Calendar-based
- Time
- Mileage
- Operating hours
Scheduled maintenance
PredictiveIQ Edge turns live or offloaded vehicle data into health, remaining useful life and diagnostic evidence, then carries it through VHMS: live on the vehicle for crew awareness, and at the platform for maintainer workflows.

Two nominally identical assets accumulate different degradation depending on load, environment, duty cycle and prior damage. PredictiveIQ Edge estimates current condition and remaining useful life from how each asset has actually operated, then delivers that judgment in two operating contexts: live on the vehicle for crews, and at the platform for maintainers. One product, one validated package, distinguished only by where the model runs and who needs the answer.
Calendar-based
Scheduled maintenance
Based on condition
Better informed maintenance window
Remaining useful life is tied to current condition and assumed future operation, not a universal countdown. Every estimate carries the assumptions behind it, so crews and maintainers know what the number depends on.

Crew / Operator
On vehicle compute, a VHMS-linked prognostic module consumes live sensor streams so the crew sees health state, developing anomalies and remaining useful life with stated confidence, without waiting for connectivity or post-mission analysis.

Maintainer
At the maintenance platform, that same module works from offloaded or available data to support PMCS, fault isolation and deeper diagnostic review. Condition and remaining life carry the same meaning the crew saw, so a maintenance window can be planned around evidence rather than an interval.

Developing problems often appear first as a change in the relationship between signals, operating conditions and expected behavior. Physics AI models what the system should be doing, compares it with what it is doing, and adds that context next to the alarms already in place. The same comparison runs in both profiles: crews see it on platform, maintainers work it at platform.
Operating conditions
+Sensor / telemetry data
+Engineering context
Expected system behavior
vsObserved system behavior
Residual / difference across signals and operating state
Normal
No action requiredAnomalous
Investigate furtherFailing subsystem or component
·Affected signals
·Likely failure mode
Crew / operator on platform
·Maintainer at platform
Diagnostic alert through VHMS with fault evidence and context
PredictiveIQ Edge executes compact model packages produced by PredictiveIQ Model Factory on the compute already available around the vehicle or platform.
Sensor measurements, telemetry, operating conditions and derived signals.
A Physics AI model optimized for the target hardware and runtime environment.
Health index, anomaly score, fault information, degradation estimate, remaining useful life and model confidence.
Vehicle compute, industrial PC, gateway or edge system at the site, plus hybrid exchange to enterprise systems when approved.
The same runtime carries whichever degradation the program cares about, because the physics and the health meaning travel inside the package from PredictiveIQ Model Factory.
Create and validate the executable Physics AI packages PredictiveIQ Edge runs.
Explore PredictiveIQ Model Factory →See how Model Factory and Edge fit together.
Open the solution →Health, diagnostics and RUL at the tactical edge.
Open Defense →Fleet health and diagnostics from design through sustainment.
Open Commercial →