AI/ML for Anomaly Detection and Health Monitoring for the NSN

Overview

Development stage

In development or acquisitionSource: a) Pre-deployment – The use case is in a development or acquisition status.

High-impact designation

Not high-impactSource: c) Not high-impact

Impact justification

Not reported

Start date

Not reported

Withheld from public reporting

NoSource: a) No

Mission

Topic area

Science

Operational functions

Summarization Derived by keyword rules; see methodology

Problem the AI is intended to solve

The project develops ML/AI methods/tools to enhance the operation and sustainment of GSFC Space Network (ACCESS-managed) assets. We analyze telemetry from TDRS and STPSat-6 to assess state of health, detect anomalies, and predict remaining useful life (RUL) of spacecraft components. The ML/AI Algorithms are implemented in Python and MATLAB and are reported and validated with engineering teams during Sustaining ML meeting. Validated models are being integrated into a GUI to provide an intuitive platform for engineers to analyze spacecraft telemetry and support real-time monitoring. Documentation and results are summarized and posted to SharePoint.

Expected benefits

Automated telemetry anomaly detection and forecasting reduce manual review and improve Space Network availability. In internal tests, the TDRS-8/10 Bus Voltage Limiter (BVL) detector achieved ~98% true-positive rate with very few false alarms and processed ~10 years of telemetry in ~5 minutes. Early, prioritized alerts (hours–days sooner) have surfaced incipient battery issues—e.g., two early-warning signals of failing cells—and the battery tool identified diverging cells on TDRS-9/10. Remaining-useful-life (RUL) and short-term forecasts enable condition-based maintenance and proactive scheduling, reducing unplanned downtime and associated costs.

System outputs

Anomaly scores/flags from telemetry (After thresholded), Predicted Telemetry (SA current, Battery Voltage), SOH index for selected subsystems, remaining-useful-life (RUL) predictions with confidence bands, Summary of the ML/AI results for review and action by engineers.

Technology

AI classification

Classical / predictive MLSource: Classical/Predictive Machine Learning: Models trained on data to make predictions or classifications based on identified patterns or relatio…

System name(s)

Not reported

Custom-developed code

Not reported

Public source code

Not reported

Sourcing

How it was built

Not reported

Vendor (as reported)

Not reported

Vendors (standardized)

None on the standard list

A vendor is the supplier named by the agency. It does not identify the underlying model or AI technology.

Data and privacy

Involves PII

Not reported

Privacy Impact Assessment

Not reported

Authorization to Operate

Not reported

Demographic features

Not reported

Training and evaluation data

Not reported

Federal Data Catalog

Not reported

Governance

0 of 8 minimum-practice questions answered. A blank answer means the agency reported nothing; it does not mean the practice is absent.

Potential impacts and how they were identified

Not reported