Site discovery
2 weeksLine inventory, sensor points, MES API.
Manufacturing · IoT & Analytics
OEE analytics platform from shop floor to ERP
Line-level real-time efficiency metrics instead of delayed manual reporting.
Problem
OEE data was collected via Excel and nightly batch reports. Maintenance planning was reactive; downtime root causes were analyzed days later.
Solution
A single data pipeline from shop floor to management with IoT gateway, MES adapter, time-series store, and OEE dashboard.
Architecture
Edge gateway collects sensor data, stream processor normalizes, analytics store computes OEE, dashboard API serves metrics.
1. Edge IoT gateway (OPC-UA / MQTT)
2. Stream processor & deduplication
3. Time-series analytics store
4. OEE dashboard & alert API
Timeline
Pilot line
8 weeksGateway, stream pipeline, OEE MVP.
Expansion
6 weeksAdditional lines, maintenance alert rules.
Go-live
2 weeksTraining, SLA, hypercare.
Challenges
· Limited digital outputs on legacy machines
· Network outages during night shifts
· OEE definition differences across lines
Performance metrics
24h → 5 min
OEE report lag
−18%
Unplanned downtime
−30%
Maintenance planning time
Business impact
Production leadership can make line-level efficiency decisions same-day. Maintenance gained early-warning intervention capacity.
ROI
Investment payback in 14 months through efficiency gains (internal calculation).
Lessons learned
· OEE definitions should be clarified per line in workshops.
· Edge gateway offline buffer is critical — prevents data loss during network outages.
Technologies
PythonPostgreSQLMQTTDockerNext.jsTimescaleDB
KVKK / GDPRNDACI/CDTestingSecuritySLA & Support
Sık sorulan sorular
Which machines were integrated?
Two production lines in pilot; expansion covered five lines total.
Is predictive maintenance included?
Anomaly detection module added in phase 2; full predictive model on roadmap.
Discuss a project like Manufacturing
Book a discovery call or explore the industry page for related solutions.