At a glance
- Client
- An automotive manufacturerName withheld for confidentiality
- Industry
- Manufacturing
- Service
- IoT Digital Engineering
- Scope
- Sensor retrofit, IoT platform and predictive maintenance models
The challenge
Unexpected breakdowns of presses, robots and conveyors stopped production lines several times a month. Maintenance was either reactive or done on fixed schedules, replacing parts that still had life in them.
Every hour of unplanned downtime meant missed delivery slots to customers.
Our approach
Instrument critical machines
Vibration, temperature and current sensors were added to machines with the highest downtime cost.
Build the data platform
Sensor data streams into an IoT platform alongside maintenance history.
Predict failures
Models learn each machine's normal behaviour and flag early signs of wear.
Act on predictions
Alerts create maintenance work orders scheduled into planned stops.
The solution
Maintenance teams now see which machines are likely to fail and when, and fix them during planned downtime. Parts are replaced based on condition, not the calendar.
Plant managers track machine health across the whole site from one dashboard.
Results
- Thousands of machines monitored continuously
- Early warnings of wear and failure
- Automatic work orders in the maintenance system
- Condition-based instead of calendar-based maintenance
Technology used
- Industrial IoT sensors
- MQTT
- AWS IoT
- Time-series database
- Anomaly detection models
Facing a similar challenge?
Tell us what you are working on. We will walk you through how we would approach it, with no obligation.