Comparing NH3 values and catalytic converter efficiency at different metering rates enables sensor-free exhaust monitoring.
Machine-learning analysis of crank time, battery voltage, and ambient temperature flags degradation early for proactive engine service.
A normalized turbine-power model estimates exhaust pre-turbine pressure without direct sensor exposure, supporting engine control and failure backup.
A controller uses combustion signals to identify abnormal cycles and adjust fueling for selected cylinders before further engine damage occurs.