From Health Care to Space Robotics

Reading time ( words)

Each year, US space agency NASA asks startups with ideas that could improve space exploration to apply for their iTech initiative. Among the ten finalists in the recent competition (October 7 and 8) is MyelinS. This startup at the Technical University of Munich (TUM) began developing software for intelligent prostheses and ended up inventing new ways for astronauts and robots to interact in outer space.


Robots are an indispensable part of space exploration. While machines like the Mars rover "Curiosity" resemble very basic model cars, there are plans to use humanoid robots during space missions. Remotely controlled by humans from inside spaceships or space stations these robonauts could be assigned dangerous tasks. Software designed by startup MyelinS could be a part of these future missions.

“Our software provides three basic features,“ says Zied Tayeb, a doctoral candidate at TUM's Institute for Cognitive Systems and, together with Samaher Garbaya, one of the minds behind MyelinS. “The first feature is navigation: We use machine learning algorithms to enable a robot to create an internal map of its surroundings and avoid obstacles.” This is important even in remotely-controlled robots, since it enables the pilot to concentrate on the actual task and let the robot deal with the navigation—a concept called shared control. “The second feature is tactile feedback,” continues Tayeb. “Our software can learn to translate input from the robots tactile sensors into output generated by vibration motors. Imagine, for example, a robot handling rock samples: The scientist in control would be able to feel the structure of these samples.”

Curiosity as a feature

The third major feature of the MyelinS software is curiosity. Many great discoveries start with a sentence like “Wait … that looks interesting“. “We use machine learning to teach our algorithm which kinds of things 'look interesting' and should be highlighted,” says Zied Tayeb. “Depending on which human expert trained the machine’s curiosity, this could be strange rocks, bent machine parts or three-eyed fish.” The MyelinS software is platform-independent and thus can be used as an interface for many different kinds of robots.

When Zied Tayeb and Samaher Garbaya first thought about founding a company, they went in a very different direction. Tayeb had created “Gumpy,” an open source software package for brain-computer interfaces. “Gumpy” included the basic machine learning algorithms for the tactile feedback. “Our original project was to create software for prostheses that let amputees feel their limbs,” says Tayeb. “We created an advanced version of Gumpy with better algorithms and more features.” Together with two other former TUM students, Tayeb and Garbaya validated the functionality of their software in cooperation with people with missing limbs.

“About a year ago, I was talking to an old mentor. When he heard about what we were doing, he pointed out that space agencies might be very interested in brain-computer interfaces for their robots,” Tayeb narrates. “It wasn't easy to let the prosthetics project go , but the prospect of being a part of space exploration was an opportunity we didn't want to miss.” While the two fields may seem very different at first, the requirements for human-brain interfaces are in fact similar, he explains.

Support from TUM

The MyelinS team was supported by the TUM Gründungsberatung.  They also received guidance and mentorship from UnternehmerTUM,  the Center for Innovation and Business Creation at TUM—among others, they took part in the Xplore Pre-Incubation Bootcamp. „Gordon Cheng, Professor of Cognitive Systems, too, was also very supportive“, Tayeb adds. In the months to come, he and his colleagues are planning to formally found their company.



Suggested Items

DARPA Funding Brings Machine Learning to BAE Systems’ Signals Intelligence Capabilities

07/08/2019 | BAE Systems
The solution provides a reconfigurable hardware platform for developers to make sense of radio frequency signals in increasingly crowded electromagnetic spectrum environments.

Copyright © 2020 I-Connect007. All rights reserved.