed, we’ve seen the FAA continue to design regulations for Unmanned Traffic Management (UTM) and companies incorporate drones into their operations for their usability in numerous settings as a subsequent result of improvement in cameras, battery life, sensors, communication, and other component technologies. In a recent CBInsights report, the use of drones wasn’t limited to logistics providers either, with the list featuring names like Alphabet (Google), IBM, Apple, GE, and the BBC joining the likes of Amazon, FedEx, DHL, and UPS, just to name a few. Among the logistics operators, Amazon is still the loudest drone advocate

with tangible commercial services in place. In 2019, we saw Amazon deliver more packages via its Prime Air service. If you’re not familiar with Prime Air, it touts delivery in 30 minutes or less for packages weighing less than 2.25 kg (5 lbs.) and within a 24 km (15 mi) range of the Washington, D.C. area. In the same year, DHL announced its first fullyautomated drone delivery service over the skies of Guangzhou, China. Finally, we covered that UPS trucks will become a mobile hub/landing pad for a counterpart drone to deliver lightweight packages in conjunction with routes to increase efficiency and capacity in our first version of this article two years ago. Since then, UPS has been the first logistics supplier to be fully approved for commercial drone delivery by the FAA in America. Considering the added complication of flight, continuous innovations, developing FAA regulations, and implementation costs, drone technology is still progressing slowly. However, as companies continue to tighten their grip on the value and logistics of effectively utilizing drones, we will begin to see more drones moving forward, and with such a broad spectrum of usage applications avail-able, we’d argue that a future where drones


are performing small- to medium-scale services, like package or pizza deliveries among others, is no longer a question of “if,” but rather a question of “when.” Robots Robots, on the other hand, seem to have proven themselves on multiple fronts over the past few years demonstrating far greater possibilities than their aerial counterparts. Since land-based robots are not constrained by the limitations of sustained flight and associated payload weight restrictions, their application areas are considerably broader. Labor shortages being experienced across the globe are also providing a major opportunity for implementing robotics of all shapes and sizes to support the manufacturing, order fulfillment, and warehouse operations of logistics providers around the globe – we’ll skip the Industry 4.0 (aka “Fourth Industrial Revolution”) lesson for now. Since 2017, we’ve seen the development of 5G (high-speed, high-reliability, mobile networks) and SLAM technology along with increasingly sophisticated WMS and OMS (Warehouse and Order Management) systems. Robots can now co-op with human counterparts to support and optimize functions for order fulfillment and inventory management across a company’s manufacturing and distribution portfolio. The data bridge between consumer and retailer has now become seamless, which allows for orders coming in to be optimized and to reroute a robot’s path through the warehouse in real-time. In some cases, robots completely take over functions that humans used to complete or alert the human to the next pick (e.g., picking SKUs for order fulfillment, replenishment, etc.) which has allowed for more efficient use of warehouse space. A few years ago, there was a lot of discussion around the elimination of jobs similar to the plight of John Henry (i.e., “man vs. machine”). However,

Robots can now co-op with human counterparts to support and optimize functions for order fulfillment and inventory management

ted to be sold in total from 2019 to 2021. In addition to companies buying robots, we’re also seeing companies sprouting up, offering “Robots as a Service” (RaaS). RaaS is where a company sets up and rents out robots to warehouse operators and manufacturers much like you would lease out commercial vehicles or equipment. This especially comes in handy for providers that want to adopt robotics when short-handed on labor during seasonal peaks in activity or simply can’t justify the initial investment of purchasing their own. Self-driving Vehicles (SDVs)

robots seem to have become more of a sidekick filling in the gaps that were previously subject to human error. According to a study completed by the UC Berkeley Labor Center, displacement of workers will vary across companies experimenting with robotics, but will not drastically impact the industry in the short- or medium-term. In fact, the robots may serve to intensify workers’ day-to-day activity since the robots will be optimizing workers’ activities and routes in the warehouse to fulfill orders more efficiently, resulting in increased worker activity. The study went on to say, “With continued growth in demand, aggregate employment levels in the warehousing industry will likely continue to rise over the next five to 10 years. That said, job growth may be tempered by the increased use of labor-saving technologies in e-commerce warehouses in particular, such as autonomous mobile robots, autobaggers and autoboxers, and sensors or RFID tags applied to goods. Honeywell, for example, has developed robotic unloading machines that reduce the offloading time and work in parallel to the role of workers.” The logistics robot market is expected to grow with an 18% compound annual growth rate (CAGR), which equates to roughly 485,000 logistics robots expec-

Since we last touched on this sector in 2017, there has been an explosion in demand for self-driving cars by consumers, parcel delivery companies, and shippers globally to take advantage of this technology. Others believe it will save lives in addition to allowing for deliveries to run more efficiently. As such, we’ve taken the step to divide this category into three sub-sections: SDVs (what you’ll find on the road), Automated Yard/ Shuttle Trucks (self-driving industrial vehicles you’ll find in your depot’s yard), and Autonomous Ships (found, well…on the water). SDVs Safety is a big driver in the move towards any automated driving technology, SDVs included. In the U.S., the NHTSA estimate that in 2017 there were approximately 91,000 reported crashes involving tired drivers which resulted in as many as 50,000 injuries and nearly 800 deaths. They go on to say, “…there is broad agreement across the traffic safety, sleep science, and public health communities that this is an underestimate of the impact of drowsy driving.” Skilled labor shortages are another significant driver motivating a shift to SDVs. The median age of truck drivers


continues to rise around the world. In the U.S. it is now 45 and in Europe, it is 44. Age is also only half of the problem. Attracting and training drivers is also a significant challenge. According to a report from the American Trucking Association (ATA), there are more drivers than ever, but in the U.S., they continue to fall short on demand. In Europe, there is a current shortage of drivers by 20% of demand, and this is only estimated to double in the coming years. When you consider transportation costs equate to roughly 45-55% of all costs across the supply chain (Labor Availability Challenges in Logistics Real Estate – Liz Dunn and John Vitou), it is impossible to ignore the difficulty facing the logistics industry. There is a need for strong solutions. In the U.S., this may come in the form of lowering federal age restrictions to allow adults ages 18-20 to drive Class 8 motor vehicles. But, there is also a huge incentive for companies like Man, Volvo, Sacnia, Daimler, and other manufacturers to come up with more technical solutions like SDVs and Platooning. And, it isn’t just suppliers bringing technology to the market. In mid-2019, the U.S. Postal Service started testing autonomous vehicles on runs between Phoenix and Dallas noting that the move comes as a way to constrain operating costs as much as it is to improve safety and efficiency.

Automated Guided Vehicles, or AGVs, were introduced to the maritime logistics sector in 1993 at ECT’s Delta/Sealand Terminal in Rotterdam, The Netherlands. Since then, many maritime terminals around the world have embraced the technology in their broader transitions towards becoming mostly or fully-automated, which in turn has seen the technology improve and the costs reduce. In the past three years, we’ve seen significant options come into the market. Just last year, two major players announced automated yard vehicles, and there are multiple Europe-based projects actively testing the feasibility of these solutions. ZF released its ZF Innovation Truck, capable of handling swap bodies (which complemented its existing Automated Terminal Yard Tracktor), and we were also were introduced to the futuristic, fully-electric Volvo Vera for more typical trailer handling. Both are purpose-designed for use in manufacturing and distribution yards, and importantly, both come in at a price that will drive a strong ROI for logistics yard operations. Dachser and the Fraunhofer Institut are testing solutions together in Dortmund, Germany, and the Austrian Post and the Technical University Graz are working together to test autonomous technologies to name just a couple more. Autonomous Ships

Automated Yard/Shuttle Trucks Most manufacturers agree that to bring a Level 4 (fully autonomous vehicle) to public roads is a medium-term goal, and it requires continuous refinement and further investments in the technology. In general, the controlled environments of internal yards are the perfect location for autonomous trucks to be introduced into logistics operations. In fact, internal logistical operations were one of the first implementations of automated vehicles.

Rolls Royce Marine and Japanese shipping company Nippon Yusen have claimed they will have un-crewed cargo ships, or autonomous ships, in operation as early as 2020. Most shipping accidents occur as a result of human error. The Maritime Journal said in 2017, “It is estimated that between 75% to 96% of marine accidents can be attributed to human error.” As we see with vehicles on land, taking human error out of the equation could improve safety across the board.

The costs to remove wrecks, impacts on the environment, and liabilities to the crew and other ships are huge, and there are major incentives to streamlining a solution that mitigates the risk that comes with maritime transportation of goods. However, instead of allowing ships to become completely autonomous, there is a consensus that they will be more of a hybrid between remote navigation/ control as needed and autonomy reliant upon sensor arrays positioned throughout each vessel. Each shipper will have, essentially, a traffic control group that will take over a ship when necessary. The crew goes from being on the ship to sitting behind a desktop in an office building somewhere. Reshaping our Industry As we move squarely into the Fourth Industrial Revolution, it seems the promises of automation are everywhere. It is important for logistics operators to consider experience as they move into the field of automation. INFORM has been working with automated ports for over 20 years now, and we’ve learned a lot along the way. We’ve also been part of an industry that, to this day, is still learning how to implement automation well. All of this knowledge and experience is built into the core of our software solution, Syncrotess. For distribution centers and post and parcel operators, Syncrotess performs as a powerful Yard Management System that is ready to support fully-automated and semi-automated yards right out of the box. And, with a powerful AI engine under the hood, our YMS will complement your investment in automated hardware with an automated software solution truly transforming your yardhandling operations and delivering a powerful ROI.


ROI FOR YOUR YMS Is your YMS a “Nice-to-have“ Tool or an Investment? By Dr. Eva Savelsberg and Matthew Wittemeier

Your Yard Management System (YMS) isn’t exciting. In fact, in the scheme of post and parcel operations, it isn’t even a big deal; harsh, but true. You’ve heard the same feature arguments over and over disguised under the vague vale of ‘digitizing your yard’. True, some of these features are ‘niceto-haves’ but do they add value to your business or simply cost you money to have? While we’ll look at a full range of features later in this paper, let’s start with a quick discussion on ROI. While YMSs are not exciting, a purpose-built post and parcel YMS with optimization built into it can deliver a 100% ROI in as little as 7 months and then drive significant value over its lifespan. How does this compare to a standard YMS? The way the numbers stack up might surprise you.


“Nice-to-have“ or Investment

A typical YMS might cost you ‘N’, and it might deliver ‘nice-to-have’ features and a little measurable value along the way. But at the end of the year, it still costs you nearly your full purchase cost of ‘N’. Annual ongoing costs will push this out to just over 2x ‘N’ over a ten-year lifespan (see figure 1). There is little added value in a nice-to-have tool. In comparison, a YMS that delivers measurable value through optimization might cost you 2x ‘N’ but can deliver as much 4x ‘N’ a year in value. After ongoing annual costs are factored in, your post and parcel business stands to earn 36x ‘N’ over a ten-year product lifespan (see figure 1). When you compare the two, the standard YMS will end up costing you more than your initial investment in a purpose-built YMS and you stand to earn minimal value with that tool. In today’s ultra-competitive post and parcel business environment – ‘nice-to-haves’ don’t contribute to the bottom line and things that don’t contribute to the bottom line need to go. You’ll find YMS software comes in all shapes and sizes – but every system is different and, not all systems have been developed equally. The trick is finding the solution that: best suits your business’s needs today, can grow with your organization for tomorrow, and will deliver the strongest ROI over its lifespan. Below are our top 11 features post and parcel operators should look for in a Yard Management System. We’ve split the list into three sections – “Nice-tohave“ features, “Added Value“ features, and features ‘Specialized’ for parcel operations.

“Nice-to-have“ Features

1. Full transparency of the on-site load units and yard resources.

With a YMS, you know where every load unit is at all times because the system logs every incoming and outgoing asset as well as every move they’ll make when they are in your yard. Beyond the time you’ll save every day, you also improve the overall safety of your yard by reducing how many people you have in the yard at any given time which in turn also improves operational efficiency through a decrease in operational disturbances. Further, a good YMS builds on the full transparency of each on-site asset and provides you with transparency of your yard resources too. Real-time knowledge of the position of each yard truck helps to improve routing and transport decisions as well as yard safety. 2. Faster gate in and out handling. Through standard, digitized gate processes, the speed of checking trucks in and out of your yard will be dramatically improved. A good YMS will support fully automatic gates as well as manual processing. It should also facilitate communications inside the software when the situation doesn’t match the plan (i.e., the load unit that arrived isn’t what was planned for) so that your site dispatchers can manage the exception inside of their larger plan and communicate that directly to gate staff without the need to switch to email or make a phone call. 3. A calm work environment. Central to any good YMS is an automated communications platform that allows dispatchers to communicate via the software directly with gate staff and drivers. This removes the need for telephone and radio chatter dramatically improving the work environment for all staff. Automated communications also decreases delays and improves on-time job performance, and improves the accuracy of communications.


4. Data standardization. Good YMS are designed and configured to support standardized processes. This allows the system to predict good solutions and adhere to constraints within your yard design. They also enable the system to be run in an automatic or partially automatic state when desired. The side effect is that standard processes means standard data. Importantly, this is the crucial first step in setting up advanced analytics for evaluating business KPIs as well as implementing Machine Learning for longer-term process improvement initiatives.

“Added Value” Features 5. Yard vehicle (shunter) optimization. A high performance YMS comes packed with optimization features, including the ability to optimize your yard vehicle operations. It doesn’t matter if you have two yard trucks or twenty, optimization delivers a range of benefits from reduced travel (fuel savings) through to decreased maintenance and increased asset lifespan (strong asset ROI). When applied to larger yards, with many yard trucks, a good vehicle optimizer can reduce the total number of trucks required to deliver the same service quality level by as much as 34%. 6. Dock door optimization. Your yard is about feeding your dock doors – and a good YMS will optimize how you manage your dock doors to improve your facilities overall performance. Further, it should be able to automate routine dock door decisions freeing up dispatchers to focus on services that add value to the organization vs. control cost.

A complete knowledge of dock door parameters and status ensures that the YMS adheres to fixed constraints as well as variable ones too. Displays inside the facility that interface to the YMS provide facility staff with real-time information about the load unit at the dock and handheld devices allow them to feed critical information about the unloading/ loading process back into the YMS for improved decision-making. 7. On-time performance management. Advanced YMS systems can receive data from your TMS about load unit contents allowing the YMS to make decisions that improve your facilities on-time performance. As in-bound load units arrive, containers with time critical content are assigned for immediate unloading while standard load units can be stored until a dock becomes available. 8. Priority load performance management. With data from your TMS, a good YMS will use load unit content data to improve how it handles in-bound load units. Containers containing priority loads will be prioritized over standard priority content. Outbound load units containing priority loads can also be prioritized for discharge from the facility. 9. AI and Machine Learning integration. AI improves the way users interact with the system (through features like Chatbot) and also enable the system to run autonomously making routine decisions so dispatchers are able to manage more complex exception handling and improve customer service outcomes. Machine Learning enables the system to

continuously learn from itself and feed those learning outcomes directly back into the system and/or to human experts for review improving the systems overall efficiency over time.

“Specialized” Features for Post and Parcel Operators When evaluating YMS software solutions, the above features can be included in a YMS designed for any industry and any yard environment, but there are also industry specific features which make a select few vendors more suitable for delivering YMS for post and parcel operators. 10. Tight integration with the sortation system. Through a tight integration with your distribution hubs sortation system, a YMS will be able to improve your sortation systems productivity through improved yard management decisions. At cross docking sortation centers, a YMS with tight integration to your sorter can balance sorting workloads to sub 85%, improve sorter efficiency, minimize shift completion times, and improve safety outcomes in the yard. These outcomes improve the efficiency and handling capacity of your centers most expensive investment, its sorter, through improved yard management decisions. 11. Post and Parcel KPI based performance measures. Real-time KPIs can transform a business’s productivity but what’s the point of measuring speed if you’re driving in the wrong direction? The best YMS for post and parcel operators provide realtime KPI based performance measures designed for post and parcel operators.




For more than twenty years, Swiss Post has optimized the transport and transshipment at their Härkingen, Daillens, and Frauenfeld parcel centers, employing INFORM’s software solution SyncroTESS. The yard management system (YMS) monitors, controls, and optimizes the yard management and handling of containers at each of the three Swiss Post parcel centers. In addition, the YMS handles all of the road and rail trunking between the three sites. As a mainstay application within Swiss Post’s IT application stack, it has stood the test of time and continues to deliver decisive support to the organizations yard dispatchers. And, year-on-year, it continues to deliver a return on their initial investment and contribute to the organizations strong bottom line.

Read the full article online. Infrm.co/YardMag2020




INFORM’s Senior Vice-President Dr. Eva Savelsberg, spoke to Post & Parcel about the diversity of optimisation, how the sector can leverage AI and the importance of thinking ahead.

P&P: Can you describe your role at INFORM? ES: I joined the company 13 years ago. At that time, INFORM had around 250 employees; so still a medium-sized software company. Today, we have over 800 employees and I’m heading the Logistics business unit. Logistics is actually quite a broad area – as it is one of INFORM’s older business units. Today the Logistics business unit provides AI and optimisation software to maritime ports, inland – or intermodal – terminals, distribution centres, and of course post and parcel operators. But as well, we also optimise the logistics of building material producers, like cement, and we offer optimisation for retailers. So it is quite broad, but at the core of the business, is always logistics, it’s always how can we work more efficiently, how can we support the processes of our customers by implementing artificial intelligence and operations research (OR) methodologies to improve their bottom lines.

P&P: It’s interesting to hear that INFORM is broader than just the post and parcel sector? ES: Oh it definitely is. We are really strong in areas starting from aviation, supporting the whole ground handling at airports and going all the way to, for example, the banking and insurance in-

dustries, where we help to detect fraud in their transactions. We try to walk around with our eyes open to new opportunities and so over the 50 years the company has existed for, we’ve met potential customers in very diverse business areas and contexts. When you consider that optimisation is so broad, that you can use AI and OR algorithms in such diverse industries and deliver a really big impact. It is easy to see how INFORM has grown over the decades to encompass so many application areas of our core optimisation competency.

P&P: How did you get into the field of logistics and what attracted you to INFORM? ES: My background is mechanical engineering. That’s also where I did my PhD thesis. But at that time, I was already working quite closely with industry and always in the, let’s say triangle of mechanical engineering, software, and logistics. So I’ve been involved in this field for a couple of decades. I thought that was pretty lucky because on the one hand it’s such a broad field, but on the other hand it also has such a vast impact on our economy and on our personal life. So for me, it was always very fascinating to be part of shaping the future of logistics and software for logistics. That’s also how I met INFORM and how I came to work with the company.

P&P: Does it help having a foot in both academia and industry? I understand that you still lecture at the university. ES: Indeed, I still lecture at the university and I enjoy that quite a bit. But my day-to-day work is predominately in industry – my work for INFORM. Once a year, I have the pleasure to run a course at the RWTH University here in Aachen. I think the nice thing about it is that you always need to reflect your work in the context of the world. I try to keep it interesting for the students and for myself. I try to think about it over the course of the year and to question, “Okay, wow, that’s interesting. What are interesting technical developments or interesting political development and how might these be interesting to pick up and integrate into course.” The course I run – Innovation in Freight Transportation – is quite broad. But, since I have a bit of freedom, I can always shape it a bit. So this year I introduced the UN Sustainable Development Goals, or SDGs for short to also use those as a framework to reflect in. Students, and myself need to continually question, “okay, what products do we think will have an impact to the future and how are those topics interrelated?” So I guess it helps me to see things from different angles and this helps me for sure in my day-to-day work at INFORM.


AI can certainly support the humans through automation to relieve them from their daily tasks so that they can concentrate on being creative, on troubleshooting, on stuff which suddenly comes up and AI is not trained to handle Dr. Eva Savelsberg

P&P: Why do you think people are fascinated with artificial intelligence? ES: I think it is a fascinating topic for people, perhaps due to the myths around artificial intelligence, because I think people just assume upfront that it’s really nerdy stuff and that it’s really difficult to understand. But actually, if you invest a little time, you can at least get a quite good grip of how it works in principle, of how you can utilise it and of what might be the impact. I think people are also fascinated because it will have a decisive impact on their life. Actually it already has, I mean we have smartphones and those are packed with artificial intelligence and we rarely consider it. But to think about what’s ahead of us, “AI is a resource”. That aside, the questions: How can we use it? How can it be a tool for us? AI is very usable and you can implement it in many and diverse applications. Data and AI are a bit like a screws and the screwdriver, respectively. Screws existed from the early first century, but it wasn’t until the invention of the first screwdrivers in the 1700’s and subsequent standardisation in the 1900’s that they came into mainstream use. Data has existed since the advent of the computer and the quantity of data has exploded exponentially since the internet enabled easy sharing of data. It’s estimated that 90% of all the data in the world was generated in the last two years alone. The human mind can, simply put, no longer comprehend all the data that’s available and that is where AI, as a tool comes in.

P&P: So, how exactly can AI be a tool for the post and parcel sector? ES: From my perspective during the last years, the post and parcel sector focused a lot on how to cope with declining post volumes and also how to use AI to gather more knowledge about their customers; which was good. So, they were able to respond more to customers’ needs. But I think now, perhaps, it’s a point in time to look at how can they use AI to make their processes more efficient and also how can they leverage AI. At least in Germany, that’s really a pressing topic. In the coming years we will have fewer and fewer skilled workers, so in general, Germany needs to embrace more automation in the post and parcel sector to make up for the labour shortfalls that are expected. But, we still need to consider, that so far, robotics and AI still have their limitations and the human brain still has its advantages. But, AI can certainly support the humans through automation to relieve them from their daily tasks so that they can concentrate on being creative, on troubleshooting, on stuff which suddenly comes up and AI is not trained to handle. P&P: Is there much resistance from the post and parcel sector to employing AI? ES: For me it isn’t a question of resistance or no resistance to AI. It is a question of where post and parcel operators are focusing their attention, and subsequently their investments, in AI currently.

As noted, there is a good deal of knowledge on the side of AI organising customer insights and knowledge. I think that operators are not so aware yet how much AI can bring to the table for the production itself, for the logistic processes. We’ve just released our first whitepaper on the impact of an AI empowered yard management system on a parcel center’s sortation system – importantly, without changing a thing in the sorter. A system like our YMS that is complimented with AI and OR technology, can deliver strong performance increases and cost reductions beyond just the yard.

P&P: Do you think that people over- or underestimate AI in its current capacity? ES: Yeah. There is the saying and I think perhaps everybody heard it already, but still it’s quite right to say… people tend to overestimate what will happen in the short-term and underestimate what will happen in the long-term and I think that this definitely applies to AI. I’d say that in the next three years we won’t see much of the hype around AI change the parcel industry greatly. But, and this is an important but… over the next 20 or 30 years I’d venture a guess that AI will reshape logistics significantly. So, it might be a more steady change, but still, 20 years is actually quite a fast change process for an industry as old as ours. I’d say that in 30 years, our life will look quite differently.

P&P: If you could give post and parcel operators one bit of advice regarding AI, or even emerging technology in general, what would it be? ES: To think ahead with two perspectives. Firstly, the short- and medium-term perspective of how can we make use of the technology today and in the near future. Secondly, also to have a kind of


think tank in their company, or even external support, about what might happen 10 to 20 years from now. To think about questions like: How can we prepare and steer the company for that future? How can we prepare the processes to support and work within our industry? And as well, how can we prepare as a society?

P&P: That’s an interesting perspective that I understand you embrace at INFORM too. We’ve had the opportunity to read a book you co-authored titled 2038: A Smart Port Story where you do just that. ES: We wrote 2038 to take this approach but also to open up our own minds and give ourselves the freedom to think ahead. I think the format of a story is an interesting approach because it not only helps people to use their imagination and think visually, but it makes it easier to think about what might come up as you aren’t limited by the preconceptions of the present day. When people are thinking about the future, the second perspective to say, is it important to be right or is it about the exercise, about the thought process? It’s definitely about the thought process. It is about thinking through different scenarios and also, I think, by just discussing these scenarios and getting these pictures of possible futures, they’ve already gained some flexibility in their brains and they get some competency in terms of adapting to an unknowable future. I think it’s a bit like elite athletes. An athlete before a competition, they can withdraw in themselves and think about the competition and kind of play it through in their minds. Mental training is quite substantial in terms of performance and I think to have scenarios about the future is a mental training for a company or for the board of a company and that will help them even if the future doesn’t

turn out as they planned for. But they will be flexible enough and will be capable enough to react. P&P: What would you like to see INFORM achieve for the sector in 2020? ES: I would like to have the opportunity to sit down with different leading companies in the sector and to have a look at how we can achieve some of the benefits from the implementation of AI in the production processes. In fact, I think that would be helpful for the whole sector.

When people are thinking about the future, the second perspective to say, is it important to be right or is it about the exercise, about the thought process? Dr. Eva Savelsberg

P&P: What would surprise people about your job? ES: I think the job is probably comparable to most jobs. A quite big stretch between thinking and strategy and thinking about the future, and on the other hand still taking care of details. And I think it’s a lot about human beings and the emotions of human beings. So that can be the emotions in a project team where you have to take care that everybody is in a position to work happily together. So although, let’s say, I have a quite technical background, actually daily business is influenced greatly by taking care of the people around me.

P&P: What are you currently working on and what do you find exciting about it? ES: If I step back to how we started the interview, talking about the broader field of logistics I think it will contextualise some of the insights I’ve shared today. We are working on some large scale, innovative logistics projects. It’s exciting to see how the outcomes from these projects will shape the future of the broader supply chain and this will no-doubt have flow-on effects into the post and parcel vertical. Also, it’s exciting because at INFORM we’re focusing a lot on the UN SDGs discussed earlier and how these play out for us as a company down into how they impact our products and how our products impact them. Outside of INFORM, in the European market, companies are also picking them up more than I’d originally thought. And so that opens doors for new developments to think about aspects also in terms of, “okay, which kind of footprint will that give?” And also if we have a new service provider, “how do we want to work together?” This is also driving a shift that leads to a place where the answer to every business decision is not simply profit. And that perhaps changes a bit in the market and I think in the perspective, and that also opens up to say, “Okay, how do I want to work in a customer-vendor relationship and how do we want to produce as a company? Do we just want to look at profit or do we also want to look at what are the consequences of our actions – good and bad.”

P&P: Thank you very much.




Understanding the Human-AI Partnership By Dr. Eva Savelsberg


“You are my creator, These chilling words from Mary Shelley’s novel but I am your master.” Frankenstein were first published on JanuAry 1, 1818, during the First Industrial Revolution; a period of great social and technological change. Considered by many to be the first work of science fiction, the story influenced not only literature, drama, and film but also the public‘s perception of science.

2018 marks Frankenstein’s 200th anniversary. And at the dawn of the Fourth Industrial Revolution, the myth of creature turning on creator seems more relevant than ever before. Having escaped the laboratories of many tech companies, Artificial Intelligence (AI) is poised to change our society for good. While human-level AI is not yet looming around the corner, we constantly carry some form of AI in our pockets today. The irony is that Siri, Alexa, and Cortana, while comparable to Frankenstein in so many ways, aren’t perceived to be frightening characters. Rather, these AI enhanced assistants have become an ordinary, if not integral part of our lives and workplaces. This article will take you on a journey to the past, present, and future of AI. To unpack this story, we need to have a few stops along the way, feel free to skip ahead if you know the background. Firstly, we need a quick reference point of what AI is. Then, it is worth identifying why INFORM is qualified to speak on the subject. From here we will explore how AI is being applied in the post and parcel market today. Finally, we‘ll discuss what the role of humans is likely to be in the future and whether any of us will have jobs.

The Simple Truth Behind AI Artificial Intelligence is an area of computer science that‘s concerned with building systems that demonstrate intelligent behavior. Most people find it difficult to agree on a precise definition of intelligence, and so people‘s view of what Artificial Intelligence means also tends to diverge. For most people, when they hear the term Artificial Intelligence, or AI, they think of a General AI, or human-level AI, that can mimic all aspects of human intelligence. The simple truth, however, is that today, AI is far away from this. Instead, AI vendors have succeeded in building niche,

Narrow AI systems that are now making their way into our industry at a rapid pace as part of the Fourth Industrial Revolution

or so-called Narrow AI systems that know how to do reasonably specific things very well (for instance, play chess, translate between languages, understand natural language, or drive autonomous vehicles). It is these Narrow AI systems that are now making their way into our industry at a rapid pace as part of the Fourth Industrial Revolution. In contrast with General AI’s goal of mimicking human intelligence, Machine Learning platforms (ML) use algorithms to iteratively learn from and adapt to data, enabling computers to find hidden insights without being instructed where to look. A beginner’s example for this can be found in your email inbox: spam filters. Simple rule-based filters are not very effective against spam, since spammers can quickly update their messages to work around them. Instead, ML enhanced spam filters continuously learn from a variety of signals and tailor themselves to the email needs of the individual user. The Hidden Secret Operations Research (OR), also referred to as the science of better, uses analytical methods (mathematical optimization, heuristic methods, etc.) to analyze


AI & ML in the Post and Parcel

Operating Data Logging

Machine Learning (ML)

BI Cockpit

INFORM AI powered YMS Expert Discussion

and consider vast amounts of data to optimize the real-time control of processes. Depending on your view, OR is either a means to an AI outcome or OR and AI are complimentary disciplines. In the first view, Operations Research is a means to an Artificial Intelligence outcome. When you think of AI as the area of computer science that is concerned with building systems that demonstrate intelligent behavior, one could say that OR is part of AI. “From a classical research perspective, this is inaccurate because OR and AI are two separate disciplines that have independently developed intelligence-based computing techniques. However, if you take the broad definition of AI, with building systems that demonstrate intelligent behavior, Operations Research could be classified as a part of Artificial Intelligence,” said Dr. Ulrich Dorndorf, Chief Technology Officer at INFORM. Alternatively, AI is a technique that makes better predictions about the data that is fed into OR optimization algorithms. Either way you choose to view the relationship between OR and AI, INFORM has been working with Artificial Intelligence

for over two decades, with commercially available products in use since the early 2000‘s. “More than 20 years ago, we started developing knowledge-based AI systems that were based on the concept of using fuzzy logic and fuzzy reasoning for representing human knowledge. Over the years, we‘ve added Machine Learning as a second area in our AI activities, and the two are now working in parallel together,“ added Dr. Ulrich Dorndorf. The Modern Prometheus The subtitle of Mary Shelley’s Frankenstein novel is “The Modern Prometheus”. In Greek mythology, Prometheus was not only a symbol of limitless ability but also the god of forethought. Forethought is a much-needed skill among yard dispatchers and planners in the post and parcel industry. OR based yard management (YMS) tools allow operators to make incredibly complex, time-critical decisions with ease. Powered by algorithms, these tools analyze a virtually endless number of decisions in real-time and identify those that are ideal for


minimizing costs and maximizing service quality – based on the business criteria defined. What‘s more, the decisions made take into account a larger range of variables than the human mind can, resulting in better overall decision quality. To further enhance the decision-making quality of OR based tools, a Machine Learning (ML) platform can be connected to the system. Looking at huge amounts of data from the past, ML can analyze the behavior of individual hauliers, shunter drivers, dock door teams, etc. and identify patterns that are not immediately apparent to human operators. Example: A time slot management system provides hauliers the ability to book allocated inbound arrival time slots. However, delayed deliveries are commonplace and actual truck arrival times may vary over the course of a business day (off-peak vs high-peak). They may vary on different weekdays and weather conditions. They may even vary for different hauliers and/or drivers. Besides data from the YMS and Optimization Modules, the ML can also be connected to various other internal and external data sources (e.g. TMS, WMS,


Sortation System, weather apps, etc.) to further enrich the data basis. The ML can be operated 24/7, on a periodic, or on an event triggered basis.

sights for humans. These insights form the basis for expert’s discussions to avoid spurious correlations. Chatbot – Is it alive?

Teaching the Monster One of Frankenstein’s gravest errors was to neglect his creation. He fled from its presence, giving up the opportunity to supervise, nurture, and educate his invention. Supervised learning is a term that is often used in conjunction with Machine Learning. Similar to the way a child learns from a teacher, supervised Machine Learning finds patterns where we have a dataset of “right answers” to learn from. For example, an algorithm will learn to identify dogs after being trained on a dataset of images that are properly labeled as “dog” and some additional identifying characteristics. The examples described above, however, use a technique called unsupervised learning. Unlike supervised learning, there are no correct answers and there is no teacher. Algorithms are left to their own devices to discover and present interesting patterns in the data. This purely algorithmic extraction of rules from data is prone to creating spurious correlations. As it turns out, US crude oil imports from Norway track nicely with the number of drivers killed in collision with railway trains. This is one of the many spurious correlations that Tyler Vigen has published on his website (ww.tylervigen.com). Just because the movements of two variables track each other closely over time, doesn‘t mean that one causes the other. With this in mind, any insights from a Machine Learning platform can be fed automatically into the OR based software. And there are many use cases where this set-up works absolutely fine. Alternatively, when the ML outcome is used to fine-tune the models and rules that are encoded in the software, findings and correlations are better presented in easy-to-digest dashboards to create in-

In the riveting laboratory scene when the monster is brought to life, Dr. Frankenstein shouts, “Look! It‘s alive.” Today, computer programmers can have similar moments when they develop chatbots. Chatbots are one of the most common AI-based applications. They are designed to sound and type like human beings and continuously learn and develop through AI and ML.

How we manage this human transition is going to define our industry To make latest technologies and applications available for the post and parcel industry, INFORM recently released a chatbot add-on for their YMS solution. It receives both voice and text-based queries from a broad range of input sources, recognizes the request, searches for the answer, and sends the answer back as a text response in real-time. The chatbot quickly allows checking the status of KPIs, containers, or equipment without calling anyone. Management, with no previous training in the system, can ask the software directly enabling them to quickly access KPI data on the fly in a manner that is convenient to them. AI – Frankenstein Reloaded? In the riveting laboratory scene when the monster is brought to life, Dr. Frankenstein shouts, “Look! It’s alive.” Two hundred years later, we find ourselves at

a prologue to a new Frankenstein story. AI can solve many problems in the post and parcel sector, but it will not replicate the human brain anytime soon. The aim of AI development should not be to make “it” better than us, but rather to make “it” beneficial to us. Technology moves ahead, but so does the human mind and our attitude towards technology. A senior operations manager from the baby boomer generation might have a different opinion on the usefulness of chatbots compared to a millennial management trainee. And a seasoned shunter driver will be more hesitant to accept decisions and work orders from an AI system than a digital native who is about to start a career in our business. By 2025, millennials will make up 75 percent of the global workforce. They have grown up with speedy communication and high-tech is woven into all aspects and areas of their life. And what’s more, the generation born after 2010 – the “AI natives” – will only know a world with artificial technology. How we manage this human transition is going to define our industry. At this point, there are more questions than answers: How do we best utilize highly skilled staff who are in traditional roles? How do you prepare these staff for the future and how their roles will change? If we retrain them, who bares the financial and social costs of retraining them? How do we attract a young, millennial aged workforce that have the new skills we’ll need in the future? To position ourselves for the future, it is the role of all stakeholders in the post and parcel industry to take a degree of responsibility. The question isn’t whether AI is coming or not, but rather, whether as an industry, we will be well prepared or caught off guard when we realize that AI is here – or, as Mary Shelley wrote in Frankenstein, “Nothing is so painful to the human mind as a great and sudden change.”


MILLENNIALS AND STAMPS Really, you lick it? By Dr. Eva Savelsberg and Matthew Wittemeier


While millennials will not only penetrate the logistics workforce, they will also be the ones who drive demand on the consumer side. “On-demand” and “digital” is their way of life. Anything else will not be accepted which begs the question, what is the future of the humble stamp?


The stamp; we all know it as that little square that you used to lick to get your letter sent. Now, it is only a sticker or in many cases not even a physical thing but rather a “something” printed on an envelope. Is the day coming when what we all know to have been a stamp will be a thing only existing in our memory and something unknown to millennials and their children? We’ve all seen the YouTube videos of kids reacting to antiquated technology – everything from cassette players to rotary phones – and there is a brilliantly, funny consistency to these videos. In the case of the rotary phone, it takes most kids anywhere from 30 to 60 seconds to understand what it might be. The idea that it is a phone is foreign to them. As they begin to physically explore it, you can see them trying to “press” the numbers. The idea that it is a dial that rotates and returns never crosses their mind. To an adult, it is certainly funny to watch, but, humor aside, it is telling of a generational gap where things that are commonplace to us are

disappearing from the lives of a younger generation. With the age of the millennial fast approaching, we should be asking ourselves if things like the humble stamp will survive beyond our generation. On to Avocado Toast By 2025, millennials will make up 75% of the global workforce, i.e., the generation of “pen and paper” operators are a dying species. Millennials have grown up with instantaneous digital communication, and “high-tech” is woven into all aspects and areas of their life. It’s been reported recently that Canadian millennials across the board don’t know how to use a stamp or properly address an envelope. So, while millennials will not only penetrate the logistics workforce, they will also be the ones who drive demand on the consumer side. “On-demand” and “digital” is their way of life. Anything else will not be accepted which begs the question, what is the future of the humble stamp?

Generation “avocado toast” (what many have nicknamed the millennials), is “stirring the pot,” “shaking things up,” “challenging the norm” or whatever you want to call it. Overall, they are exceptional at not accepting the status quo – “just because.” Things that we consider normal, they often disregard. To the frustration of their more senior colleagues, this is often seen as a negative – but, is it always? If it Ain’t Broke, Don’t Fix it When approaching the challenge millennials can bring, try and remember that change is often resisted as is summed up in the age-old saying, “If it ain’t broke, don’t fix it.” But change is required to remain relevant, and now more than ever. In our previous piece which was centered around technology and humans, we noted that the pace of technological change will never be slower than it is today. If you recall from that piece, it was estimated that only 54% of major change projects are successful. Those that fail are plagued by higher than expected costs and lowered employee morale. Studies also show that when employees see major projects fail, or fail to deliver major elements, cynicism sets in, which, in turn, further undermines adoption, utilization, and worse – company culture. So, while organizations see a need for new technology, implementation often goes astray. People Have to Drive all Change Before pinpointing all software and hardware solutions on your roadmap to transformation, decision-makers should plan to have two internal stopovers and address the human and organizational aspects of the change process first. The sequence of the HOT approach (Human, Organization, Technology) ensures “transformation readiness” before new technology moves in. Many leaders un-


derestimate the consequences of inadequate readiness and, at the same time, overestimate the current capabilities and culture of their own organization. Fully understanding the impact of this process on business and people helps to avoid the pitfalls into which so many repeatedly fall. A common mistake many organizations make is to label digital transformation an IT project. It is seen as the responsibility of the IT team to take the lead, while the necessary business inputs are provided only half-heartedly or not at all. Consequently, the project takes a wrong turn at an early stage, and the finished product eventually falls short of internal expectations and customers’ needs. A Word on Transformation As an industry, we’ve identified the value of digital technology to drive business results. But when it comes to actually putting them into motion, most companies pay lip service to digital transformation. Many believe it is about using shiny new technology to do more of the same things they have done before (i.e., digital stamps vs. printed stamps). In the worst-case scenarios, this may mean doing the wrong things even faster, leading to the simple formula “old process + new technology = expensive old process.” Today, many digital transformation projects are focused on the “digital” and not so much on “transformation.” Instead, real digital transformation requires change at a much deeper level. It calls for action that cuts across every aspect of how postal organizations operate internally and engage externally. This process is less about technology and more about cultural change. It includes elements of understanding how to interpret data and leverage technology so that it shifts every corner of the business. But equally, transformation involves understanding how to implement those shifts so that

the organization can evolve. Millennials are the generation best positioned to challenge your thinking and enable true digital transformation. If you accept that millennial talent will help position postal organizations for the future, then the question becomes, “How do you attract them?” Postal organizations should consider a few things along the way.

parcel operators. Embracing the thing that many see as a negative will enable millennials inner need for shared value and purpose which will increase the likelihood that they will stay with your organization. Remember, attracting them is only step one. Retention is important, too. As for the Stamp

Location Matters Many distribution centers are located on city fringes with easy access to main transport roads. These remote locations, while great for moving goods, are far away from the urban centers with their deep, millennial talent pools. In the race for millennial staff, neither such a working environment nor location qualifies as a good starting position. Digital transformation offers organizations the ability to centralize digitally based planning and operational support teams. Instead of planning independently at remote locations, centralization allows for synergies across the entire network of distribution hubs, depots, and, of course, the broader organization. Furthermore, a centralized office can be located close to any urban hotspot offering easy access to a high density of top talents. Instead of a remote and foreign environment, a centralized urban office offers a working environment that will not only attract millennials but also allow them to prosper. Embrace the Challenge Instead of clinging to aging processes and tools, transformation is needed for post and parcel operators to survive and prosper in a digital world. It is not a question of “if” or “when” to start the journey but rather “how.” Millennials are inherently poised to drive transformation. Their natural tendency to disrupt and challenge the norm is the very thing that will deliver lasting value to post and

You made it to the end of an article on millennials – good on you! However, there’s a bit more… You must remember that millennial’s prominence in the workforce is only a reflection of their prominence as purchasers of goods and services. In 2012, it was estimated that 50.5% of the world’s population was under 30. Their buying power will only strengthen over the next decade. Organizations that embrace the “never licked a stamp” generation now will be the first to reap the benefits as they can align their organizational thinking with the future purchasers of their services. And, as for the fate of the stamp? This comes from the millennial age coauthor: Isn’t it rather odd that we still have something as disposable as a single-use sticker on a single-use envelop, that serves a single purpose at the core of our industry? While stamp sales are on the rise is it not time for us to evolve past the traditional stamp/envelop combo to something more sustainable, more in line with the digital nature of communication? Perhaps, it’s time for a YouTube video featuring kids who have no idea what a stamp is or how to use it…

“Really, you lick it?”



EMERGING TECH The technology world around us is constantly changing. While the pace of change for post and parcel operators is a bit slower, it isn’t entirely isolated from emerging tech and its potential impact. Having only just gotten our heads around the last round of game changers like big data, cloud, and the Internet of Things (IoT) you might be discouraged by the idea of more technology. But stay the course as some of this tech could significantly impact how you run your post and parcel business in 5-10 years’ time. By Dr. Eva Savelsberg and Matthew Wittemeier


Emerging Technologies There are five emerging technologies that are worth keeping an eye on. Let’s define each and then look at how they might impact the post and parcel industry.

look. AR also has a role to play in optimizing last-mile delivery. From augmented delivery maps laid onto the drivers actual view to applications that assist with packing delivery vans and finding parcels – major gains can be made here in the coming years.

Virtual Reality (VR)

tomorrow. As you implement tech like drones, robots and self-driving vehicles – these technologies will make decisions on their own. We need to be comfortable with that or face a bandwidth bottleneck that inhibits your organization. Digital Twins

Blockchain Virtual reality is defined as an artificial, computer-generated simulation of a three-dimensional environment requiring dedicated hardware (Google Cardboard, headset, controller, etc.) to generate an immersive user experience not dissimilar to first-head reality. Major retailers around the world are looking to VR as the next big thing in online shopping. The advantage of VR over traditional online shopping is products that are hard to buy online now (i.e., fashion products) will be a breeze in the digital environment. With the rise in VR shopping, parcel operators should expect a subsequent rise in parcel volumes. It’s probably too early to call it for sure, but we think VR-commerce will see an even greater growth in parcel volumes over what will become its antiquated predecessor, e-commerce. Augmented Reality (AR) Augmented reality layers digital elements and data onto the real-world blending digital information and real-world imagery together to create an enhanced experience. Its main advantage over VR is that it doesn’t require dedicated hardware. And with major smart-device manufacturers like Apple announcing both hardware and software level support for AR, we’re likely to see an increase in its adoption across the board. Again, the impact to parcel operators lies in the growth of e-commerce and parcel volumes. The proof is in the pudding. Just this year L’Oreal invested big in AR allowing their products to be superimposed onto you in real-time allowing you to preview how a beauty product would actually

Blockchain is a technology platform that provides a secure, digital ledger in which transactions are recorded in chronological order and made public across multiple nodes that hold a shared copy of the ledger. Blockchain must be the tech buzzword of 2018 – and is showing no signs of slowing down as we look to 2019. We have all heard of it, likely because of the extravagant value of Bitcoin at the end of 2017. Nevertheless, blockchain stands to transform every industry in the world and post and parcel is no exception. Whether we’re talking end-toend traceability (real accountability) of post and parcel items or postal operators fulfilling id management services with the blockchain; it is coming.

Edge Computing Edge computing allows data produced by IoT devices to be processed closer to where it is created instead of sending it across networks to data centers (or the cloud). In short, edge computing leverages the massive increases in computing power that is now common place. We carry more computing power in our pockets to make phones calls (our smartphones) than was available to send men to the moon in the 1960’s. This coupled with the fact that reliable data transfer and bandwidth have remained challenges and one can see why edge computing is on the rise. For post and parcel operators, it means that the smart devices you’re investing in today may become peripheral decision-makers

A digital twin is a virtual model of a process, asset, product, or service. IoT devices with digital twins pair the virtual and physical worlds together allowing an in-depth analysis of data and real-time monitoring of systems to improve one’s ability to respond to real-world scenarios. It isn’t hard to imagine a parcel network in 5-10 years where every vehicle, load-unit, staff member, and even parcel has a digital twin inside your planning and optimization systems allowing you to predict problems and make better decisions in real-time. Powering the Future Some of the above five technologies will come to play a significant role in the future. Exactly how and where will, of course, remain to be seen. But they each possess the ability to add value to post and parcel operators – whether that value be monetary, social, or cultural. Like it or not, the future is one that will be powered by technology and there are two more technologies to add to the mix. But they are special because they enable each of the technologies introduced above in some way shape or form as well as offering their own value. Artificial Intelligence (AI) Artificial Intelligence is an area of computer science that‘s concerned with building systems that demonstrate intelligent behavior. Most people find it difficult to agree on a precise definition of intelligence, and so people‘s view of what Artificial Intelligence means also tends


to diverge. For most people, when they hear the term Artificial Intelligence, or AI, they think of a General AI, or human-level AI, that can mimic all aspects of human intelligence. The simple truth, however, is that today, AI is far away from this. Instead, AI vendors have succeeded in building niche, or so-called Narrow AI systems that know how to do reasonably specific things very well (for instance, play chess, translate between languages, understand natural language, or drive autonomous vehicles). It is these Narrow AI systems that are now making their way into our industry at a rapid pace as part of the Fourth Industrial Revolution. Machine Learning (ML) In contrast with General AI’s goal of mimicking human intelligence, Machine Learning tools use algorithms to iteratively learn from and adapt to data, enabling computers to find hidden insights without being instructed where to look. A beginner’s example for this can be found in your email inbox: spam filters. Simple rule-based filters are not very effective against spam, since spammers can quickly update their messages to work around them. Instead, ML enhanced spam filters continuously learn from a variety of signals and tailor themselves to the email needs of the individual user. ML is already making its way into hubs around the globe as a means to mine data to improve processes. Closing Thoughts – Technology and Millennials Matter Some of the technology discussed may never impact the logistics industry. Some might take longer than 5-10 years. In discussing them here, the point isn’t to say that they are all going to revolutionize your operation. Some of it may be too radical for some operators or their management teams. That’s ok. Rather, the point is to put it on your radar, to bring

it to the front of your mind. One should consider the plight of the taxi industry globally when choosing to ignore or embrace new technology. Smartphones and apps didn’t appear overnight but had taxi companies the world over chosen to be innovative and be progressive, UBER, and the taxi companies fight to remain relevant, probably wouldn’t exist today.

When you combine the technological innovations and an entire generation of digital natives – the Millennials – we are headed directly into an era where technology will not support post and parcel operations, but rather define them On the whole, the post and parcel industry is comprised predominately of baby boomers with a small proportion of Gen X, and Gen Y workers. In short, we have an aging workforce problem.

Attracting millennials is as much about attracting young workers as it is about attracting skilled workers and underlying the skills young people bring is tech. By 2025, millennials will make up 75% of the global workforce, i.e. the generation of “pen and paper” operators is a dying species. Millennials have grown up with instantaneous communication and “high-tech” is woven into all aspects and areas of their life. Millennials will not only penetrate the logistics workforce in the maritime industry, they will also be the ones who drive demand on the customer side. “On-demand” and “digital” is their way of life – anything else will not be accepted; technology is at the core of this generation. When you combine the technological innovations and an entire generation of digital natives – the Millennials – we are headed directly into an era where technology will not support post and parcel operations, but rather define them. We’re seeing the start of this with automation, where processes have been redefined to suite robotic equipment. The addition of technologies like AI and Machine Learning (ML) will see more significant changes to come. The innovators in the industry have the power, now, to define what the operators of the future will look like and those that lag behind their innovation will have no choice but to meet that standard or risk becoming irrelevant.




A Win-Win

Pushing Your Sortation System Further By: Eva Savelsberg, Markus Sekula, and Matthew Wittemeier


Pushing your Sortation System Further Learn how to balance your sortation system’s efficiency to sub-90% levels while also improving the safety and productivity of your yard environment. By utilizing a yard management system (YMS) empowered with AI algorithms, you can derive significant benefits for your sortation system and your overall operations including up to an 18% reduction in total task time to unload, sort, and load parcels all from better yard-based decision-making.

More Info https://infrm.co/winwin


TO BOLDLY GO Understanding the Sceptical User By Dr. Eva Savelsberg and Matthew Wittemeier

In 2016 Pokemon Go was launched and the app eached 50 million users in 19 days! The pace of technological adoption is quickening. The challenges that hindered the adoption of the telephone are all but non-existent today.


The world is moving forward quickly. What was once science fiction (the internet, robots, artificial intelligence, etc), is increasing commonplace. Underlying these innovations are challenges around both technology and how humans interact with new technology. Understanding both is crucial to addressing why users are resistant to technological innovations. Equally, it is paramount in fostering a path forward so that new technology solutions can drive value instead of floundering in the hands of sceptical users.

Time to Reach 50m Users TV 1950s

13 years

Telephone 1876

75 years

Technology is Here to Stay “Come here. I want to see you,” were the first words communicated over the telephone by Alexander Graham Bell to his assistant in 1876. After that first call, Bell penned a letter to his father where he noted, “... the day is coming when [telephone cables] will be laid on to houses just like water or gas - and friends converse with each other without leaving home.” Despite its revolutionary ability to connect people anywhere, anytime, it took approximately 75 years for the telephone to reach 50 million users. A lack of infrastructure and technological constraints are generally the two factors noted when discussing the very slow adoption of the technology. Fast forward to the 1950’s and the TV was introduced; it took TV about 13 years to reach 50 million users. Fast forward again to the late 80’s and the first commercially available internet hit the market - it took approximately 4 years to reach 50 million users. In 2016 Pokemon Go was launched and the app reached 50 million users in 19 days! The pace of technological adoption is quickening. The challenges that hindered the adoption of the telephone are all but non-existent today. Today, the internet serves as a common backbone for almost all technological innovation. While not perfect, it’s common use architecture allows anyone,

Internet 1980s

4 years

Pokemon Go 2016

19 days

Time to Solve a Planning Problem from 1990... ...to today 100 years

1 second


anywhere to develop and distribute new technologies with ease. Further, since the mid 1900’s Moore’s law has seen the steady doubling of technological capability every two years or so. Today, most of us carry a mini-supercomputer in our pocket. Ironically, these supercomputers take us back to the beginning of the story; our smartphones are designed to supersede the, now, outdated telephone system. Today, consumer facing technology companies are able to innovate on timeframes measured in months and users can adopt those innovations in mere days. While the pace of innovation and adoption is slower in the enterprise IT world - it too has dramatically increased over the past two decades. Over the past 25 years, improvements in computer hardware have resulted in an increase in computing power by a factor of 2,000 times. This seems impressive until one compares it to the advances in optimization algorithms over the same period. For instance, Linear Programming algorithms, considered the most important class of optimization techniques by many experts, have improved by a factor of 1.4 million times. When combined, the effects of both advances generate a tremendous 2.8 billion times improvement in processing capability. To better understand this, a planning model, using linear programming, that takes us a second to solve today, would have taken almost 100 years to solve in the 90’s. When you combine the technological innovations we are capable of today with an entire generation of digital natives – the Millennials – we are headed directly into an era where technology will not support hub operations, but rather define them. We’re seeing the start of this in Industry 4.0, where processes are redefined to suite robotic equipment. The addition of technologies like AI and Machine Learning (ML) will see more significant changes to come.

We’re Creatures of Habit After decades of implementing technology systems, we’ve learned a lot about what users like and dislike about systems that offer decision support. Predominantly, most users are sceptics, at least to begin with. Why? What are their concerns underpinning the use of advanced technology? And, how do you implement change that does not cause disruption? In psychology, the measure that best aligns with a willingness (or lack thereof) to try new things is called “openness” – it is one of the “Big Five” personality traits and is well researched and well documented. In most cases, openness follows a normal distribution, meaning that some people are very open to new experiences, some are really closed to new experiences, and the rest lie somewhere in the middle. Those in the middle are likely to have areas of their life where they are more open, and areas where they are more closed.

Job loss, or even the fear of job loss, is an extremely powerful fear. Some psychologists argue that the emotions surrounding job loss are on par with those surrounding any kind of major emotional loss Interestingly, those who are more open to new experiences are more likely to progress up the corporate ladder into leadership roles. These are also the individuals who are often responsible for specifying and selecting new systems. On the flip side, their colleagues are ge-

nerally more set to routines – a trait that is strongly associated with moderate to low openness. Experiences that challenge those routines are almost always interpreted in a negative context. The saying, ‘we’re creatures of habit’ isn’t a saying because it sounds good, but rather, because it is true. The largest concern users have when first working with decision support systems is that the system will replace them. Job loss, or even the fear of job loss, is an extremely powerful fear. Some psychologists argue that the emotions surrounding job loss are on par with those surrounding any kind of major emotional loss. This fear can be exaggerated depending on how the project is introduced to the users. When users are not actively involved in defining the criteria and in the selection process of the system they are often taken off-guard when the system is ready to be implemented. It takes time for users to trust that an advanced decision support system isn‘t going to take their jobs, however, this is only their first challenge to overcome. New, complex systems come with the challenge of learning a new system. Senior staff, who often already feel challenged by technology are often stressed by the idea of learning how to use a new system. This fear is exacerbated when younger employees are able to learn the new system more quickly – in these cases, senior staff can often revert to fears of job loss. There are also feelings of distrust. “If it isn’t broke, don’t fix it.” Many users are confused as to why new systems are implemented. They believe that they’ve managed to run the hub successfully without a technology system and feel as though their experience and skills are underrated by its introduction – especially with decision support systems that are capable of making the majority of the decisions they would normally make. These feelings of distrust often lead to features within systems being disabled or ignored.


When implementing a new system, parcel center operators need to understand that the project is more than a technology project that will deliver specific management KPIs by improving operational parameters ‘A’, ‘B’, and ‘C’ leading to an ROI of ‘X’ over ‘N’ years. Implementing a new system is as much a project about cultural change as it is about operational change and to implement cultural change, change management processes can be of great benefit. It is estimated that only 54% of major change projects are successful. Those that fail are plagued by higher than expected costs and lowered employee morale. Studies also show that when employees see major projects fail, or fail to deliver major elements, cynicism sets in, which in turn, further undermines adoption, utilization, and worse – company culture. Change management is a well research branch of social and business science with many models and techniques that can be implemented. Of the many available, there are some common elements such as: involve every layer of your organization throughout the entire process, work from within your culture to implement change, and continuously assess and adapt your project to suit the combined technological and cultural needs of your organization. Getting Technology Implementation Right With the onslaught of technological innovation, one could be left wondering whether technology should just replace humans to avoid the challenges a human element adds to a technology project. It is an easy conclusion to draw when you consider that today, decision support systems can run a hub’s daily operations automatically with little, to no human intervention. Add to this equation the advancements in AI and Machine Learning (ML), which are only enhancing an intel-

ligent system’s learning and decisionmaking capabilities, and one can easily envisage a future where the human operators of post and parcel centers will be obsolete in due course. History can lend a hand here. A technology we all consider commonplace today revolutionized its industry and the world after it was introduced in 1967 in Enfield, London. When it was conceived, the same questions that face the post and parcel industry today were present – but how the story unfolded isn’t as straight-forward as you might think. The technology, of course, is the Automated Teller Machine, or ATM as it became commonly known to the world. When you move past the marketing spin behind the development of the ATM (offering customers convenience), banks pursued the technology for two primary reasons: addressing workforce limitations (banking unions in the UK wanted banks to close on Saturday) and reducing costs (operational labor costs to be more precise). Well we all know that the story of the ATM was a success for the technology, but how about the humans? ATMs achieved their goals of allowing banks to reduce their operational hours and reduce the total number of tellers required per branch (less humans). The twist is, that because banks could allow for greater cost controls per branch, the number of branches increased dramatically which in turn lead to an increase in bank teller positions in the market (more humans). Further, their roles evolved from completing the mundane, simple task of dispensing and collecting money, to ones which added increased value to banks, such as selling services or improving customer service outcomes. There is one more plot twist with the ATM story. ATMs lead to the creation of an entirely new service industry. Never before had a technology needed to be so exposed to the elements – the mean time between failures was low and hu-

mans were the intervention to resolve the issues that arose. The rise of the ATM also led to the creation of the ATM service technician and the birth of an industry. We see the same story play out across other industries too. The introduction of automated weaver technology in the 19th century led to an increase in the number of weavers. The introduction of electronic discovery (e-discovery) software in legal offices in the 1990’s led to an increase in paralegals. The moral of these stories – technology enables humans to achieve. When humans are relieved of mundane tasks, they are enabled to focus on higher level problems that technology isn’t yet capable of automating. The future of humans in post and parcel operations isn’t bleak. In fact, it is likely to be better than ever before. As operators begin to allow intelligent systems to assume the day-to-day decision-making in hub operations, they will be free to focus their skills and experience at solving larger operational challenges, managing by exception, and improving customer service outcomes. The Operator of the Future Logistics operators who excel in the future will be the ones who develop partnerships with their technology partners. As the pace of innovation continues to quicken, traditional specification, development, and delivery business models will not suit the post and parcel industry. New business models, based on mutual trust, shared risk, and joint reward will be the defining characteristics of successful logistics operators of the future. These new contracts will allow for flexibility in project delivery that does not exist today, while also allowing for improved outcomes for end users of the systems and better cultural outcomes for the organization.



MACHINE LEARNING IN TERMINAL OPERATIONS A Practical Review of MLs Impact on Yard Optimization By Dr. Eva Savelsberg, Ulrich Dorndorf, and Matthew Wittemeier



In 2018, INFORM‘s Machine Learning (ML) assessment project, aimed to achieve two results. Firstly, could INFORM’s broader ML algorithms, developed for use in other industries such as finance, be applied to our Optimization Modules used in terminals around the world. And secondly, if so, apply them to real-world terminal data and identify areas where improvements could be made.

Last year we wrote about how Artificial Intelligence (AI) had made, and was continuing to make, its way into the terminal industry and further, how Machine Learning (ML), as a branch of AI could be implemented. In a piece that likened AI’s current position to that of Frankenstein, the article closed by saying that AI was coming and that as an industry we can either be prepared or caught off-guard when it does. For INFORM, as a leading AI solution provider, the question wasn’t how to prepare for AI, but rather, how could we leverage the promise of Machine Learning and build it into our core AI driven solution? As such, in 2018, INFORM undertook a Machine Learning (ML) assessment project, looking at maritime container terminals and how ML could be used to improve operational and optimization outcomes. The assessment aimed to achieve two results. Firstly, could INFORM’s broader ML algorithms, developed for use in other industries such as finance, be applied to our Optimization Modules used in terminals around the world. And secondly, if so, apply them to real-world terminal data and identify areas where improvements could be made to para-meters that influence the optimization calculations of INFORM’s add-on Optimization Modules. Working with a randomized sample of 1 million containers handled in the 2017 calendar year with 50 data variables (explanatory variables) at a selected terminal, we set off to answer these two questions. The dataset was split further; using a time slicing method, a training dataset (75% of the dataset) and a testing dataset (25% of the dataset) were created in accordance with good ML practices. Further, we worked with an human export to review and identify variables amongst the 50 explanatory variables that would prove meaningful in the assessment. We identified 16 variables that have been used to build the random forest ML models presented later in this paper (see figures 2 and 6).

In short, the answer to the first question was a resounding yes; INFORM’s ML algorithms could be applied to work with our solution for the maritime terminal industry. To answer part two, it is worth exploring some of the areas of terminal optimization where it was identified that we could further improve our solution offering through the implementation of Machine Learning. While we identified many areas, we will focus on container dwell time and outbound mode of transport. Predicting Outbound Mode of Transport We started with the outbound mode of transport predictions. The mosaic plot (figure 1) below shows what was expected based on the TOS information received upon container arrival versus how the container actually departed the terminal. The areas represented are proportional to the volume of containers handled. “NA” areas in grey, were unknown to the TOS when the container arrived. The data the TOS is configured to use to make decisions is accurate in only 62.9% of the total 1 million boxes sampled. A random forest with 500 trees ML model was trained using the training dataset and subsequently tested against the testing dataset. The forest was subsequently tasked with identifying the importance of the pre-identified 16 variables (see figure 2). Random forest ML models are very flexible algorithms that produce balanced results for classification and regression tasks. They are created by building a multitude of decision trees, in our case 500 trees, and then outputting the mode (classification) or mean (regression) of the individual trees. Importantly, random forests correct for a single decision trees’ tendency to overfit to the training data. Finally, random forests improve upon the predictive power of single decision trees by making clever use of random chance.




The findings: using the revised ML generated prediction model opposed to the data available from the TOS upon container arrival would increase prediction accuracy to 83.6% corresponding to a relative improvement in prediction accuracy of 33%. Figure 3 below maps the improvement in accuracy from the TOS to the ML model against each outbound mode of transport (OMT). Looking into the data more closely, the TOS had a good accuracy at predicting OMT for containers leaving by ship (81.9%), average accuracy for truck OMT (65.5%), and poor accuracy for feeder (42.4%) and rail (4.7%). In comparison, the accuracy for all OMT shifts to average or much better: ship (94.3%), truck (87.2%), feeder (76.3%), and rail (53.0%). Predicting Container Dwell Time Container dwell time is used within INFORM’s Optimization Modules to assist with container yard positioning calculations. The basic logic is straight-forward; when building stacks in your yard, place containers with longer dwell times at the bottom and containers with shorter dwell times at the top. In this way, you minimize the number of rehandles needed to retrieve containers for their outbound journey. Given its relevance, dwell time is a central variable in optimizing the placement of containers in one’s terminal. However, the data point used to calculate dwell time – expected departure time – is frequently missing. In our dataset, 47% of containers were missing an expected departure time. This is visually represented in figure 1; data was available for ship (orange) and feeder (blue). Traditionally, for these instances, our optimization modules use a strategically calculated and pre-configured dwell time variable as a stand-in. Working with the dataset, we drew up an empirical model to determine that the mean dwell time for loaded containers where there was no expected departure time upon container arrival. The mean was calculated at 84 hours (see figure 4). Not too far off of the selected systems pre-configured 96 hour stand-in variable, so using basic statistical modelling, we can already achieve a small improvement. From there, we decided to see what would happen when we factor in the dwell time versus the expected departure mode. A different picture emerged (see figure 5). Containers leaving by ship stay longer, while containers leaving the terminal by truck and rail have, on average, a significantly shorter dwell time. There was a correlation between the complimentary storage duration offered by the terminal and the associated outbound mode of transport; that said, it was not the aim of this assessment to evaluate this finding and further assessment is needed to confirm causation vs association. Working from the hypothesis that a better stand in dwell time could be predicted if the system took into consideration


the OMT, a random forest with 500 trees ML model was trained on the training dataset and subsequently tested against the testing dataset. Again, the forest was tasked with identifying the importance of the pre-identified 16 variables (see figure 6). Interestingly, weekday proved to be highly relevant. The ML model found that containers arriving on Thursday or Friday were likely to remain longer than those arriving Saturday through Wednesday. Our human expert attributed this to reduced operational hours over the weekend period. Using the revised ML generated prediction model for dwell time instead of the standard dwell time variable resulted in relative improvement in prediction accuracy of 26.8%. Opportunities for Further Assessment From there, the assessment should aim to review available data from 2018 and run it against the same process to assess whether the findings from the 2017 data are consistent with more current data or what alternative patterns are seen. Further, considering external data sets could add additional insights beyond that of the core container data. For instance, vessel ETA vs ATA and ETD and ATD differences could reveal additional patterns that would improve the dwell time prediction model. It is reasonable to assume that seasonal (monthly or quarterly) differences are also plausible due to container traffic patterns or the impact of weather. Application in Terminal Operations It is expected that the ML models will lead to improved container location selection (OMT) and improved precise stack location (dwell time) both resulting in fewer rehandles. Let’s assume you can reduce rehandles by 1% in our example dataset of 1 million containers that is 10,000 fewer moves. Further, let’s assume that the rehandle cost in a typical EU terminal is approximately 80 Euro, this would result in an annual savings of 800,000 Euro for every 1% decrease in rehandles. Using the Machine Learning Module within INFORM’s optimization solution, the application of ML to review and improve predicted outbound mode of transport, or OMT, and container dwell times should be run on a regular basis. Further, the output should be reviewed by a human expert before being used to modify optimization parameters. As noted in our previous paper, this expert discussion review process will assist operators in firstly, understanding the changes made inside of the optimization systems. And secondly, allow human operators to both learn from the output as well as gain confidence in the ML Module’s ability to recommend appropriate parameter improvements for future use.






A digital twin is a virtual model of a physical asset, process, or system. Pioneered by NASA in the early years of space exploration, it allows many industries today to understand and manage the operations of their remote machines and assets. This article will review use cases and benefits for digital twin technology in bulk material logistics.



In April 1970, NASA sent astronauts Jim Lovell, Jack Swigert, and Fred Haise on America‘s third landing mission to the moon. Two days into the flight, and more than 200,000 miles from earth, disaster struck Apollo 13: an explosion rocked the spaceship, and soon its oxygen and power began draining away. “Houston, we have problem“, were the famous words that the crew radioed to mission control immediately after they heard the loud bang. NASA’s engineers solved the problem by constructing a twin of the component they were trying to fix, using only physical parts that the astronauts in the capsule had available to them. While there was nothing digital about the process, mirrored systems became the precursor of digital twins. And nearly 50 years later, this technology allows many industries to understand and manage the operations of their remote machines and assets. Down to Earth A digital twin is a virtual model of a physical asset, process, or system. As conditions change, the digital twin reports those changes in real-time, whether it is a bearing in a roller mill, a chain in a bucket elevator, or a cement truck stuck in traffic. Combining the virtual and physical world allows cement producers to avoid problems before they occur, prevent downtime, and even plan the next steps using simulations. The ultimate goal is to have a digital twin running for every real-world asset in the field, with the digital replica updating its status as it receives operational data. Apollo-era data acquisition technologies got mankind to the moon, and back, nine times. But sensors have advanced dramatically since then and with the rise of the Internet of Things (IoT) they have become connected too. Computing has emer-

ged as a cheap and abundant resource that can be deployed against any problem, making large-scale digital twin modelling cost-effective for a wide range of applications. A Giant Leap for Dispatchers The daily mission of a dispatcher in a cement company is to determine the delivery schedule and fleet configuration for the following shift(s); decision-making at this stage is complex and intricate. Each decision has multiple flow-on decisions that, in turn, impacts future decisions. Coming up with an optimized plan is a real brain teaser. However, it offers the potential for great savings if done right. State-of-the-art planning tools use algorithms and Artificial Intelligence to analyse a virtually endless number of scheduling decisions in real-time and identify those that are ideal for minimizing costs and maximizing service quality – based on the business criteria defined. The software allows dispatchers to make incredibly complex, time-critical decisions with ease. What’s more, it offers great visibility into all logistics assets. With a digital copy of each truck it allows dispatchers to drill down to the deepest level of detail to analyse each transaction and move. At the push of a button, dispatchers can run different “what-if-scenarios” and model the outcome of even minor changes to the truck/order set-up. The decisions made also take into account a larger range of variables than the human mind can, resulting in better overall decision quality. And like the men on the ground in Houston, it enables dispatchers to come up with actionable plans and steps that get the job done. Cement producers who use intelligent optimization software powered by algorithms typically achieve: • • •

Reality vs Real-time

A reduction in truck fleet size by 10 to 30% A reduction in empty mileage by almost 9% An increase in loads/truck/day by up to 30% Failure is Not an Option


Fig. 1: 2: Digital twins and real-time capabilities of telematics systems. systems

NASA flight director Gene Kranz was the man behind the team that got the Apollo 13 crew home safely. Portrayed in the blockbuster movie “Apollo 13” and best known for his flattop haircut and white vest, he was also author of the book “Failure is not an option” in which he recounts the details of this mission. One of the most critical decisions he had to make was to choose between firing the spacecraft’s rockets and returning it home immediately as it drifted away from earth, or using the


moon to slingshot the Apollo 13 capsule back to earth. Despite the longer route, he picked the latter option which proved to be the key to success. Cement plants are often found in remote locations and cement truck drivers face long hauls to reach their final customer destinations. Telematics systems form the communication backbone between the trucks and the supporting customer service center where new orders are taken and transport plans are updated automatically by the planning software. Telematics allow dispatchers to track every mile on the road and every heartbeat of the engine is captured. The real-time data is vital to keep a digital twin system of the entire fleet running. For many, the term “real-time” means immediate response. However, uptime of the system, accuracy of the status messages and sampling rates as well as accuracy of GPS readings due to external factors may vary between telematics providers. So depending on the type of telematics services used, the term “near real-time” might be more appropriate. Figure 1 shows what effect this might have for the planning process. The orange truck is crossing the geofence around the cement plant ahead of the green truck. But due to different GPS signal sampling-times, the green truck’s digital twin appears to be ahead of the orange truck’s twin in the “real-time” view of the planning software. If the green truck is running late for his next shipment, the software assumes that loading will start shortly and time can be made up. In reality, the orange truck will enter the loading bay first – adding further delay to the green truck. The telematics market is very competitive and crowded. It is easy to get dazzled by fancy features. So buyers should make sure they get the best out of their investment. Telematics products that are tailored to the needs of the construction material industry make it easier to access relevant and actionable data for the transport planning software. Houston, We Have a Program NASA’s mission control center is based in Houston, Texas and was built to coordinate the US manned spaceflight program. It centrally manages space flights from point of launch until landing. Flight controllers and other support personnel monitor all aspects of the mission using telemetry and send commands to the spacecraft. A key element of digital twin technology in logistics is the centralization of all planning and dispatching units. Instead of planning independently at a local level, centralization unlocks synergies across the entire network of cement plants, terminals, depots, and, of course, planning teams. A centralized customer service office can be located close to any urban hotspot with a high density of top talents. Instead of dust, heat, and noise, centralized offices offer a call center atmosphere, which in


Fig. 2: Centralized planning is a key element of digital twin technology.

turn attracts more female staff. Like in many traditional industries, female workers are hugely underrepresented in the cement industry. This hurts even more, since research has shown that gender-balanced teams outperform homogenous teams by means of productivity and financial performance. The Dusty Road to Digital Twins The Lunar Roving Vehicle (LRV), or simply called the “moon buggy”, was a battery-powered four-wheeled vehicle used in the last three missions of the Apollo program. LRVs allowed the astronauts to explore beyond their landing site, although the non-rechargeable battery limited their range. It marked the beginning of a new technology to overcome many challenging problems for which there was no precedent in vehicle design and operations. Logistics and IT have come a long way since then. And when it comes to planning and optimization, digital twin technology goes far beyond traditional tools. Moving forward, there is more to explore and discover for the bulk materials industry. Artificial Intelligence and Machine Learning are about to further enhance the decision-making quality of the planning software. Semi-autonomous vehicles and trucks equipped with platooning technology will hit the logistics industry soon and on-site charging terminals or battery swapping stations for electric trucks may become a common sight at many cement plants. And undoubtedly, connected vehicles will transform mobility soon. The next generation of telematic systems will feature vehicle-to-vehicle and vehicle-to-infrastructure communications, allowing trucks to exchange data between nearby vehicles as well as roadway infrastructure. This has the potential to move telematics for digital twin technology from data capture and reporting to on-board actions based upon real-time conditions.



SHIFTING THE DYNAMICS OF WORKFORCE MANAGEMENT Algorithms are a major efficiency driver for logistics assets in the cement industry. However, many producers still lose traction when it comes to creating optimal shift schedules for their human assets. This article will explore latest developments and technology in workforce management and discuss how they can be applied across the cement sector.


In the “olden days,” shifting gears was hard work. Back then, commercial trucks came with unsynchronised manual transmissions and drivers had to use a method called “double-clutching” to prevent damage to the gearbox when changing gears. It took some timing and practice, and it came at the expense of extra work for your left leg. Today, auto-shift gearboxes are commonplace in heavy trucks, and changing gears requires virtually no effort at all. However, when it comes to managing shift work, many manufacturing companies still use old technology that is suited neither to synchronising the competing interests of employee needs and operational objectives, nor providing the planning comfort or intuitive logic of the latest workforce management tools. This often results in expensive overtime, non-productive idle times, lower employee morale, poor customer service and, in a worst case scenario, loss of production. Failing to adequately schedule your workforce can become extremely costly in the long-run. Before we review some potential application areas and benefits within the cement industry, let’s take a look under the hood to explore the basic technology that powers the latest workforce management tools. DI Technology Not all gearboxes are built the same, which is why there’s a myriad of different and often confusing brand names. The same applies to workforce management: Rostering, staff scheduling, employee logistics, shift planning, resource planning... Almost every organisation has a different term and approach. Employee logistics, however, is quite a fitting term since the latest workforce management tools are based on the same technology that has been deployed in the building materials industry


for over 25 years to optimise the use of logistics assets: Operations Research (OR) and algorithms. In the mid-1990s, Redlands in France (now Lafarge-Holcim) was an early adopter in the aggregates and ready-mix business. Six years later, Hanson Australia (part of the Heidelberg Cement Group), followed. Both have been using algorithms, real-time information and automated decision-making to run their fleets of trucks ever since. But even if there is a large number of petrolheads among your workforce, human specs are quite different compared to trucks and other logistics assets. While the logistical processes of a cement producer are usually programmed into the transport optimisation software, workforce management tools need to be more flexible to accommodate the requirements of human assets. With so-called Deductive Intelligence (DI), the representable logic and structure remain flexible. This allows experts to easily formulate requirements without touching the programming level.

The first major challenge to sharing data is an industry wide acceptance that data is important Deduction is an important area of artificial intelligence (AI) and many systems rely on deduction to solve problems. With this top-down logic, conclusions are reached by applying general rules to observations. Or as the Greek philosopher Aristotle, considered by many to be the father of deductive reasoning, might say: “All cars with manual transmission have a gear stick. My car has a gear stick. Therefore, my car has a manual transmission.” Workforce management tools equipped with DI technology enable planners to easily weigh factors according to their priorities, e.g., by costs, service level, shift ergonomics, or employee satisfaction. With this unique technology in mind, let’s review some potential application areas and benefits within the cement industry. Central Shifts

Fig. 1: Shifting the dynamics of workforce management.

A transmission control unit is a device that controls modern electronic automatic transmissions. It centrally collects vehicle data and by evaluating information about speed, acceleration, road grade and torque demand, it applies extreme precision to every shift. In contrast, shift scheduling in the cement industry is often managed by local teams and within their specific departments, e.g., logistics, manufacturing, maintenance, research and development and others. Some are lucky enough to have an ERP system to support them, but many still rely on



MS Excel or, let’s face it, pen and paper. This ‘silo approach’ has its limitations, including the fact that shift leaders are experts in their field of expertise, but usually lack the time and skill set to create optimised shift schedules. A centralised tool equipped with DI technology can analyse a larger range of variables than the human mind is able to, resulting in better overall decision quality. What’s more, it finds the best possible balance for all legal, operational and individual requirements. But centralised scheduling does not stop at the gate. It can span over several cement plants, quarries, depots and terminals and integrate to other vertically-integrated assets such as concrete batching plants. Again, transport planning in our industry can serve as an example: centralised planning has been an integral part of the truck fleet optimisation cases mentioned above – driving synergies and unlocking value across all corners of the business. Flexible Shifts Fully-laden trucks accelerate more slowly than cars, take up more space for manoeuvering and need more time to come to a stop. The same applies to the corporate world: The larger the business, the slower the movement. Traditionally, many cement producers use rigid shifts and simple rotating patterns,

Fig 2: DI Technology for workforce management.

e.g., week one early shift, week two mid-day shift, week three night shift. Rotating shifts are popular among shift managers since they can be managed easily by spreadsheet tools. However, they do not cater for the shifting needs of a younger workforce generation.1 By 2025, millennials, workers who have only been adults in the 21st Century, will make up 75% of the global workforce. Our industry needs to find ways to be attractive for this digital-savvy generation. One thing they take for granted is flexible schedules that help them to achieve a healthy worklife balance. However, flexible shifts are the stuff of nightmares for any shift planner. Nevertheless, with software tools based on OR and algorithms, more granular start, break, and finish times can be assigned to each individual worker, while keeping the overall staffing at an optimised level and in-sync with targeted production goals. Seasonal Shifts In cold weather, the effort it takes to shift gears can increase due to the higher viscosity of the transmission fluid. This may result in higher wear and tear of the components. When temperatures drop, cement producers are faced with the prospect of lost time due to employees who call in sick. But warm



Fig 3: Algorithms help plants avoid throwing spanners into the workforce planning process.

weather also places its challenges onto the workforce planning process. A summer vacation schedule that everyone can live with is hard to find. The legal position on this is clear. In most countries, cement producers are legally entitled to restrict annual leave for their staff, e.g., at high-peak periods. They can also tell their employees to take leave at certain times, for example during a planned kiln or plant shutdown. Beyond these legal guidelines, however, the key priority for any employer should be to ensure that they find a fair and consistent solution for everyone involved, that also meets the staffing requirements and shift demands. The dilemma starts with finding a consensus on which criteria vacation requests will be approved or denied. “Seniority”, where long-term employees get a first pick of the most wanted days/weeks, is a classic example. Elsewhere, employees with school-aged children might have a higher priority during the school summer holidays. You cannot please everyone, but software tools powered by algorithms allow you to add more constraints to the calculation, e.g., social factors, while providing a higher level of transparency at the same time. Evaluating Shifts An electronic logging device, also called an E-log, is a piece of hardware that is installed on an engine to record a truck driver’s hours of service (HoS). An E-log cannot be tampered with and it provides full transparency between drivers, hauliers and shippers In manufacturing environments, time and attendance systems are used to track when employees start/stop their work or take a break. Some systems also allow to record the type of work they carried out. Time recording data needs to be managed and evaluated to process the payroll. Shift work with

its many different allowances and premiums, however, is prone to inaccuracies. Add overtime or paid time-off compensation to it and there’s enough reason for workers to throw a spanner into the payroll works. Payroll errors can be very costly and time-consuming to rectify. What’s more, a single mistake can erode trust. Integrating workforce management tools and payroll software helps to reduce the amount of work required for the time evaluation process. This approach allows to automatically assess and correct deviations that fall within a specified tolerance range. Only cases outside this tolerance range need to be evaluated by the payroll accountant, while the system takes care of the routine work. Technology Shifts More than 30 years after its debut, MS Excel is still an important cog in the wheel of many cement producers. It is no secret that workforce planners and accounting professionals are among the most loyal users of the iconic spreadsheet program, mostly because it is easy to configure. However, technology has evolved dramatically over the years. When it comes to preparing complex shift schedules, even the best macro cannot compete with an optimisation engine that is powered by algorithms. Instead of clinging to ageing processes and tools, transformation is needed for cement producers to survive in a world of IoT and Industry 4.0. Producers who are content with their status quo and hesitate to invest into latest digital planning tools, can find further advice from Henry Ford, who shifted America’s Industrial Revolution into overdrive, “If you need a machine and don’t buy it, then you will ultimately find that you have paid for it and don’t have it.”


IT SYSTEMS FOR INTELLIGENT DECISIONS INFORM specializes in intelligent, decision-making IT systems. These systems optimize complex operational and logistical workflows. Integrated into the existing IT environment, they ensure that companies always make the best decision from an unmanageable number of alternatives while under great time pressure. Whereas data management software merely provides information, INFORM systems can analyze huge quantities of data, cost-out numerous decision-variants, and suggest the best-possible solution to the user for implementation in a matter of seconds. Consequently, companies can swiftly respond to market requirements, create transparency, and optimize the entire sequence of all business processes. As a result, they increase their productivity in a sustainable manner.


PROJECTS IN MEDIUM-SIZED COMPANIES AND CORPORATIONS – WORLDWIDE Our projects take us to every country in the world. How is an airbus in Toronto de-iced? How often do trucks need to supply concrete to large construction sites in Hong Kong? Our employees are experts in their specific fields, because in order to optimize operational workflows through software, we need to understand them. Ultimately, our customers expect time and cost savings in highly-complex decision-making situations – in industries like container terminals, passenger airports, financial service providers, industrial operations, wholesalers, storage and transshipment hubs, and shipping companies.


Headquarter in Aachen, Germany More than 1,000 installations worldwide Partners worldwide Office in Atlanta, GA, USA Office in Sidney, Australia


STAFF DEVELOPMENT From 5 to more than 850 software engineers, data analysts, and consultants.

1969 1985 5 15

2000 250

2010 400

today 850


Contact: Markus Sekula Key Account Manager Phone +49 (0) 2408 9456-6000 m.sekula@inform-software.com

INFORM GmbH Pascalstr. 35, 52076 Aachen, Germany inform-software.com