An American university has created a robotic arm which sorts recycling waste simply by squeezing it.
Graduate Lillian Chin and colleagues at the Computer Science and Artificial Intelligence Lab, at the Massachusetts Institute of Technology (MIT), have developed the robot which they call RoCycle.
It sorts between metal, plastic and paper depending on its size and stiffness, using two pincers fitted with pressure sensors.
Tests at the Massachusetts university showed it correctly identify 27 objects with an 85% accuracy rate.
Traditional methods of sorting recycling include magnets, moving paper with air flow and the laborious task of identifying pieces by hand.
Visual recognition machines use artificial intelligence to identify objects.
The ZenRobotics 'Fast Picker' can sort 4,000 objects an hour.
MIT says their sensory method is much slower but also more accurate.
It is working towards an upgrade that uses visual checks first, with a sensory check for more difficult items. MIT says its RoCycle could work in apartment blocks and on university campuses.