Project Lead: Matthew Daigle
Develop algorithms that predict end-of-charge and end-of-life for batteries (prognosis) based on rapid assessment of state-of-charge (SOC) and state-of-health (SOH) (diagnosis) coupled with anticipated environmental and load conditions. Perform subscale experiments on batteries to demonstrate prognostic capability.
+ Visit Battery Prognostics
Project Lead: Kai Goebel
Develop algorithms that are not based primarily on physics-based models but that instead learn remaining life from training data. Issues tackled here are the need to deal with sparse time series data, to provide a fair uncertainty estimate, and to deal with the validation problem.
+ Visit Data-Driven Prognostics
Project Lead: Chetan Kulkarni
Investigate damage propagation mechanisms for critical electrical components in select avionic equipment. Specifically, understand the impact of aging due to thermo-cycling, electric overstress, and vibration on MOSFETs, IGBTs, etc., and develop models for damage propagation.
+ Visit Electronics Prognostics
Project Lead: Matthew Daigle
Apply physics-based modeling methodologies for prognosis of control valves. Develop model-based prognostics algorithms for state and parameter estimation and end of life prediction. Perform comprehensive simulation studies.
+ Visit Valve Prognostics
Project Lead: Shankar Sankararaman
This project focuses on quantifying the different sources of uncertainty that affect prognostics and estimating their combined effect on prognostics by calculating the probability distribution of remaining useful life of different types of engineering components and systems. Since prognostics deals with the prediction of future, it may not be possible to precisely predict the future behavior of such engineering components and it is important to quantify the confidence in such predictions by estimating the uncertainty in such predictions. The estimated uncertainty can be useful for making risk-informed decisions with regard to several activities such as life extension, fault mitigation, mission re-planning, etc.
+ Visit Uncertainty Prognostics
Project Lead: Edward Balaban
This project concentrates on detecting and classifying incipient fault conditions in Electro Mechanical Actuators (EMA) at any point during their lifetime which can be used to provide an accurate picture of EMA component health to maintenance crews, enabling on-demand, selective servicing.
+ Visit Actuator Prognostics
Automated Contingency Management (ACM)
Project Lead: Abhinav Saxena
The Automated Contingency Management (ACM) technology aims at accommodating impending failure conditions in an automated fashion. In general it performs a multi-objective constrained optimization problem for resource reallocation and system reconfiguration of low-level and high-level controllers in a hierarchical manner. The ACM considered here focuses in particular how prognostic information can be integrated and processed to carry out such reconfiguration tasks more efficiently.
+ Visit Automated Contingency Management (ACM)
Diagnostic Algorithm Benchmarking
Project Lead: Scott Poll
A framework for systematic benchmarking of diagnostic technologies has been developed and implemented using the Electrical Power System testbed in the ADAPT lab. Our benchmarking approach generates realistic data sets for diagnostic benchmarking and emphasizes the use of standardized vocabularies and protocols which together enable “apples to apples” assessments of the effectiveness of different diagnostic technologies. The framework has been used in two diagnostic competitions hosted at the International Workshop on Principles of Diagnosis (DX).
+ Visit Diagnostic Algorithm Benchmarking
Diagnostic Inference using Probabilistic Computation
Project Lead: Ole J. Mengshoel
The goal of this project is to investigate probabilistic approaches to diagnostic inference. Our research includes the development of new methods and algorithms as well as development of cutting edge applications and demonstrations of importance to NASA.We investigate a real-world electrical power system (EPS), namely the Advanced Diagnostics and Prognostics Testbed, which is representative of EPSs found in aerospace vehicles, and demonstrate how probabilistic approaches to diagnostic inference offer a scalable approach with potential for real-time evaluation in aircraft and spacecraft.
+ Visit Diagnostic Inference using Probabilistic Computation
Project Lead: Sankalita Saha
Distributed prognostics is the next step in evolution of ISHM systems. They involve a coordinated health management of a system by using a distributed architecture of smart sensor devices. These devices monitor the health of individual subsystems. When any component/sub-system triggers the possibility of a failure requiring more attention, these devices work in coordination with each other to estimate the RUL (remaining useful life) and the health implications for the whole system.
+ Visit Distributed Prognostics
Project Lead: Sriram Narasimhan
Develop a model-based diagnosis engine that uses candidate generation and consistency checking to diagnose discrete faults in stochastic hybrid systems. The system uses hybrid (combined discrete and continuous) models and sensor data to deduce the evolution of the state of the system over time, including changes in state indicative of faults.
+ Visit HyDE Diagnostics
Prognostics Performance Evaluation
Project Lead: Abhinav Saxena
Develop universal performance metrics that allow objective assessment of prognostic algorithm performance.
+ Visit Prognostics Performance Evaluation
Rover Decision Making
Project Lead: Adam Sweet
A testbed that supports the development of reasoning algorithms at the component and system levels and decision-making algorithms based on system health information.
+ Visit Rover Decision Making
Water Recycling System Prognostics
Project Lead: Indranil Roychoudhury
The water recycling system cleans graywater and recycles it into clean water. The underlying technology is based on forward osmosis. The work described here is developing a physics model of both nominal and faulty system behavior of the WRS for several different fault scenarios and employs diagnostic and prognostic techniques to detect faulty behavior and estimate remaining life.
Idaho National Lab
Iowa State University
Penn State ARL
Qualtech Systems, Inc.
Scientific Monitoring, Inc.
University of Connecticut
University of Maryland
Jonny da Silva