Actuator defect detection: use case

Context

Actuators can suffer from oil leaks, which can lead to fire, tool slippage, poor weld quality, and so on. They can block the production line. Cyclic actuator data (11 columns) were provided to detect and anticipate failures.

Need

Detect and anticipate failures using actuator time series.

Solution

DiagFit generated a prediction model from actuator time series.

Results

DiagFit detected 100% of cycles labeled as unhealthy, and detected an additional cycle not previously labeled as unhealthy. This potentially corresponded to an anticipation of failure.