Good morning. In this week’s news mix: MIT researches next level human-robot collaboration, Boston Dynamics & Otto unveil stunning warehouse project and cobot sous chef raises USD5.5m. We also marvel at robots learning to walk, meet a ‘Stranger Things’-inspired robot and much more!
Cobots & manufacturing
Could wearable muscle sensors help enable next level human-robot collaboration?
Researchers at MIT CSAIL’s Distributed Robotics Lab are exploring the feasability of humans and cobots lifting loads together by kitting humans out in wearable sensors that send signals to a cobot, which it then interprets and acts upon; mostly by mirroring the human’s motions, but also by responding to specific gestural controls…
A different form of collaboration now, with the news that Boston Dynamics (BD) and OTTO Motors have been working on a stunning integration of BD’s wheeled Handle bot with OTTO’s powerful autonomous mobile robots… (H/T The Robot Report)
Meanwhile, California-based precision screw machine shop Swiss Productions produces around two million precision parts per month with a little help from a 7-axis OB7 cobot from Productive Robotics fitted with a 2-finger adaptive gripper from Robotiq…
On Thursday, commercial kitchen automation firm Dexai Robotics announced that it had raised USD5.5m in its latest funding round. The company recently launched ‘Alfred,’ a cobot sous chef and plans to expand its engineering, sales, and product teams to both “serve new cuisines and fuel its growth in the foodservice space,” according to the press release.
Cobots in the kitchen. Via: Business Wire
- How are robotics meeting today’s packaging challenges? (Packaging Europe)
- This week’s poll: obstacles to automation (The Engineer)
- That robot next to you is helping create a safer workplace (EHS Today)
- Metal fabricator runs lean to deal with the skilled worker shortage (Tube & Pipe Journal)
- Esys Automation completes robotic glass urethane system for Ford Motor Company in record time (PRNewswire)
In a paper published this week on arxiv, researchers at Google unveiled a new set of reinforcement learning algorithms that can be used to teach a quadruped to walk in a matter of hours. Crucially, the new system also means that much less human intervention is required during these types of experiments.
The robot learning to walk on different types of surfaces. Credit: Google
MIT Technology Review reports:
“I think this work is quite exciting,” says Chelsea Finn, an assistant professor at Stanford who is also affiliated with Google but not involved with the research. “Removing the person from the process is really hard. By allowing robots to learn more autonomously, robots are closer to being able to learn in the real world that we live in, rather than in a lab.”
Wevolver released its 2020 Autonomous Vehicle Technology Report this week. The report provides a really useful overview of current technical topics in autonomous vehicle hardware and software, presented by a team of expert contributors. It’s also free.
Different levels of vehicle automation. Credit: Wevolver
From the report:
Over 22% of the engineers visiting the Wevolver platform do so to gain more knowledge on autonomous vehicle technology. Despite how much topics like market size and startup valuations have been covered globally by the media, many engineers have expressed to our team at Wevolver that comprehensive knowledge to grasp the current technical possibilities is still lacking.
Finally, Somatic’s BCR-01 autonomous cleaning bot has been in the news, following its recent commercial release. All bathrooms used for commercial purposes built in the United States after 1994 must comply with the Americans With Disabilities Act, which creates a useful degree of environmental conformity for the robots to work with.
The bot costs around USD1,000 per month on lease. Credit: Somatic
Nevertheless, BCR-01 requires a little human-assisted toilet training, as CEO Michael Levy told TechCrunch:
The static nature of most commercial restrooms means that robots only have to train on a space once. The team does the work remotely now, using a VR simulation of the bathroom to show the robot where to spray and wipe chemicals, vacuum and blow-dry. It’s an activity the team affectionately refers to as “the worst video game, ever.”
- RAS 2020 Award Recipients Announced (IEEE)
- Liquid metal that floats on water could make transformable robots (NewScientist)
- DeMoBat project developing robot-assisted dismantling for EV batteries and motors (Green Car Congress)
- Robots key part of succession plan on Tasmanian dairy (Farm Online)
- Team CoSTAR takes first place in underground robot competition (I-Connect007)
Come back next week for more of the latest robotics news! Until then, please enjoy…
Five vids for Friday
1. NASA unveiled a stunning 360-degree panorama of the Martian landscape this week. Composed of more than 1,000 images captured by NASA’s intrepid mars rover Curiosity, the almost 1.8 billion pixel panorama –the highest resolution achieved by Curiosity so far– took several months to assemble.
2. A collaboration between a clinical psychologist, robotics engineers and a robotics entrepreneur could help improve robots’ walking skills. The team created algorithms based on perceptual control theory –a model usually employed in therapeutic settings– and tested them against standard alogorithms, with very promising results. (H/T University of Manchester)
3. Researchers have developed an interface that amplifies nerve signals, enabling amputees to precisely control a robotic hand by simply thinking about their intended finger movements. (H/T Science)
4. Engineers at Boston University’s Morphable Biorobotics Lab unveiled a new soft robot this week. Dubbed “Dart” due to its resemblance to a terrifying creature kid’s pet from the Netflix series ‘Stranger Things,’ the bot is able to grip and lift objects using its unique six-fingered mouth.
5. To reduce human exposure to radiation during inspection tasks, researchers from the Electric Power Research Institute have been reviewing autonomous drone technologies. New video shows one such system operating without human intervention inside Exelon’s Peach Bottom Atomic Power Station in Pennsylvania, USA.