Proactive Data-driven Scheduling of Business Processes

Authors: Francesca Meneghello, Arik Senderovich, Massimiliano Ronzani, Chiara Di Francescomarino, Chiara Ghidini

IJCAI 2025 | Venue PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We evaluate our approach using synthetic datasets with varying levels of uncertainty and size. In addition, we apply the approach to a real-world dataset from an outpatient cancer hospital, demonstrating its effectiveness in optimizing the process Makespan by an average of 5% to 14%.
Researcher Affiliation Academia 1Fondazione Bruno Kessler, Via Sommarive, 18, POVO 38123, Trento, Italy 2Sapienza University of Rome, Via Ariosto, 25, 00185, Rome, Italy 3York University, 4700 Keele St, Toronto, ON M3J 1P3, Canada 4University of Trento, Via Sommarive, 9, 38123 Trento, Italy 5Free University of Bozen-Bolzano, via Bruno Buozzi, 1 39100, Bozen-Bolzano, Italy EMAIL, EMAIL, EMAIL, EMAIL
Pseudocode No The paper describes the methodology and model construction in paragraph form and through mathematical equations. It does not contain any clearly labeled pseudocode or algorithm blocks.
Open Source Code Yes In the repository https://github.com/francescameneghello/ IJCAI2025-Proactive-Data Driven-Scheduling-Business-Process provides further evidence of the differences between applying q U and q C.
Open Datasets Yes As synthetic data we use three problems from a set of publicly available JSP beanchmarks of different size small (10 10), medium (20 20), big (50 20) [Reijnen et al., 2023]
Dataset Splits Yes we use an entire year (2021) of data from Day Hospital to learn the BPSP parameters and minimize the Makespan for the subsequent five months (January-May 2022), which serve as the test set.
Hardware Specification Yes The experiments are conducted on a PC with 16 GB of RAM and an M2 processor.
Software Dependencies Yes We use the Python version of Google OR-Tools (v9.11.4210) as our CP solver.
Experiment Setup Yes A 3-minute time limit is set for the CP solver, and the evaluation with RIMS, thanks to the k-solution selection, takes an average of 5 minutes, with a maximum of 15 minutes for the largest problem in the Day Hospital experiment. ... Specifically, we simulate the BPSP problem 1,000 times using stochastic activity durations and identify the maximum critical path, π.