Hardware-health AI roster

Continuously searches for repeated fault codes across devices that share firmware or model. Surfaces clusters before they become incidents.

Compares fault rates by firmware version. Flags regressions and recommends roll-forward / rollback windows.

Tracks z-score deviations on temperature, contactor, and heartbeat signals. Triggers Maintenance Dispatcher when threshold trips.

Drafts and updates field runbooks based on resolved incident patterns. Each new step needs human approval.

Drafts work orders for predictive-maintenance findings and routes them to the closest field tech.