Case study · Automotive supplier
One data platform for 18 plants
Machine data from 18 plants now lands in one governed platform that drives maintenance planning.
- Client
- Automotive supplier
- Industry
- Manufacturing & mobility
- Practices
- Cloud platforms, Data & AI
- Duration
- 7 months
Context
A tier-one automotive supplier with 18 plants across three continents.
Challenge
Each plant kept its own historian, maintenance ran to a fixed calendar and no one could compare plants on the same measures.
Approach
How the work ran
- MO 0–1FrameChose one production line and one failure mode worth predicting, with plant managers.
- MO 1–3ProveStreamed machine data from the pilot line and flagged bearing failures nine days ahead.
- MO 3–7BuildConnected all 18 plants through a common data model and routed alerts to maintenance crews.
- Azure IoT
- OPC UA
- Databricks
- ISA-95 model
- Python
Results
What changed
18
Plants connected
−22%
Unplanned downtime
3mo
Payback period
“We finally compare plants on the same numbers.”
Related work
More from this route
Regional insurer · Financial services
Claims settled in 3 days instead of 19
19→3days to settle an average claim
Hospital network · Healthcare & life sciences
A bed forecast planners act on every morning
92%forecast accuracy at 72 hours
National bank · Financial services
Payments settled in 4.2 seconds, not overnight
4.2send-to-end settlement
Start with one line, scale to every plant.
We will help you pick the first failure mode worth predicting.
Plan a plant data pilothello@speey.com