Description: EPS-MAESTRO (EPS Management through intelligent, AdaptivE, autonomouS, faulT identification and diagnosis, Reconfiguration/replanning/rescheduling Optimization) substantially leverages previous NASA investments to assemble the correct set of technologies to implement all aspects of the intelligent, autonomous EPS manager. We have significant experience in all required technologies and have already integrated them into a general MAESTRO architecture designed to be easily applied to spacecraft subsystems. Montana State University (MSU) has designed, built, launched, and operated several satellites and has specifically studied in-space PV degradation. In addition to providing substantial knowledge, expertise and practical experience, MSU will also provide real satellite telemetry data and set up a laboratory hardware testbed, using spare MSU satellite hardware, for testing our EPS-MAESTRO prototype in Phase I. They also plan to field an actual EPS-MAESTRO prototype onboard one of their future satellites, in-space, during Phase II. The eventual, ultimate goal is the ability of an onboard autonomous intelligent system to manage the spacecraft EPS itself through the development of EPS-MAESTRO, which can be easily adapted to the EPSs of different spacecraft. EPS-MAESTRO must be sufficiently powerful, general, and computationally efficient and be easily adapted by developers. This will be accomplished using open standards, clearly defined open interfaces, use of Open Source software, and leveraging several previous NASA investments. Phase I research goals are to explore the spacecraft EPS management domain for small satellites and large manned spacecraft, elaborate the AI techniques useful for EPS characterization, diagnosis, replanning/rescheduling/adaptive execution/safing, prove the feasibility of these techniques through prototype development (by prototyping two applications), and develop a complete system specification for the Phase II EPS-MAESTRO system.

Benefits: The most direct targets for transition of this proposed effort are the large number of various future manned and unmanned, deep-space and near-earth spacecraft that would significantly benefit from autonomous, intelligent EPS management. By showing its ability to autonomously create high-quality responses to EPS events, EPS-MAESTRO will clearly illustrate its advantages over the status quo. Because it will be an open system that other developers could use to create intelligent EPS management systems, a large number of EPS-MAESTRO applications can be quickly developed. Since EPS-MAESTRO is specifically designed to easily interface with Diagnosis, Adaptive Execution, Planning, and Scheduling engines, such developers will have their choice. And additional interfaces can be developed over time to increase the number of such options. There is a potential to automate the majority of EPS management decision-making at NASA, even for low Earth orbit, with a corresponding savings in highly skilled manpower. Additional applications are various types of ground processing at KSC that also have EPS management needs. The planned Phase II demonstration of EPS-MAESTRO in space onboard an MSU satellite, will greatly aid its adoption.

Stottler Henke already sells Aurora to private companies. Commercial product and service sales related to Aurora have already resulted in over $10 million in revenue. EPS-MAESTRO improvements can be readily incorporated into Aurora and sold through existing sales channels, especially to the power generation industry, which we are already pursuing. And beyond NASA there are a large number of real-time diagnosis, replanning/rescheduling, and execution problems that EPS-MAESTRO could be readily adapted to, such as oil refineries, factories of all types, etc. And many of these potential EPS-MAESTRO users are already Aurora customers. Current Aurora customers tend to be aerospace manufacturers, partly due to our early conversion of the Boeing 787 Dreamliner production line to being an Aurora customer. Companies like Learjet and Bombardier quickly followed suit as well as some Boeing suppliers. Other customers tend to have high-value applications both requiring a high-quality solution and justifying the relatively high price. Examples include Massachusetts General Hospital, which is saving a huge amount of manpower scheduling residents, Honda, which is realizing large savings through reduced destruction of prototype vehicles during safety (e.g., crash) testing, and Clipper Windpower, which has greatly shortened their production time for individual, custom wind turbines. A MAESTRO-enhanced version of Aurora would presumably tend to have a similar diverse base of customers.