Developing ePartners for human-robot teams in space based on ontologies and formal abstraction hierarchies

Telling autonomous systems what to do

“…Humans may think differently than machines and, as a result of diversity, the perspective taking and solution space for off-nominal (wicked) problems can be larger when machines and humans share their perceptions and cognitions. For example, a goal and work model has been developed and instantiated in ePartners for future space missions to Mars which have to deal with potentially non-nominal situations, such as a malfunctioning rover [7]. Due to the Mars rover's malfunctioning, the original mission goal of an extravehicular activity may not be attained (e.g., setting up a specific cache at a remote location).…”

“…The ePartner supports the exploration of alternative mission goals from which the team can select one. Evaluations with the Eurobot rover and with a virtual reality environment showed that the model-based ePartner support can enhance the mission team's resilience substantially [7].…”

“…In general, ePartners, as described above, are being developed to advance the effectiveness, efficiency and resilience of work processes in a responsible way, for example, for the prevention and management of lifestyle related diseases [31], robot-assisted disaster management [28], train traffic control [18] and manned space missions [7]. In these domains, generic computational models of the human's social, cognitive, emotional and physical processes are being developed, so that the ePartner can be aware of and reason about these processes.…”