AI Enhances Business Process Autonomy with Multi-Perspective Constraint Planning

Traditional Business Process Management Systems (BPMS) are being augmented with AI to handle more complex process structures. A key concept in this evolution is "Framed Autonomy," where systems can execute processes independently while strictly adhering to predefined constraints. Prior work primarily focused on control-flow constraints, often transforming them into automata-based representations. This new study expands on framed autonomy by integrating multi-perspective constraints, such as data-aware and temporal conditions, into the process framework. It proposes a novel "what-if" analysis tool that, given a partial process execution, leverages this enriched constraint set to suggest optimal next steps. Empirical evaluations demonstrate the technique's scalability and effectiveness in supporting autonomous, constraint-aware decision-making within ABPMS.