Contour
Predict failures from the plant data you already hold.
Contour reads your historians, SAP and SCADA, maps every sensor to the equipment it measures, and builds reliability and failure models for each asset. The result is maintenance planned on the condition of the equipment, with every date traced to the records behind it.
Most plants already record what a reliability study needs. The equipment master and failure history sit in SAP, the process signals in a historian, the alarms in SCADA. They were never joined, so the analysis is assembled by hand, one asset at a time, and the maintenance calendar comes from the manufacturer's manual instead.
Contour does the joining. It extracts the asset tree, links each asset to its history and instruments, describes how assets depend on each other in series and in parallel, and fits statistical and machine-learning models over that structure. Your engineers check the method; they no longer build the dataset.
Figure 1 One pump, from records to plan. Contour joins your systems into a model of the plant, fits P-101 on its own history and the instruments attached to it, and states the chance it fails before the planned outage. The dashed line is provenance: the date points back to the records it was built from. Values illustrative.
Bring us your expensive questions
We help you deliver the impossible.
If your plant data should answer it and still can't, reach out.
You have years of failure history in SAP and a historian full of sensor data, and the reliability study is still built by hand, one asset at a time. Or a pilot predicted one failure well and never made it to the rest of the plant. Talk to us: we bring the reliability engineering and the modelling to make your plant data answer.
Product
Symbolic structure, statistical estimates.
The plant is described explicitly: assets, components, variables, and the series and parallel dependencies between them. Reliability statistics and per-asset models are fitted over that description, so a prediction knows which asset it is about and what fails with it.
Contour is built from separate components: connectors, the asset tree, the model context layer, reliability block diagrams, life data analysis, model training and monitoring, and an interface for AI assistants. The same platform serves funds and hospitals too, wherever records exist and their meaning has never been written down.