Traditionally, processes are modelled and discovered primarily considering their control-flow dimension, while disregarding other key dimensions that affect the control flow, such as the data dimension. Consequently, the resulting models are unable to suitably represent real-life, widespread processes where behaviour arises from the complex interplay among multiple business objects and their one-to-many/many-to-many relationships. ​New paradigms that combine data and processes, such as object-centric processes, present new perspectives to the field of business process management, but also bring new challenges. 

On the one hand, object-centric processes toned to be correctly specified and modelled. Such multi-perspective models are intrinsically difficult to analyse. This calls for a suitable trade-off between expressiveness and feasibility of analytic techniques. In addition, object-centric models can span a complex network containing many processes and objects. Thus, they bring great potential to create models that cross process and organisational boundaries that current modelling techniques impose. 

On the other hand, process mining focus on the discovery and analysis of object-centric models from event data, which poses a twofold challenge. First, the current notion of event logs fail in representing event data for object-centric processes. Conventional event logs contain isolated traces, whereas object-centric processes require richer, relational and graph-structured representations of the event data. Second, novel process mining techniques and suitable event data/log formats have to be studied so as to operate over such complex data and fully unleash the insights hidden therein. This also brings the need to ensure that modelling constructs in object-centric process notations can be effectively and efficiently discovered and analysed.