For socio-technical system, which require a tight collaboration among humans and machines, a flexible process support is crucial. However, the digital workflows driving the interactions between humans and machines need to be highly adaptive and customizable in order to meet practical demands, e.g., to enable a flexible order management for manufacturing products with many variants or to dynamically trigger and execute maintenance processes to cope with disruptive events. In the context of project P6, an adaptive process management technology, developed in previous work by Ulm University, shall be enhanced with the ability to automatically detect deviations between digital workflows on one hand and the processes actually taking place in the physical world on the other. Respective deviations shall be detected based on advanced techniques (e.g. sensor data analysis). Furthermore, novel techniques shall be developed to flexibly handle them, e.g., by automatically adapting the digital workflow in order to realign it with the real-world process. Note that such dynamic and automated workflow adaptations need to be accomplished in a controlled, correct and efficient way.