D2.1 - Variable Autonomy control for mirrored teleoperation

Brief description of functionality/utility:

A mirrored teleoperation setup for sorting waste, using two Franka Emika Panda Manipulators.

Three levels of autonomy have been established:

  1. manual control with no feedback
  2. manual control with virtual force fields for object avoidance
  3. semi-autonomous mode with programmed movements

Smooth teleoperation using wireless communications has been developed together with Toshiba Telecommunications Research Lab.

Experiments have shown a reduction in cognitive loading for operators when using the semiautonomous mode, with a small reduction in cognitive load when working with virtual force-fields. Software has been developed in a specific environment for the Panda arm, using a high-speed communication protocol developed especially for this task. The concept on the control is transferable to other remote robotic arm systems.

Currently at TRL 5, having been tested at small scale using industrial-relevant tasks and items for sorting.