Enabling process mining on sensor data from smart products

In this paper we address the challenge of applying process mining to discover models of human behaviour from sensor data. This challenge is caused by a gap between sensor data and the event logs that are used as input for process mining techniques, so we provide a transformation approach to bridge this gap. As a result, besides the automatic discovery of process models, the transformed sensor data can also be used by various other process mining techniques, e.g. to identify differences between observed behaviour and expected behaviour. We discuss the transformation approach in the context of the design process of smart products and related services, using a case study performed at Philips where a smart baby bottle has been developed. This case study also demonstrates that the use of process mining can add value to the smart product design process.