Predictive Maintenance Platform for Manufacturing

I led a digital transformation initiative for a multi-site manufacturer by deploying a predictive maintenance platform powered by IIoT sensors and machine learning. The goal was to reduce downtime and extend the lifespan of high-value industrial assets.

Project Snapshot:
The company suffered unplanned outages from wear-part failures in CNC, stamping, and casting lines. No telemetry or trend analysis existed. They needed a real-time dashboard and alerts engine tied to actual machine health.

Key Challenges:

Zero real-time telemetry from machines
Reactive maintenance culture
Technicians had no mobile interface or alerts
Significant lost revenue from downtime

What I Did / Led:

Installed edge IIoT sensors on critical assets (temp, vibration, current draw)
Built real-time alerts and condition-based maintenance logic
Integrated factory PLCs with Azure IoT Hub and Power BI
Deployed predictive models trained on failure history
Designed technician mobile app to receive alerts and log resolutions

Impact:

Reduced unplanned downtime by 58%
Increased equipment lifespan by 27%
Enabled shift supervisors to reroute jobs dynamically
Achieved ROI in 7 months

Tech Stack:
Azure IoT Hub, Power BI, Python ML, MQTT, SQL Server, React Native (tech app), OPC UA