Project background
Older machines without native telemetry are often the productivity hot spots in a plant. The client wanted a non-invasive way to measure machine utilization across their older equipment.
Challenge
Inferring machine state — running, idle, stopped, in setup — from indirect signals, and doing so consistently across machine types and duty cycles.
Approach & solution
We combined current clamps and vibration sensors with a classifier calibrated per machine. Utilization feeds into a dashboard showing trends, shift comparisons, and outliers. Calibration is handled through a guided tool rather than requiring specialist tuning.
Results & benefits
Previously opaque assets became transparent, and the data supported concrete decisions — sometimes to invest more in an underutilized machine, sometimes to retire one. Utilization rose on monitored assets within the first quarter.
