AMP Robotics provides autonomous solutions for sorting trash in recycling plants. Their robots push, suck or grasp metals, plastics, and cardboard recyclables from a fast-moving conveyor belt. To identify objects, AMP employs computer vision and deep learning techniques. A key metric of the success of AMP Robotics’ system is how many recyclables are successfully removed from the stream of trash.
A key challenge AMP is facing is moving a manipulator empty in case a grasp fails. In this case, the conveyor keeps moving, possibly missing valuable recyclables. Robotic Materials is working with AMP to integrate a ruggedized version of our combined proximity and force sensors to detect the presence of an object in the gripper. Here, vision is excluded due to the amount of background clutter and required robustness of the system, which moves at high speed and experiences high impact forces by trash hitting sideways.
Contact us for Services or subscribe to our newsletter to stay tuned!