Interactive/Reactive Robot Mission Scheduling For Human-in-the-Loop Mission Control

Abstract

One inherent problematic in the development and operation of complex robotic systems is the considerable skill barrier level associated with them. This is particularly problematic in the context of human-robot coop- eration, especially with regards to non-roboticist human experts from other fields. In order to mitigate some frustrations associated with mission generation and modification involved for remotely operated scientific exploration missions, we developed and tested an interactive and reactive scheduling program capable of enforcing system and precedence constraints while providing human-understandable feedback as part of an iterative human-in-the-loop refinement process. An early failure warning during mission execution to enable preventive recovery behavior was also provided. As part of this development, we formulated a “good enough” search heuristic and decision process for clear result presentation to the user. The pro- gram demonstrated reasonably good results with simulated test missions and is being developed further for deployment on a real robotic system.