A situated reasoning architecture for space-based repair and replace tasks

Abstract An area of increasing interest within AI and Robotics is the integration of techniques from both fields to the problem of controlling autonomous systems. Space-based systems, such as NASA's EVA Retriever, provide complex, realistic domains for this integration research. Space is a dynamic environment, where information is imperfect, and unexpected events are commonplace. As such, space-based robots need low level control for collision detection and avoidance, short-term load management, fine-grained motion, and other physical tasks. In addition, higher level control is required to focus strategic decision making as missions are assigned and carried out. Throughout the system, reasoning and control must be responsive to ongoing change taking place in the environment. This paper reports on current MITRE research aimed at bridging the gap between high level AI planning techniques and task-level robot programming for telerobotic systems. Our approach is based on incorporating situated reasoning into AI and Robotics systems in order to coordinate a robot's activity within its environment. Thus, the focus of this research is on controlling a robot embedded in an environment, as opposed to the generation and execution of lengthy robot plans. We present an integrated system under development in a “component maintenance” domain geared towards repair and replacement of Orbital Replacement Units (ORUs) designed for use aboard NASA's Space Station Freedom . The domain consists of a component-cell containing ORU components and a robot (manipulator and vision system) replacing worn and/or failed components based on the collection of components available at a given time. High level control reasons in “component space” in order to maximize the number operational component-cells over time, while the task-level controls sensors and effectors, detects collisions, and carries out pick and place tasks in “physical space.” Situated reasoning is used throughout the system to cope with, for example, nondeterministic component failures, the uncertain effects of task-level actions, and the actions of external agents operating in the domain.