Embedded and Distributed Machine Learning for Prognostics Monitoring
Project Description
Scheduled maintenance is inefficient and costly with no ability to take into account actual hardware degradation. The goal of this project is to develop a generic cost-effective embodiment that is relatively independent of the type of physical equipment being monitored by employing machine learning for prognostics monitoring. Prototype units will be developed with embedded novel machine learning algorithms for cryogenic equipment in the engine test complex as pilot demonstration systems. Energy harvesting technologies will also be integrated to further demonstrate low powered energy harvesting health monitoring capabilities.
More »Anticipated Benefits
Beyond the initially intended use in propulsion test facilities, the machine learning for diagnostics and prognostics has real applicability to support autonomy throughout NASA facilities and for missions employing autonomy in flight computing and robotic and system autonomy to handle failures and readapting when independent from Earth. This capability additionally has potential for missions that need to reduce the necessity for human intervention in maintaining equipment such as with Deep Space Missions.
More »Primary U.S. Work Locations and Key Partners
| Organizations Performing Work | Role | Type | Location |
|---|---|---|---|
Stennis Space Center
(SSC)
|
Lead Organization | NASA Center | Stennis Space Center, Mississippi |
| Co-Funding Partners | Type | Location |
|---|---|---|
| Rocket Propulsion Testing | NASA Other |
Primary U.S. Work Locations
-
Mississippi
