MIT researchers help robots navigate uncertain environments

MIT CSAIL

MIT researchers have developed a trajectory-planning system for autonomous robots in unpredictable environments. | Source: Jose-Luis Olivares, MIT based on figure courtesy of the researchers

Researchers at MIT have developed a technique that can guide an autonomous robot through unknown environmental conditions. The technique helps a robot avoid obstacles without knowing the size, shape or location of what it could encounter. 

The research team hopes that its findings could help autonomous robots explore remote exoplanets where the robot, and the researchers who programmed it, don’t know what it will encounter on the planet. 

“Future robotic space missions need risk-aware autonomy to explore remote and extreme worlds for which only highly uncertain prior knowledge exists. In order to achieve this, trajectory-planning algorithms need to reason about uncertainties and deal with complex uncertain models and safety constraints,” co-lead author on the paper, Ashkan Jasour, said. 

MIT’s team couldn’t use typical trajectory planning methods that make assumptions about the vehicle, obstacles and environment. These methods are too simplistic for real-world settings. Instead, the team developed an algorithm that could determine the probability of observing different conditions or obstacles at different locations.

The algorithm determines the probability of these events based on a map or images the robot collects with its perception system. This approach formulates trajectory planning as a probabilistic optimization problem, a mathematical programming framework which lets the robot achieve planning objectives while avoiding obstacles. 

“Our challenge was how to reduce the size of the optimization and consider more practical constraints to make it work. Going from good theory to good application took a lot of effort,” Jasour said.

The researchers then used higher-order statistics of probability distributions of the uncertainties to convert the probabilistic optimization problem into a more straightforward deterministic optimization problem. This kind of problem could be solved efficiently with off-the-shelf solves. 

MIT’s team tested their technique with simulated navigation scenarios. In an underwater model where the algorithms needed to chart a course from an uncertain position, around obstacles and to a goal region. The system could safely reach the goal 99% of the time. Depending on how complex the environment is, the algorithm can plan a safe course in seconds or minutes. 

The next step for the team is to create more efficient processes that significantly reduces runtime. Co-authors on the paper include Jasour, former Computer Science and Artificial Intelligence Laboratory (CSAIL) research scientist who now works at NASA’s Jet Propulsion Lab, and Weiqiao Ham, a graduate student in the department of electrical engineering and computer science and member of CSAIL. Senior author on the paper was Brian Williams, a professor of aeronautics and astronautics and a member of CSAIL. 

The post MIT researchers help robots navigate uncertain environments appeared first on The Robot Report.

Efforts to deliver the first drone-based, mobile quantum network

Hacked bank and Twitter accounts, malicious power outages and attempts to tamper with medical records threaten the security of the nation's health, money, energy, society and infrastructure. Harnessing the laws of nature—namely quantum physics—a cutting-edge teleportation technology is taking cybersecurity to new, "unhackable" heights using miniscule particles of light, or "beams."

Intuition Robotics partners with NY State Office for the Aging

ElliQ

ElliQ is a robotic companion designed specifically to fit the needs of older adults. | Source: Intuition Robotics

Intuition Robotics has partnered with the New York State Office for the Aging (NYSOFA) to put ElliQ, a robot designed to help older adults gain independence, in the homes of more than 800 older adults. 

ElliQ is a robotic companion designed specifically for aging adults that suffer from loneliness or social isolation. Its technology combines psychology, behavioral sciences and advanced cognitive artificial intelligence capabilities to provide proactive care. 

“ElliQ was really initially designed to help with companionship and loneliness,” Grace Andruszkiewicz, the director of marketing at Intuition Robotics, said. “But along the way, a lot of other sort of helpful features have been built into the product. Lots of communication features to help people stay connected to their loved ones, and health and wellness, so they can achieve their goals.” 

According to Intuition, ElliQ can proactively suggest activities or start conversations. The more users interact with ElliQ, the more the robot is able to build context and inform follow-up conversations, resulting in a more natural relationship than with other robotic companions. 

“Throughout the day, she’ll sort of chime in, and this is really where one of the key differentiators is. Where a lot of other voice activated devices require the individual to prompt the device using a call word, ElliQ can start to understand when might be a good time to interact and engage with the individual,” Andruszkiewicz said. 

On average, users interact with ElliQ 20 times a day. ElliQ users meet with nurses on Intuition Robotics’ team that will help users set goals so the robot knows how to help each user. The robot can remind users to take health measurements, like their blood pressure, and medications. 

Intuition Robotics’ partnership with SilverSneakers, a fitness and wellness program geared towards seniors, means ElliQ is equipped with a library of fitness videos to help users stay active. 

The company’s partnership with NYSOFA comes just months after the robot made its commercial debut. Intuition Robotics doesn’t have a strict timeline for the program, but initially each county in the state will be able to opt into the program. To qualify for the program, users must be residents of the state of New York and speak English, as ElliQ doesn’t know any other languages yet. 

“Despite misconceptions and generalizations, older adults embrace new technology, especially when they see it is designed by older adults to meet their needs,” Greg Olsen, director of NYSOFA, said. “For those who experience some form of isolation and wish to age in place, ElliQ is a powerful complement to traditional forms of social interaction and support from professional or family caregivers.”

Intuition Robotics is continuing to develop ElliQ with feedback from users. The company is currently working on updating its caregiving facing application, to allow caregivers to more easily interact with their family members with ElliQ, according to Andruszkiewicz. 

The post Intuition Robotics partners with NY State Office for the Aging appeared first on The Robot Report.

DeepMind’s open-source version of MuJoCo available on GitHub

Shadow hand MuJoCo

The Shadow hand from Open AI was built in part using the MuJoCo physics engine. | Credit: OpenAI

DeepMind, an AI research lab and subsidiary of Alphabet, in October 2021 acquired the MuJoCo physics engine for robotics research and development. The plan was to open-source the simulator and maintain it as a free, open-source, community-driven project. According to DeepMind, the open sourcing is now complete, and the entire codebase is on GitHub.

MuJoCo, which stands for Multi-Joint Dynamics with Contact, is a physics engine that aims to facilitate R&D in robotics, biomechanics, graphics and animation, and other areas where fast and accurate simulation is needed. MuJoCo can be used to implement model-based computations such as control synthesis, state estimation, system identification, mechanism design, data analysis through inverse dynamics, and parallel sampling for machine learning applications. It can also be used as a more traditional simulator, including for gaming and interactive virtual environments.

DeepMind said the following are some of the features that make MuJoCo attractive for collaboration:

  • Full-featured simulator that can model complex mechanisms
  • Readable, performant, portable code
  • Easily extensible codebase
  • Detailed documentation: both user-facing and code comments
  • We hope that colleagues across academia and the OSS community benefit from this platform and contribute to the codebase, improving research for everyone.

Here is more from DeepMind:

“As a C library with no dynamic memory allocation, MuJoCo is very fast. Unfortunately, raw physics speed has historically been hindered by Python wrappers, which made batched, multi-threaded operations non-performant due to the presence of the Global Interpreter Lock (GIL) and non-compiled code. In our roadmap below, we address this issue going forward.

“For now, we’d like to share some benchmarking results for two common models. The results were obtained on a standard AMD Ryzen 9 5950X machine, running Windows 10.”

As for the near-term roadmap, DeepMind said it will unlock MuJoCo’s speed potential with batched, multi-threaded simulation, support larger scenes with improvements to internal memory management and introduce a new incremental compiler with better model composability. DeepMind also said it will build out support for better rendering via Unity integration and add native support for physics derivatives, both analytical and finite-differenced.

Before the acquisition, DeepMind used MuJoCo as a simulation platform for various projects, mostly via its dm_control Python stack. It highlighted a few robotics examples, which you can watch via the playlist below.

The post DeepMind’s open-source version of MuJoCo available on GitHub appeared first on The Robot Report.

Festo introduces pneumatic cobot arm

festo cobot

Festo’s pneumatic collaborative robot will be available in 2023. | Source: Festo

Festo announced its pneumatic collaborative robot (cobot) arm at the Festo TechTalk 2022 earlier today. The company plans to make the cobot commercially available in 2023. 

The cobot uses six pneumatic direct drives, instead of the typical electric motors and mechanical transmission, to move. Each of the six drives consists of a circular chamber with a moveable partition. Differences in air pressure on either side of the partition wall in the chamber cause the it to shift, which then moves the joint. 

Festo’s pneumatic cobot has many advantages over typical cobots. The high energy density of compressed air means that the cobot can be moved precisely even without complex force-torque sensors.

The arm is equipped with precise pressure regulators in the joints, meaning the robot knows when it’s touched by a human and can respond accordingly, according to Festo’s Head of Robotics Christian Tarragona. 

The cobot has a 670 mm reach and a 3 kg payload. It weight around 17 kg, due to its use of die-cast aluminum. Because all of its relevant systems are integrated into the foot section of the robot, it doesn’t require an additional control cabinet. 

Festo’s cobot can be programmed similarly to many other cobots on the market. The company’s robotic suite software offers the option of programming the arm with an operating device and predefined skills. The robot can also be programmed with hand-guiding. Getting the cobot ready for pick-and-place tasks can take less than an hour, according to the company. 

While Festo did not reveal any exact price information during the presentation, Festo CEO Frank Mezler said that the company plans to keep the price lower than an electrically driven cobot. The product comes with the cobot itself, a hand modules, the Robotic Suite and software for intuitive commissioning and programming.

Festo was founded in 1925 in Esslingen, Germany, and has been family-owned for three generations. The company offers around 33,000 different products ranging from automation technology to learning systems, training and consulting. 

The post Festo introduces pneumatic cobot arm appeared first on The Robot Report.

Advanced cable management lets robots depaint airplanes

depainting a plane

Titan’s robots use lasers to strip the paint from airplanes. | Source: Titan Robotics

The aviation industry is no stranger to automation. The first “gyroscopic automatic pilot” dates back to the earliest days of powered flight. In this century, automation has all but completely transformed the nature of aircraft manufacturing. 

In the maintenance hangar, essential safety procedures traditionally relied on the “keen, experienced eyes” of human maintenance personnel. Now, however, automation is gaining ground there as well.  

One specific aspect of aircraft maintenance has always posed challenges for human health and safety: the depainting of airplanes. According to Boeing, planes need to have their coatings of paint removed about every five years. Aircraft are painted to prevent corrosion as well as for aesthetic purposes, but periodically, it is important to inspect the plane’s bare surface.  

However, aerospace paint typically contains hexavalent chromium. This compound helps to create certain colors and aids in reducing corrosion, but it is also associated with serious health risks. According to the Occupational Health and Safety Administration (OSHA), exposure to hexavalent chromium “can cause severe health effects to workers, including lung cancer.”

This threat has made depainting both dangerous and costly.  In the past, every time human workers blasted the paint off an F-16 with hoses spraying plastic beads, the process created no less than a full ton of hazardous waste.  

Robotics to the Rescue

With the advent of a new laser-based depainting system from Titan Robotics, it is now possible to execute this essential work with far less risk to human workers.

Based in Pittsburgh, PA, Titan is a spinout from the National Robotics Engineering Center (NREC) at Carnegie Mellon University (CMU).  Founded by CMU faculty in 2014, its new approach to aircraft surface processing reflects the success of NREC research funded by the U.S. Air Force.  

Laser ablation removes paint from airplanes by “exciting” and vaporizing the molecules.  By comparison to blasting or chemical removal methods, this process reduces the production of hazardous waste by more than 90%. Furthermore, Titan’s strong focus on software development has allowed it to develop a system in which robots operate the lasers, keeping humans out of harm’s way. 

“We add more complexity and software control on to typical development of automated solutions,’’ noted Alex Klinger, a program engineer at Titan. “With our software, the robots figure out where the plane is, how much paint is on the plane, what type of paint and how to burn the paint off.”  

Precision, Mobility, and Safety

To create a 3D map of the surface to be depainted, Titan’s robots use LiDAR (“light detection and ranging”), a sensor technology widely used in autonomous vehicles.  In two-robot full aircraft systems, the robots are mounted on mobile bases which allow them to drive themselves around the aircraft.  For work on off-aircraft components, rail-based systems allow the robots to reach very long parts of the aircraft.  Titan can also design systems in which the robots are fixed in position.  

While the robots use continuous wave lasers to remove the paint, humans monitor the process from a control room, where they are protected from exposure. 

“It’s really about safety,’’ Klinger said. “They’re really just being the supervisor of the system, making sure it’s doing what it’s supposed to be doing.”

Another critical requirement is the need to maintain the structural integrity of the plane.  Damage to the surface of a $130 million fighter jet would be no small matter from a cost point of view.  But there is the ultimate priority of flight safety to consider as well.  

“There is a tremendous amount of rigor on our side for our control systems and our robotics in making sure that we’re not hitting the aircraft and making sure that our laser processes are not damaging the aircraft,’’ Klinger said. 

Connection with Confidence

To guide and protect the power supply and other essential connections such as laser fibers, Titan’s design required specially engineered cable carriers.  Manufactured by igus, the triflex system can hold up to 16 cables at a time. One important advantage of triflex is its capacity to handle the weight and protect the cables.

“The weight and the volume are substantial,’’ Klinger said. “I think it’s 12-to-16 distinct cables and they can be held very nicely.” 

The other advantage of triflex is the range of motion. “It can hold the cables, but it also has a three-degree freedom of movement,’’ Klinger noted. “A lot of cable chains move in a two-dimensional space, not a three-dimensional space. When you get to complex motions, where a robot is working in three-dimensional space, we need a flexible link that can contain the cables while the robot operates with six degrees of freedom. Triflex can do that.”

Multi-axis cable carriers such as triflex are used in welding, packaging, material handling and automotive applications. A built-in torsion stop reduces mechanical stress on cables, and a defined bend radius ensures the bend radius of cables is not violated. 

Klinger noted, “For us, it’s all about having confidence in those cables. We know the product will contain them, will limit how they bend, and it will hold them exactly where we want them to for the entire time.”   

New Horizons

The sensing and control methodologies of Titan’s systems have many other potential applications, with the robots deploying different tools.  “The robotics are the same, the software is the same,’’ Klinger noted. The cable carriers provided by igus can also solve the same problem in other systems. Regardless of the specific context, it is all about delivering utilities to the point of use. It’s important that we know where those cables are, know that they are protected and not worry about them getting snagged.”

For now, however, Titan Robotics has made a meaningful contribution to the safety of aircraft maintenance personnel.  As summed up by Klinger, “you had the human factors of a person working in a terrible environment for a long period of time.’’  Now, with human oversight, robots are helping keep planes safe to fly with far less risk to crews on the ground.  

Katherine Bonamo and Thomas Renner write on engineering, construction and other trade industry topics for publications in the United States and Canada.

The post Advanced cable management lets robots depaint airplanes appeared first on The Robot Report.

AB Dynamics’ robots at use crash testing NASCAR cars

NASCAR crash test

NASCAR performed a crash test with cars equipped with AB Dynamics’ technology. | Source: AB Dynamics

AB Dynamics and NASCAR have partnered to perform crash-test of NASCAR’s Next Gen race car using AB Dynamics’ robots. The car was equipped with steering, shifting and pedal (throttle, brake and clutch) robots, as well as sensors and a crash test dummy. 

The plan for the test crash was to send the race car into a Steel and Foam Energy Reduction (SAFER) barrier at 130 mph. The car needed to hit the barrier precisely at an angle of 24 degrees.

“The challenge was trying to get this extremely complex machine to do a very precise test without a human driver piloting the car,” Craig Hoyt, AB Dynamics Business Development Manager, said. “AB Dynamics robots allowed NASCAR to use a fully running race car and conduct the test at a real race track at real race speeds. There is no better data than replicating crash tests in a real environment and our robots enable us to do that accurately and repeatedly.”

NASCAR used AB Dynamics’ SR60 for steering, CBAR600 for pedals and its Gearshift Robot to drive the car. The company’s path following software ensured the robot was able to steer the car into the SAFER barrier at exactly 130.015 mph and within one degree of the determined angle. The vehicle hit the barrier within 2 cm of the desired impact point. 

“This is a truly innovative way to test the safety of vehicles in motorsport. The data we obtained from the test was extremely important and was not possible to get from any crash test facilities at the time,” John Patalak, the managing director of safety engineering at NASCAR, said. “The test provided valuable information for correlation with our computer crash simulations and confirmed that the predicted vehicle impact performance from the simulation was duplicated in this real-world crash test.”

AB Dynamics was founded in 1982 as a vehicle engineering consultancy. The company offers automotive test systems for a variety of applications, including highly-efficient durability testing to precision control for new areas of technology development. 

The post AB Dynamics’ robots at use crash testing NASCAR cars appeared first on The Robot Report.

Tiny drone based on maple seed pod doubles flight time

A trio of researchers at City University of Hong Kong has developed a tiny drone based on the maple seed pod. In their paper published in the journal Science Robotics, Songnan Bai, Qingning He and Pakpong Chirarattananon, describe how they used the maple seed pod as an inspiration for increasing flight time in under 100-gram drones.