The introduction of the Industrial Internet of Things (IIoT) has led to major changes in the
industry. Thanks to machine data, business process management methods and techniques could also
be applied to them. However, one data source has so far remained untouched: The network data of
the machines. In the business environment, process mining, for example, has already been carried
out based on network data, but the IIoT, with its particular protocols such as OPC UA, has yet to be
investigated. With the help of design science research and on the shoulders of CRISP-DM, we first
develop a framework for process mining in the IIoT in this paper. We then apply the framework to
real-world IIoT network traffic data and evaluate the outcome and performance of our approach in
detail. We find tremendous potential in network traffic data but also limitations. Among other things,
due to the dependence on process experts and the existence of case IDs.