The challenge
UMPI3D uses compressed air in a large part of its processes. It had no real-time data on the compressor’s temperature, vibration, power consumption or pressure, which made it hard to spot problems early and to make informed decisions. Equipment availability was 85%, power consumption was high and maintenance costs were significant.
What we did
- We fitted the compressor with temperature, vibration and pressure sensors and a power consumption meter.
- An Arduino Machine Control, by Arduino PRO, gathers the data and serves it over Modbus TCP to a gateway, which uploads it to the cloud.
- The data is stored in the Amazon cloud (AWS).
- We built the dashboards in Grafana, to see and understand the data collected.
The result
- Equipment availability: from 85% to 95%, with higher productivity and less unplanned downtime.
- 9% less electricity consumption.
- 37% lower maintenance expenses, thanks to real-time monitoring, which allows more effective preventive maintenance.
After this project, training a machine learning model for predictive maintenance is being considered, to detect anomalies before they occur. The company plans to bring IoT to other equipment in its facility.
