NASA Logo, National Aeronautics and Space Administration

Overview

The Prognostics Center of Excellence (PCoE) at Ames Research Center provides an umbrella for prognostic technology development, specifically addressing prognostic technology gaps within the application areas of aeronautics and space exploration. The PCoE is currently investigating damage propagation mechanisms on select safety-critical actuators for transport-class aircraft, damage mechanisms on aircraft wiring insulation, and damage propagation mechanisms for critical electrical and electronic components in avionic equipment. We are also in the process of extending a testbed that will allow the comparative analysis of different prognostic algorithms. In addition, data collected from aging processes will be made available to the research community (see link to data repository below).

smart plane

Next-generation aircraft such as this morphing wing concept will experience new and unknown faults and failure modes, and will benefit from integrated health management.

The common thread among the various avenues of prognostic technology development is the investigation of physics-of-failure at the component level. Modeling damage initiation and propagation at this level is a key element in describing component health. Just as important is the investment of resources into algorithm development to provide the estimates for remaining component life and for uncertainty management.

Some of the challenges that we are interested in tackling include:

  • Uncertainty management: How can the information from multiple uncertainty sources be properly captured and processed?
  • Autonomic control reconfiguration: How can local prognostic information be translated into changes at the controller level such that controller objectives are satisfied in the long term?
  • Integration: How should information from different, interacting subsystems be combined and processed?
  • Validation and verification of prognostics: How can the proper operation of prognostic algorithms be validated, especially on new systems?
  • Post-prognostic reasoning: How can the information from a prognostic reasoner be turned into an action, also factoring in other considerations such as logistics information, mission information, and fleet management?

To that end, we will employ tools from engineering, statistics, and machine learning. Specifically, we draw upon expertise in:

  • Electronics and mechanical systems modeling
  • Risk assessment and failure analysis
  • Statistics
  • Machine learning and soft computing
  • Classification
  • Optimization

Prognostic Data Repository
One of the common bottlenecks in prognostic algorithm development is the availability of data that allows the comparison and benchmarking of algorithm performance. This data repository is geared towards easing that bottleneck by making available prognostic data sets to the research community.
+ Visit Prognostic Data Repository

Systems Health, Analytics, Resilience, and Physics-modeling (SHARP) Lab
The PCoE makes use of laboratory facilities designed to test, measure, evaluate, and mature diagnostic and prognostic health management technologies. A number of hardware-in-the-loop testbeds and associated measurement equipment allow for controlled, repeatable, safe injection of faults.
+ Visit SHARP

Members

Center Coordinator
Goebel, Kai, Ph.D.

FYI:
Industry-Day Presentations

Open Source Products Prognostic Models Prognostic Algorithms

Conferences
Annual Conference of the PHM Society
Denver, CO
Oct 2 - Oct 6, 2016

Center Members
Bajwa, Anupa
Balaban, Edward
Daigle, Matthew
Frost, Susan
Gorospe, George
Iverson, David
Kulkarni, Chetan
Martin, Rodney
McIntosh, Dawn
Nishikawa, David
Oza, Nikunj
Patterson-Hine, Ann
Poll, Scott
Roychoudhury, Indranil
Sankararaman, Shankar
Sweet, Adam
Teubert, Christopher
Timucin, Dogan
Wheeler, Kevin

Alumni


Collaborations & Associations

First Gov logo
NASA Logo - nasa.gov