BedreFlyt Artifact: A Self-Adaptive Digital Twin Software System for Hospital Resource Allocation

Abstract

Socio-technical systems can benefit from self-adaptive digital twins that combine runtime models, simulation, formal reasoning, and operational data to support continuous monitoring, analysis, and runtime decision making. This paper presents the BedreFlyt Artifact, a reusable software system implementing a self-adaptive digital twin for hospital resource allocation. The artifact integrates semantic reflection, lifecycle management, simulation, optimisation, and a human-in-the-loop dashboard within a modular architecture. Together, these components maintain a runtime representation of the hospital and generate bed bay allocation recommendations under changing operational conditions.

The paper describes the architecture, deployment, configuration, and execution of the artifact, and illustrates its use for hospital bed bay allocation across multiple wards through both allocation and simulation workflows. The BedreFlyt Artifact provides an executable and reusable software exemplar for reproducing, evaluating, and extending research on self-adaptive digital twins and runtime decision support in hospital resource allocation.