NASA Logo, National Aeronautics and Space Administration

+NASA Home

+Ames Home

Innovative Capabilities Project Selected for Artemis Program Ground and Launch Systems
Intelligent Systems Division Banner

Innovative Capabilities Project Selected for Artemis Program Ground and Launch Systems

The Exploration Ground Systems (EGS) Program solicited proposals from the NASA@work community to address their strategic thrust areas and capability needs for ground and launch systems processing supporting each Artemis mission. 49 proposals were submitted and evaluated in a two-phase process. Of these, eight projects were ultimately selected for FY21 project execution, including the ARC proposal, “Capabilities for Online Anomaly Detection”.

BACKGROUND: Current EGS maintenance operations rely on preventive maintenance, where maintenance is performed on a schedule regardless of condition, and reactive maintenance, which occurs when something goes wrong. Both approaches are rather costly. These approaches are supplemented by intermittent manual condition-based maintenance, where the system is inspected and maintenance is performed on an as-needed basis.

The KSC Integrated Health Management Remote Monitoring (IHM RM) project seeks to reduce reliance on scheduled and reactive maintenance by instrumenting the monitored system, ensuring a constant stream of health data that could be inspected at any time. However, manual condition assessment would still be intermittent, and automated limit checks can miss trends and more subtle degradations. Could anyone help the IHM RM project realize its ultimate maintenance cost reduction vision?

Indeed, a group of researchers at ARC have developed an anomaly detection pipeline, well suited to automatically monitor streams of health data and identify anomalies. The first stage of the pipeline, the Inductive Monitoring System (IMS), characterizes expected regions of nominal operations from past normal operations. From this nominal baseline, IMS monitors a system during operations and quantifies how far the system has deviated from normal operations. The second stage of the pipeline, the Meta Monitoring System (MMS), models the likely profile of IMS output for nominal and off-nominal systems. Given these models, MMS monitors trends in the IMS output and classifies the system state as nominal or off-nominal. The final stage of the pipeline, Active Learning, judiciously quizzes a Subject Matter Expert (SME) to build a classifier of MMS-identified anomalies and is future work. In operation, it will distinguish between operationally-significant anomalies (or not) and provide rationales based on the SME’s previous responses.

NASA PROGRAM FUNDING: Exploration Ground Systems (EGS) Program, Human Exploration Operations Mission Directorate (HEOMD)

TEAM: Kevin Bradner, René Formoso (KSC), David Iverson, Nikunj Oza, Adwait Sahasrabhojanee, and Shawn Wolfe,


First Gov logo
NASA Logo -