Client Context
A mid‑size manufacturing company operating multiple facilities across Europe needed real‑time visibility into machine performance, energy consumption, and downtime events.
The Problem
Their existing monitoring relied on manual logs and isolated PLC data. No unified dashboard, no predictive alerts, and no remote access.
Objectives
- Centralize machine telemetry
- Enable real‑time dashboards
- Reduce downtime
- Provide remote diagnostics
- Support future automation workflows
Architecture & Approach
I designed a distributed IoT architecture using ESP32 edge devices, MQTT messaging, and a cloud‑native backend built with Java microservices and Docker containers. Data was streamed into a time‑series database and visualized through custom dashboards.
Technologies Used
- ESP32
- MQTT
- Docker
- Java + Spring
- TimescaleDB
- Grafana
- Azure IoT Gateway
Challenges
- Legacy PLC integration
- Unstable WiFi environments
- High‑frequency sensor data
- Multi‑facility synchronization
Solution
A modular IoT platform capable of ingesting sensor data, machine states, and environmental metrics. Edge devices performed local filtering and batching to reduce bandwidth. Cloud services handled analytics, alerting, and visualization.
Outcome & Impact
- 40% reduction in unplanned downtime
- 3× faster incident response
- Unified monitoring across all facilities
- Predictive alerts for vibration and temperature anomalies
Metrics
- 99.9% uptime
- 12,000+ data points/min
- 5 facilities connected
- 18 dashboards deployed