Critical Advancement in Telerobotic Servicing Vision Technology

Future space missions will rely heavily on telerobotic servicing capabilities. One of the most difficult scenarios envisioned for these missions is to provide servicing capabilities beyond low earth orbit. In addition to highly variable lighting and other environmental considerations as well as data latency issues, solutions for these missions must be low mass and low power due to system resource limitations. A key functionality for implementation of these missions is vision systems to provide situational awareness and visual feedback cues to the teleoperator, and to support remote tool selection, dexterous object manipulation and manipulator path planning to the telerobot. Recent advancement in hardware, software and algorithms associated with vision sensors, autonomous feature recognition and high performance processing are critical to mission implementation plans. Critical sensors addressed in this paper include high resolution two dimensional imagers as well as three dimensional vision systems such as flash LIDAR. Both kinds of vision systems benefit from recent technology advances that yield higher resolution images, reduce sensitivity to dynamic lighting variation and are robust to space radiation effects. Real time embedded processing capabilities leverage advances in high throughput processing technology to fuse multiple data sources for the autonomous identification of critical features to support motion and manipulation system operations regardless of communications latency between the servicing system and other critical assets. These high speed processing capabilities also provide near real time visualization for the teleoperator. This paper addresses key development methodologies, results and findings for advancement of these technologies in context of future in-space servicing systems.