Starting point
In sequenced manufacturing — automotive assembly, for instance — containers have to reach the line in exactly the right order at exactly the right takt time. The consultancy had mastered this just-in-sequence logic but introduced it at each client from scratch by hand: paper lists, mental arithmetic at the line, no traceability. It wanted to cast its method into software it could deploy at a first automotive supplier — and reuse at further plants afterwards.
Solution
We translated the consultancy's takt model into a real-time platform: every container is a movement with a sequence position on the belt; the software continuously compares belt progress against the moment the container must be emptied and projects that onto a traffic light operators can act on instantly. Shift and break schedules shift every deadline automatically, so the traffic light stays reliable across breaks and shift changes.
- Real-time traffic light per product group with a takt countdown
- Scanner-driven operator flow — the wrong container or wrong gate is caught before the mistake
- Shift- and break-aware takt control that re-projects deadlines automatically
- Control center for supervisors with belt stop/start and master-data management
- SOS function for cross-station help in real time
- Built as a product core — a dedicated, isolated install per client
Results
Mental arithmetic and paper lists became a glanceable real-time board that ran in production for three years in multi-shift operation — across test, stage, and production environments. The consultancy now has more than a methodology: a demonstrable product, built once and rollable out to each further client as its own install. Because such line-side systems run close to the plant, we set up the install per client on dedicated infrastructure — on-premises at the client's plant on request.
- Methodology as a reusable product instead of project-by-project rollout
- Three years of production use in a multi-shift environment as a reference
- A dedicated, isolated install per client — on-premises possible
- Error prevention at the line instead of after-the-fact root-cause hunts
