← Back to VizionEdge

Case · Belgian compressor manufacturer

Compressors that report their own health.

Industrial IoT · Real-time telemetry · Predictive maintenance

20198 monthsBrian's work as a consultant, before VizionEdge

The goal

Sell maintenance before the breakdown, not after.

A Belgian compressor manufacturer wanted to sell predictive maintenance to its customers: know that a compressor needs attention before it stops, and turn that knowledge into a service customers pay for. That starts with one thing: live data out of the machines, in one place, that people can read.

What was built

A real-time pipeline, from the machine to the alert.

Compressors, then Node-RED, Kafka, Elasticsearch and Kibana. Deployment automated with Ansible and Jenkins. Today we would add a failure prediction model, AI that explains each alert, and a service contract to sell.
  • Node-RED flows collecting telemetry from the compressors.
  • Kafka streaming every reading in real time.
  • Elasticsearch storing and indexing the readings.
  • Kibana dashboards and alerts on machine behaviour.
  • Deployment automated with Ansible and Jenkins, so every new flow goes out the same way.

What we would add today

Step one was getting the data out. Step two is selling what it says.

  • A failure prediction model trained on the history you already store.
  • AI that writes each alert in plain language for the service team, with guardrails: it only uses your machine data, and a person approves before a customer sees it.
  • The offer packaged as a contract tier: what the customer gets, the response time, the price.

That is offer 02, From monitoring to prediction.

Your turn

Your machines already send data?

Then most of step one is done. Let's talk about step two.