Engine Sustainment IntelligenceAI-enabled predictive sustainment prototype
A015 SYNTHETIC PROTOTYPE • V2
HUMAN-LED • AI-ENABLED

Sense → Contextualize → Predict → Diagnose → Plan → Source → Learn

Condition-based sustainment that accounts for telemetry, operator usage, mission load, environment, and adjacent component health.

Fleet Risk by Depot

Highest-Risk Engines

Recommended Command Actions

120-Day Health Trend

Current AI Assessment

Recent Fluid Analysis

Maintenance History

What is driving risk?

Important: operator data is pseudonymous synthetic usage context. It is not a personnel-performance score.

Baseline PM vs. AI-Informed Review

Recent Operating Context

Adjacent Component Health

Guided Diagnostic Questions

Select an engine, then use these prompts. Responses are deterministic explanations from synthetic evidence—not an LLM service.

Red Cedar Diagnostic Copilot
Choose a guided question to begin.
Synthetic decision support only. Human maintenance authority remains responsible. Operator context is not a personnel evaluation.

AI-Projected 30-Day Parts Risk

What is real, public-source-inspired, and synthetic?

Everything operational in this demo is synthetic. No Caterpillar proprietary data and no actual Army maintenance, readiness, operator, or supply records are included.

Public references used to shape the simulation

  1. NASA C-MAPSS / Prognostics data: informed run-to-failure degradation, multivariate sensor history, and Remaining Useful Life concepts.
  2. UCI AI4I 2020: informed synthetic industrial predictive-maintenance structure, operating conditions, failure mechanisms, and machine-failure labels.
  3. PHM Society engine-maintenance challenges: informed sensor-to-maintenance-event prediction and time-to-event framing.

Team-provided maintenance concept incorporated in V2

The prototype now demonstrates a four-tier, hourly-based planned-maintenance concept plus daily/seasonal inspection logic, while adding factors that fixed hourly schedules alone may not capture: operator usage, mission/duty severity, environmental exposure, and adjacent system condition.

PM disclaimer: The PM1/PM2/PM3/PM4 hour intervals in this prototype are illustrative synthetic values. They are not Caterpillar service intervals. Production deployment must use approved CAT/Army technical publications for the exact engine and application.

What we intentionally invented

Absolute sensor ranges, fluid-analysis thresholds, failure timings, pseudonymous operator profiles, driving-event counts, duty profiles, terrain severity, component-health values, PM intervals, parts identifiers, inventory, costs, technical text, depot asset assignments, and model results.

A015 alignment

CBM: telemetry + fluid analysis + contextual stress + personalized maintenance review. Technical Support: diagnostic evidence + knowledge transfer + remote-support packet. Supply Chain: predicted parts demand + inventory and lead-time risk.

Open docs/DATA_PROVENANCE.md, docs/SYNTHETIC_DATA_METHOD.md, and docs/MODEL_CARD.md for detailed documentation.