The situation
The manufacturer faced unplanned equipment downtime and manual quality checks that missed defects. Maintenance was reactive and costly, and there was no early visibility into failing machinery.
What we did
We audited operations, then deployed IoT sensors feeding predictive-maintenance models built in PyTorch and TensorFlow, plus a computer-vision system for automated quality control on the line. Grafana dashboards give operators live insight into equipment health.
The outcome
Downtime fell 35%, the vision system detects 99.5% of defects, and predictive maintenance saves roughly $200k a year.