Dipartimento di Ingegneria informatica, automatica e gestionale

Increasing industrial resilience is a big challenge for manufacturing enterprises that are continuously facing severe accidents causing injuries, casualties, and economic losses. Assessing industrial resilience requires the analysis of production processes in order to find possible safety flaws. Sociotechnical process management suffers often from misalignments of process descriptions according to formal organization documents or manager views (Work-As-Imagined) and actual work practices as performed by sharp-end operators (Work-As-Done). Furthermore, existing modelling approaches leveraging on techniques such as process mining from digital traces cannot be used to solve such misalignments as these traces are often hardly available. In this context, we propose a computational creativity approach for a semantics-driven transition from Work-As-Imagined to Work-As-Done process models based on the functional resonance analysis method (FRAM). In particular, through formalized semantics, it will be possible to use automatic reasoning for identification of criticalities and prioritization of normal work analyses. To this aim, we introduce some examples of rule patterns, inspired by typical data quality issues, which can be automatically applied to guide such a transition. An explorative case study on chemical cleaning for industrial process is presented to clarify the proposed approach.