SlashGear

 While we are years away from the sort of sentient AI depicted in the movies, that isn’t stopping scientists from working on getting some of the mechanics down now. Of particular concern is somehow preventing robots in all their metallic goodness from hurting people.

Scientists at the University of Pisa’s Faculty of Engineering are currently focusing on fool proofing robot arms. What sounds like a straightforward operation is actually quite complicated in that it must account for malfunction, software bugs and glitches, problems with sensors and more. Even human error must be taken into consideration to completely make robot arms safe.

Special arm coverings and motor decouplers are being considered to help take the potential injury factor out of robotic arms. Something else included in this is proprioception, which is a human type of awareness of the world that surrounds the robot so that it can react to what is near it. Cool stuff, all except for the project’s name, physical human-robot interactions, or Phriends.

[via Gearlog]