Trustworthy delayed teleoperation in-orbit and on the Moon
Teleoperation allows us to bring human decision-making to hard-to-reach places. Long-range teleoperation will be a key enabler for the next steps of the space industry, such as in-orbit servicing and manufacture of spacecraft, orbital debris management, and in-situ resource utilisation on the Lunar surface. The problem, still yet to be overcome, is how best to perform tasks with the significant communication delays over these distances.
When delays are present, there is an unavoidable delegation of responsibility from the operator to the robot, and for this is to occur smoothly, the operator must trust that the robot will behave as intended. This project includes three main ongoing activities:
Through conversations with teleoperation experts from other applications, we have identified a list of requirements for what would constitute a trustworthy teleoperation system. Surprisingly, technical features such force-feedback were not prioritised by operators. Instead, they preferred user-centric approaches to build trust, such as providing low risk training environments.
We are exploring these requirements experimentally to determine their impact on trustworthiness and performance. We have conducted a user study which confirmed that force-feedback in delayed systems can be counterproductive. Following this, we will investigate whether other forms of haptic feedback are more compatible with delays.
We have experimentally verified a virtual model of lunar regolith to use as a low risk training environment.
In collaboration with the European Space Agency, we have integrated the regolith model with a Model Mediated Teleoperation system to handle regolith. This could be used to provide stable haptic feedback which is robust to communication delays.
Team
Joe Louca
Kerstin Eder
John Vrublevskis
Antonia Tzemanaki
Funding
EPSRC ICASE
Thales Alenia Space, UK
