Edge Accelerated AI for Robotic Teleoperation

This paper presents a body of work on the development of edge accelerated AI hardware and software components related to robotic teleoperations. In particular, the paper proposes a system based architecture incorporating teleoperations communications layers in the form of a local to remote edge based stack, along with a Time of Flight (ToF) sensor layer, delivering user controls via real time hand and gesture signals for robotic teleoperations. Typical use cases include, remote skills delivery, precision robot manipulation, factory of the future, dangerous work environments and many other Industry 4.0/5.0 scenarios.