Global Spatial Technology Solutions Inc.

Published purpose

Computer vision (CV) tasks (e.g. image classification and object detection) enable autonomous decision-making for a range of space applications and platforms, such as rovers, landers, and service vehicles, and reduce the workload and risks associated with humans in any decision-making loop. Such applications rely on high-performance Artificial Intelligence (AI) capability. Recent advancements in AI have leveraged the nearly limitless computational resources of cloud computing, and as such the resultant solutions are not inherently transferrable, or optimally designed, to be hosted onboard satellite platforms which have very constrained resources. This project will develop a reconfigurable AI capability that supports CV tasks and is optimized to be hosted as part of a portable system using hardware with a path to space-qualification. The results of this project will support a range of future scenarios including imaging support for autonomous space vehicle docking, decision support for autonomous probes or landers, autonomous robotic systems, satellite star trackers as well as remote sensing targeting decision support.