Reactive Synthesis for Expected Impacts

TL;DR

This paper extends BPMN with probabilistic and impact modeling to synthesize strategies ensuring expected impacts stay below thresholds, addressing complex business process optimization.

Contribution

It introduces BPMN+CPI, a formal extension for modeling impacts and probabilities, and provides a PSPACE algorithm for strategy synthesis to control expected impacts.

Findings

The problem of impact-constrained strategy synthesis is in PSPACE.

An effective procedure for computing such strategies is proposed.

The approach enables better resource and impact management in business processes.

Abstract

As business processes become increasingly complex, effectively modeling decision points, their likelihood, and resource consumption is crucial for optimizing operations. To address this challenge, this paper introduces a formal extension of the Business Process Model and Notation (BPMN) that incorporates choices, probabilities, and impacts, referred to as BPMN+CPI. This extension is motivated by the growing emphasis on precise control within business process management, where carefully selecting decision pathways in repeated instances is crucial for conforming to certain standards of multiple resource consumption and environmental impacts. In this context we deal with the problem of synthesizing a strategy (if any) that guarantees that the expected impacts on repeated execution of the input process are below a given threshold. We show that this problem belongs to PSPACE complexity class; moreover we provide an effective procedure for computing a strategy (if present).