Optimizing Runtime Business Processes with Fair Workload Distribution

Abstract

Predictive process monitoring, which utilizes historical event log data from previously executed business processes to provide support for processes currently in progress, has emerged as a prominent area of research in recent years. By leveraging this technology, organizations can improve work efficiency by allocating optimal human resources to tasks based on predictions. However, most of these studies primarily focus on minimizing task completion time, which often results in an imbalance of workload among the human resources executing the business processes. This imbalance can lead to overburdened employees and decreased overall productivity in the long term. In this study, a predictive model generated from event logs is used to forecast activities to be executed in the future and estimate their execution times. Based on these predictions, we propose an online human resource allocation strategy that prioritizes workload leveling. Validation using simulation-generated hospital event logs demonstrates that the proposed method successfully equalizes human resource workloads, albeit with a slight reduction in immediate work efficiency.

Received: 2 Januray 2025 | Revised: 28 February 2025 | Accepted: 27 March 2025

Conflicts of Interest 

The author declares that he has no conflicts of interest to this work.

Data Availability Statement

Data available on request from the corresponding author upon reasonable request.

Author Contribution Statement

Hiroki Horita: Conceptualization, Methodology, Software, Validation, Formal analysis, Investigation, Resources, Data curation, Writing – original draft, Visualization, Writing – review & editing, Supervision, Project administration.