Multimodal Process Analysis

Traditional process analysis and process mining techniques are primarily based on structured event logs produced by enterprise information systems (van der Aalst and Berti 2020 ). While effective for highly automated, well-instrumented workflows, these approaches often fail to capture informal practices, tacit decision-making, physical-world actions, and contextual factors that shape process execution (Corradini et al. 2025a , b ; Gavric et al. 2024a ). Processes have long been understood as socio-technical systems; what is new today is the growing availability of additional observational channels – video, audio, text, sensor streams, and interaction traces – through which this socio-technical reality can be captured alongside classical event logs (García-Bañuelos et al. 2025 ).