Skip to content

Track record · Machinery · printing presses

Turning machine data into a service business

A major printing-press manufacturer (over 10,000 employees, 2.5 billion euros in revenue) wanted to put its machine data to work. The result: an ML solution that predicts unplanned downtime from sensor values, failure history and BOM data, sold to end customers as a value-added service.

The challenge

The problem

Maintenance contracts were in price competition with third parties. The manufacturer needed an offering third parties cannot copy: predictions from its own machine data.

Strategically, the solution had to support a subscription-based business model.

Predictive Maintenance

The solution

How it was solved

Failure prediction ML models learn the relationship between sensor values, history and failures.
Root-cause analysis When a machine is going to fail, and what the cause is.
Live integration Integration with customer systems, value delivered at the end customer.
Spare-part triggers Time-sensitive prompts for needed parts create additional revenue.

The results

~€5M annual recurring revenue (subscription)
€2.2M additional parts revenue
5,000 customers in the service worldwide

Technology

Methods used

Stack

  • Machine Learning
  • IoT/Sensorik
  • Time-series analysis
  • Subscription-Service

End customers benefit from less unplanned downtime, optimised machine performance and longer lifetime.

Led by Beyonetix founders and senior engineers. Figures per the respective project report.

A comparable problem in your operation?

We bring this experience to your project. Let's talk.