BPMN to PDDL: Translating Business Workflows for AI Planning

TL;DR

This work addresses the gap between descriptive BPMN diagrams and executable planning models by translating BPMN 2.0 diagrams into PDDL domains and problems. It proposes an end-to-end pipeline that maps core BPMN constructs to PDDL, including tasks, events, gateways, and message flows, and employs a Fully Observable Non-Deterministic planner to generate execution traces. The main contributions are an operational BPMN-to-PDDL translator, a formal encoding of gateway semantics with non-determinism, and demonstrative results on a credit-scoring workflow, with a GitHub repository of outputs. The work enables formal analysis, simulation, and validation of business processes under uncertainty, bridging descriptive process models with automated planning tooling and paving the way for broader practical adoption.

Abstract

Business Process Model and Notation (BPMN) is a widely used standard for modelling business processes. While automated planning has been proposed as a method for simulating and reasoning about BPMN workflows, most implementations remain incomplete or limited in scope. This project builds upon prior theoretical work to develop a functional pipeline that translates BPMN 2.0 diagrams into PDDL representations suitable for planning. The system supports core BPMN constructs, including tasks, events, sequence flows, and gateways, with initial support for parallel and inclusive gateway behaviour. Using a non-deterministic planner, we demonstrate how to generate and evaluate valid execution traces. Our implementation aims to bridge the gap between theory and practical tooling, providing a foundation for further exploration of translating business processes into well-defined plans.

Figures (6)

• Figure 1: Examples of parallel, inclusive, and exclusive BPMN gateways
• Figure 2: BPMN Diagram
• Figure 3: Generated PR2 Policy
• Figure 4: PDDL domain header and type declaration.
• Figure 5: Action encoding of request_credit_score.