AI Based Self-Correcting, Self-Reporting Edge Sensors

This Phase I SBIR project will establish the feasibility of a new class of super-enhanced edge sensors for segmented mirror telescopes. These sensors may be used to deploiy, align, and phase match the primary mirror segments of space based instruments such as NGST. They will be suitable for operational environments ranging from moderately hot (=373¿K) to cryogenic (well below 30 ¿K). Many innovations will be implemented in this new technology. For example, fuzzy logic will be used to provide health and status monitoring and equip each sensor with a self-reporting capability. Artificial neural networks will be employed to provide self-correcting and self-tuning capability. In addition, new error compensation methods will be devised for super accuracy, and multi-mode measurements of both phasing errors and gap separation between neighboring segments. This research is considered critical to both NGST and future NASA missions requiring large segmented primary mirrors. Phase I will entail both experimental testing and computer simulation and modeling. In Phase II the results of Phase I will be reduced to practice and at least two standard model edge sensors will be developed, fully characterized, and documented.