ROM3SA: Improved Situational Awareness for Remotely Operating Versatile Semi-Autonomous Robot Systems

The next generation of NASA#39;s robotic assets capable of performing tasks on the Lunar surface, in intermittently crewed spacecraft such as Gateway, or longer term deep space habitats will need to work effectively with remote human team members. nbsp;While fully autonomous systems are unlikely to be available soon, NASA#39;s robots will need to operate under the supervision and guidance of their human counterparts. To make informed run-time decisions regarding how muchnbsp;autonomy a robot should be permitted, even in the presence of non-trivial communication delays, or to assess how well a robot is performing its tasks once commanded, it is imperative that human operators have sufficient feedback to orient them about the robot#39;s state and intentions, and what responsibilities the robot can be trusted to accomplish.To address this challenge, TRACLabs, with Research Partner University of California Davis, will investigate data summarization tools to support shared autonomy and supervisory control when performing extended mobile manipulation tasks such as station or habitat caretaking and operations amp; maintenance, and surface activities such as ISRU production and regolith excavation. These tools will be used to help the human operator effectively assess robots status, informing them of whether they need to help nbsp;robot systems out as necessary. We will call the proposed effort the ROM3SA (Robot Operation Metrics for Mobile Manipulation and Shared Autonomy) system. This work will focus on the design and implementation of suitable metrics that can be evaluated at run-time to summarize robot status to orient remote human operators and provide improved situational awareness to allow them to make better decisions. nbsp;