The Use of IoT Sensor Data to Assess Maintenance Risk in Service Contracts

The rapid development of computer-based monitoring technologies has found its way into industrial maintenance applications. In recent years, many original equipment manufacturers have collected data on the running conditions of their installed machinery through sensors. This led to high-dimensional datasets where each dimension pertains to an operational condition of a machine. We use the operational sensor data to capture a customer's machine usage and assess its maintenance risk.

We explore their value for full-service service contracts. In a full-service contract, the customer pays a contract fee at the start of each contract period while the service provider bears all maintenance risk. Customers with high maintenance risks incur higher maintenance costs on average. It is vital to assess this maintenance risk, and thus the expected costs, as accurately as possible to ensure the profitability of these service contracts. We propose a usage-based risk assessment based on its effective machine usage. This results in a tailored and more refined risk assessment compared to models relying only on risk factors known at the contract's start.

We validate our approach on a portfolio of about 4,000 full-service contracts of industrial equipment and show how dynamic sensor data improves risk differentiation.

This seminar will take place in-person in room T09-67. Alternatively, click here to join the seminar online.

Meeting ID: 960 5028 6242