Category robots in business

The elephant in the room loomed large two weeks ago at the inaugural Internet of Things Consortium (IoTC) Summit in New York City. Almost every presentation began apologetically with the refrain, “In a 5G world” practically challenging the industry’s rollout goals. At one point Brigitte Daniel-Corbin, IoT Strategist with Wilco Electronic Systems, sensed the need to reassure the audience by exclaiming, ‘its not a matter of if, but when 5G will happen!’ Frontier Tech pundits too often prematurely predict hyperbolic adoption cycles, falling into the trap of most soothsaying visions. The IoTC Summit’s ability to pull back the curtain left its audience empowered with a sober roadmap forward that will ultimately drive greater innovation and profit.

The industry frustration is understandable as China announced earlier this month that 5G is now commercially available in 50 cities, including: Beijing, Shanghai and Shenzhen. In fact, the communist state beat its own 2020 objectives by rolling out the technology months ahead of plan. Already more than 10 million cellular customers have signed up for the service. China has made upgrading its cellular communications a national priority with more than 86,000 5G base stations installed to date and another 130,000 5G base stations to go live by the end of the year. In the words of Wang Xiaochu, president of China Unicom, “The commercialization of 5G technology is a great measure of [President] Xi Jinping’s strategic aim of turning China into a cyber power, as well as an important milestone in China’s information communication industry development.” By contrast the United States is still testing the technology in a number of urban zones. If a recent PC Magazine review of Verizon’s Chicago pilot is any indication of the state of the technology, the United States is very far from catching up. As one reporter complains, “I walked around for three hours and found that coverage is very spotty.”

Last year, President Trump donning a hardhat declared “My administration is focused on freeing up as much wireless spectrum as needed [to make 5G possible].” The importance of Trump’s promotional event in April could not be more understated, as so much of the future of autonomous driving, additive manufacturing, collaborative robotics, shipping & logistics, smart city infrastructure, Internet of Things (IoT), and virtual & augmented reality relies on greater bandwidth. Most experts predict that 5G will offer a 10 to 100 times improvement over fourth generation wireless. Els Baert of NetComm explains, “The main advantage that 5G offers over 4G LTE is faster speeds — primarily because there will be more spectrum available for 5G, and it uses more advanced radio technology. It will also deliver much lower latency than 4G, which will enable new applications in the [Internet of Things] space.” Unfortunately, since Trump’s photo op, the relationship with China has worsened so much that US carriers are now blocked from doing business with the largest supplier of 5G equipment, Huawei. This leaves the United States with only a handful of suppliers, including market leaders Nokia and Ericsson. The limited supply chain is exasperated by how little America is spending on upgrading its telecommunications, according to Deloitte “we conclude that the United States underspent China in wireless infrastructure by $8 billion to $10 billion per year since 2015.”

The current state of the technology (roadblocks and all) demands fostering an innovation ecosystem today that parallels the explosion of new services for the 5G economy. As McKinsey reports there are more than 25 billion connected IoT devices currently, which is estimated to grow to more than 75 billion by 2025 with the advent of fifth generation wireless. The study further cites, “General Electric projects that IoT will add $10 to $15 trillion to a worldwide Global Domestic Product (GDP) growth by 2030. To put that into perspective, that number is equivalent to China’s entire current economy.” Regrettably, most of the available 5G accelerators in the USA are built to showcase virtual and augmented reality instead of fostering applications for the larger opportunity of business-to-business services. According to Business Insider “IoT solutions will reach $6 trillion by 2021,” across a wide spectrum of industries, including: healthcare, manufacturing, logistics, energy, smart homes, transportation and urban development. In fact, hardware will only account for about one-third of the new revenues (and VR/AR headsets comprise considerably less).

It is challenging for publicly traded companies (like T-Mobile, Verizon & AT&T), whose stock performance is so linked to the future of next generation wireless. Clearly, market makers are overly excited by the unicorns of Oculus (acquired by Facebook for $2 billion in 2014) and Magic Leap (valued at $4.5 billion in 2016) more than IoT sensors for robotic recycling, agricultural drones, and fuel efficient rectors. However, based upon the available data, the killer app for 5G will be found in industry not digital theatrics. This focus on theatrics is illustrated in one of the few statements online by Verizon’s Christian Guirnalda, Director of its 5G Labs, boasting “We’re literally making holograms here using a dozen of different cameras in a volumetric capture studio to create near real-time images of what people and products look like in 3D.” A few miles north of Verizon 5G Labs, New York City’s hospitals are overcrowded with patients and data leading to physical and virtual latency issues. Verizon could enable New York’s hospitals with faster network speeds to treat more patients in economically challenged neighborhoods remotely. Already, 5G threatens to exasperate the digital divide in the United States by targeting affluent communities for its initial rollout. By investing in more high-speed telemedicine applications, the telecommunications giant could potentially enable less privileged patients access to better care, which validates the need for increased government spending. Guirnalda’s Lab would be better served by applying the promise of 5G to solve these real-life urban challenges from mass transit to food scarcity to access to healthcare.

The drawback with most corporate 5G incubators is their windows are opaque – forcing inventors to experiment inside, while the real laboratory is bustling outside. The United Nations estimates by 2050 seventy percent of the world’s population will be urban. While most of this growth will take place in developing countries (i.e., Africa and Asia) already 80% of global GDP is generated in cities. The greatest challenge for the 21st century will be managing the sustainable development of these populations. At last month’s UN “World Cities Day,” the diplomatic body stated that 5G “big data technologies and cloud-computing offer the potential to enhance urban operations, functions, services, designs, strategies and policies.” The UN’s statement did not fall on deaf ears, even President Trump strained to comfort his constituents last month with the confession, “I asked Tim Cook to see if he could get Apple involved in building 5G in the U.S. They have it all – Money, Technology, Vision & Cook!”

Going to CES? Join me for my panel on Retail Robotics January 8th at 10am, Las Vegas Convention Center. 

MIT researchers have invented a way to efficiently optimize the control and design of soft robots for target tasks, which has traditionally been a monumental undertaking in computation.

Soft robots have springy, flexible, stretchy bodies that can essentially move an infinite number of ways at any given moment. Computationally, this represents a highly complex “state representation,” which describes how each part of the robot is moving. State representations for soft robots can have potentially millions of dimensions, making it difficult to calculate the optimal way to make a robot complete complex tasks.

At the Conference on Neural Information Processing Systems next month, the MIT researchers will present a model that learns a compact, or “low-dimensional,” yet detailed state representation, based on the underlying physics of the robot and its environment, among other factors. This helps the model iteratively co-optimize movement control and material design parameters catered to specific tasks.

“Soft robots are infinite-dimensional creatures that bend in a billion different ways at any given moment,” says first author Andrew Spielberg, a graduate student in the Computer Science and Artificial Intelligence Laboratory (CSAIL). “But, in truth, there are natural ways soft objects are likely to bend. We find the natural states of soft robots can be described very compactly in a low-dimensional description. We optimize control and design of soft robots by learning a good description of the likely states.”

In simulations, the model enabled 2D and 3D soft robots to complete tasks — such as moving certain distances or reaching a target spot —more quickly and accurately than current state-of-the-art methods. The researchers next plan to implement the model in real soft robots.

Joining Spielberg on the paper are CSAIL graduate students Allan Zhao, Tao Du, and Yuanming Hu; Daniela Rus, director of CSAIL and the Andrew and Erna Viterbi Professor of Electrical Engineering and Computer Science; and Wojciech Matusik, an MIT associate professor in electrical engineering and computer science and head of the Computational Fabrication Group.

“Learning-in-the-loop”

Soft robotics is a relatively new field of research, but it holds promise for advanced robotics. For instance, flexible bodies could offer safer interaction with humans, better object manipulation, and more maneuverability, among other benefits.

Control of robots in simulations relies on an “observer,” a program that computes variables that see how the soft robot is moving to complete a task. In previous work, the researchers decomposed the soft robot into hand-designed clusters of simulated particles. Particles contain important information that help narrow down the robot’s possible movements. If a robot attempts to bend a certain way, for instance, actuators may resist that movement enough that it can be ignored. But, for such complex robots, manually choosing which clusters to track during simulations can be tricky.

Building off that work, the researchers designed a “learning-in-the-loop optimization” method, where all optimized parameters are learned during a single feedback loop over many simulations. And, at the same time as learning optimization — or “in the loop” — the method also learns the state representation.

The model employs a technique called a material point method (MPM), which simulates the behavior of particles of continuum materials, such as foams and liquids, surrounded by a background grid. In doing so, it captures the particles of the robot and its observable environment into pixels or 3D pixels, known as voxels, without the need of any additional computation.     

In a learning phase, this raw particle grid information is fed into a machine-learning component that learns to input an image, compress it to a low-dimensional representation, and decompress the representation back into the input image. If this “autoencoder” retains enough detail while compressing the input image, it can accurately recreate the input image from the compression.

In the researchers’ work, the autoencoder’s learned compressed representations serve as the robot’s low-dimensional state representation. In an optimization phase, that compressed representation loops back into the controller, which outputs a calculated actuation for how each particle of the robot should move in the next MPM-simulated step.

Simultaneously, the controller uses that information to adjust the optimal stiffness for each particle to achieve its desired movement. In the future, that material information can be useful for 3D-printing soft robots, where each particle spot may be printed with slightly different stiffness. “This allows for creating robot designs catered to the robot motions that will be relevant to specific tasks,” Spielberg says. “By learning these parameters together, you keep everything as synchronized as much as possible to make that design process easier.”

Faster optimization

All optimization information is, in turn, fed back into the start of the loop to train the autoencoder. Over many simulations, the controller learns the optimal movement and material design, while the autoencoder learns the increasingly more detailed state representation. “The key is we want that low-dimensional state to be very descriptive,” Spielberg says.

After the robot gets to its simulated final state over a set period of time — say, as close as possible to the target destination — it updates a “loss function.” That’s a critical component of machine learning, which tries to minimize some error. In this case, it minimizes, say, how far away the robot stopped from the target. That loss function flows back to the controller, which uses the error signal to tune all the optimized parameters to best complete the task.

If the researchers tried to directly feed all the raw particles of the simulation into the controller, without the compression step, “running and optimization time would explode,” Spielberg says. Using the compressed representation, the researchers were able to decrease the running time for each optimization iteration from several minutes down to about 10 seconds.

The researchers validated their model on simulations of various 2D and 3D biped and quadruped robots. They researchers also found that, while robots using traditional methods can take up to 30,000 simulations to optimize these parameters, robots trained on their model took only about 400 simulations.

Deploying the model into real soft robots means tackling issues with real-world noise and uncertainty that may decrease the model’s efficiency and accuracy. But, in the future, the researchers hope to design a full pipeline, from simulation to fabrication, for soft robots.

Aircraft Tooling, a Texas-based repair center for the aviation industry, was surprised to find that Universal Robots could withstand the high temperatures and harsh environment while performing metal powder and plasma spray processes. The UR “cobots” have now been in operation for three years without breakdown or service requirements.

New 2019 cobot market report from Interact Analysis reveals: • The growth rate of collaborative robots is leading the robotics industry • Logistics will surpass automotive to be the second largest end user of cobots by 2023, with electronics in first place • In the next five years, the fastest growing regions for collaborative robot shipments will be China and the USA

“We chose a combination of solutions from OnRobot, Universal Robots, and EasyRobotics because they are easy to program, and the investment will pay for itself in just nine months. It’s one of the best business decisions we’ve ever made.”

We provide motion control solutions in the form of electric motors, drives and motion control solutions for OEM Customers who are unsatisfied with having to design around inflexible off-the-shelf products.

Mobile robots, in the form of AGVs, have been on the market for nearly two decades. But a topical innovation in the sector is creating power requirements that the battery market must respond to.

Where motion is the key to a great cup of coffee, duplicating the precision and reliability of the motion of a person’s hand, wrist, and elbow require a unique robotic design.

In “Boston Ivy”, two Hachi Auto vehicles shuttle between the community’s bus stop and its food court, passing its underground parking, main garden, as well as its activity, marketing, and children’s centers along their route.

e-commerce is changing the way warehouses are constructed and operated. One way or another, warehouses must become more adept and efficient in handling multi-item instant order fulfillment. The use of automation is an essential part of the answer to this requirement.

With Exyn’s fully autonomous aerial drones, the team was able to navigate old workings and map them more accurately with drones that are able to independently navigate, visualize and record the whole space, even in the challenging environment.

The impetus was that we wanted to get more power out of the machine and wanted to pack the pallets cleaner and more dimensionally stable. The technology, the know-how and the service convinced us of Piab, after we looked at several manufacturers.

In this episode Lilly Clark interviews Marlyse Reeves, PhD student at MIT, about her work in cognitive robotics and hybrid activity-motion planning. Reeves discusses the role of robotics in space, the challenges of multi-vehicle missions, planning under uncertainty, and her work on an underwater exploration mission.

Marlyse Reeves

Marlyse is a third-year PhD student in the Computer Science and Artificial Intelligence Laboratory at MIT. She received her B.S. in Aeronautics and Astronautics from MIT in 2017. Her current research in the Model-based Embedded and Robotic Systems Group focuses on multi-vehicle online planning, incorporating complex dynamics and constraints. She is also interested in risk-aware planning, fault protection and diagnosis, and adaptive sampling. Outside of the lab, she enjoys playing soccer, dancing, and reading science fiction.

Links

• Download mp3 (14.2 MB)
• MIT Model-based Embedded and Robotic Systems Group
• Subscribe to Robots using iTunes
• Subscribe to Robots using RSS
• Support us on Patreon

Bonding is becoming more and more a new alternative to welding, screwing or riveting. The reasons are simple: safety and economy. But to be able to guarantee this, a high-precision adhesive and sealant application is necessary.

This thought experiment shows why robot localization, the task of maintaining a location position as robots move around, and obstacle avoidance, is so critical to the success of all non-stationary robots.