Robots can manipulate fabrics better with tactile sensors and machine-learning

Roboticists have spent the last few years trying to improve robots’ interaction with objects in real-world environments. While some of these efforts have yielded promising outcomes, most robotic systems are still behind humans in terms of manipulation.

Fabrics have proven to be among the most difficult objects for robots to interact with. Fabrics and cloth can be folded, stretched and moved in different directions, which can lead to complex material dynamics.

Researchers from Carnegie Mellon University’s Robotics Institute recently developed a new computing technique that would allow robots better handle and understand fabrics. This technique is described in a paper that will be presented at the International Conference on Intelligent Robots and Systems and has been pre-published on ArXiv. It relies on a tactile sensing device and a machine-learning algorithm known as a \”classifier\”.

Source:
https://techxplore.com/news/2022-08-tactile-sensors-machine-robots-fabrics.html