Organizers: Christoffer Heckman, Jay McMahon, Daniel Szafir, Torin Clark and Steve McGuire
Website: http://www.spacerobotics.xyz/
Space exploration represents a complex and multifaceted problem domain for autonomous systems. Harsh and distant environments, long time delays, high costs, and other challenges highlight the value of autonomy, but make it difficult to achieve in practice. The cost of failure is high, as robots will often be required to perform mission-/safety-critical operations, sometimes in novel environments for which little a priori knowledge or human oversight is available. At the same time, such autonomy must remain sufficiently trustworthy, accessible, and comprehensible to human stakeholders (mission scientists, engineers, astronauts, etc.) to add value in achieving mission objectives. On Earth, recent developments in perception, manipulation, and learning are rapidly improving the capabilities of terrestrial robotic systems for a wide range of applications. However, integration into the space domain is occurring at a more gradual pace, despite this context being the absolute test of autonomy. This workshop aims to highlight new developments in the field of autonomous space robots. Last year’s workshop identified several grand challenges facing adoption of autonomy in space, namely informative testing, reliability guarantees, and operational confidence. This workshop builds on last year’s successes and will bring together top researchers, engineers, scientists, and practitioners from the space exploration, space robotics, and autonomous robotics communities to identify science and technology ready for deployment.