Trusted Remote Operation of Proximate Emergy Robots (TROOPER): DARPA Robotics Challenge

Recent robotics efforts have led to automating simple, repetitive manipulation tasks to speed up execution and lessen an operator's cognitive load, allowing them to focus on higher level objectives. However, the robot will eventually encounter something unexpected, and if this exceeds the tolerance of automated solutions there must be a way to fall back gracefully to teleoperation. To address this challenge, we present our human-guided autonomy solution in the context of the DARPA Robotics Challenge (DRC) Finals. We describe the software architecture that Team TROOPER developed and used on an Atlas humanoid robot. Our design emphasizes human-on-the-loop control where an operator simply expresses a desired high level goal for which the reasoning component assembles an appropriate chain of subtasks. We employ perception, planning, and control automation for execution of subtasks. If subtasks fail, or if changing environmental conditions invalidate the planned subtasks, the system automatically generates a new chain. The operator is also able to intervene at any stage of execution, enabling operator involvement to increase as confidence in automation decreases. We present our performance at the DRC Finals as well as lessons learned.