Toward Intuitive Robot-to-Human Error Reporting to Enhance User Awareness in Space (Tele)Operation

摘要

Employing a team of robots for space exploration offers benefits such as increased spatial coverage and optimized work distribution. However, limited bandwidth, communication delays and environmental uncertainties require the use of an intuitive multi-modal user interface (UI) to control these robotic assets. This interface enables the crew to command the robots with various modalities, including joysticks and force-reflection input devices for direct teleoperation, and an intuitive graphical user interface (GUI) for task-level command supervision. How-ever, during task executions, errors may occur and robots may fail to carry out the commanded tasks. In the absence of careful consideration for conveying error information to users, the GUI may display a message notifying the occurrence of an execution failure, providing little to no context leaving a non-expert user with unclear and uninformative details. This lack of sufficient information makes the management of the robotic assets less certain, which causes an increase of mental workload on the crew. To reduce the astronauts's cognitive load and ultimately enhance their situational awareness, it is crucial for robots to communicate the details of the encountered errors through text and visual infographs. In this paper, we propose a novel robot agnostic framework based on the OpenUSD standards to support error-related knowledge exchange from robot to astronaut. This concept was tested in the German Aerospace Center (DLR) - European Space Agency (ESA) space technology demonstration mission, Surface Avatar. Results from the International Space Station (ISS)-to-ground telerobotic experiments in July 2024 validated our approach.