Solving workflow scheduling problems with QUBO modeling

TL;DR

This paper models workflow scheduling as a QUBO problem, compares quantum and classical algorithms on real instances, and introduces a decomposition method to handle complexity.

Contribution

It presents a novel QUBO formulation for workflow scheduling and a decomposition technique to improve scalability and solution quality.

Findings

QUBO complexity depends on input size

Decomposition improves problem tractability

Quantum and hybrid algorithms show promising results

Abstract

In this paper we investigate the workflow scheduling problem, a known NP-hard class of scheduling problems. We derive problem instances from an industrial use case and compare against several quantum, classical, and hybrid quantum-classical algorithms. We develop a novel QUBO to represent our scheduling problem and show how the QUBO complexity depends on the input problem. We derive and present a decomposition method for this specific application to mitigate this complexity and demonstrate the effectiveness of the approach.