Designing an Autonomous Manipulator Controller for Zero-G Repair Tasks Year Two

Currently, when man-made objects in orbit around the Earth need to be disassembled or repaired, a humancontrolled robotic manipulator is used. The object in question is often not designed to be modified in this manner, and only 1 in 4 human operators at MDA (the partner organization) are able to successfully complete these notoriously challenging tasks. The goal of the proposed project is to design a computer program to learn from the successful human operators with the end result being a novel artificial intelligence that can perform these repairs autonomously, i.e. without human input or oversight. This is a project that MDA has been considering for several years, and if successful, will reduce the costs associated with such operations while increasing success rates and throughput, thus providing a distinct competitive edge.

Faculty Supervisor:

Gabriele D'Eleuterio

Student:

Paul Grouchy

Partner:

MacDonald Dettwiler and Associates Ltd.

Discipline:

Aerospace studies

Sector:

Aerospace and defense

University:

University of Toronto