A Metaheuristic Approach to Routing and Scheduling Hospital-at-Home Visits

Details

Supervisors

Chevalier, Philippe ; Daniel Rebelo dos Santos ; Diana Jorge

Faculty

Louvain School of Management

Degree label

Master [120] : ingénieur de gestion, à finalité spécialisée

Abstract

enThe rapid ageing of the population and the growing pressure on hospital capacity have accelerated the adoption of Hospital at Home (HaH) as an alternative to conventional inpatient care. Yet, the daily organisation of teams and visiting routes is still often performed manually. Nursing and doctor teams must visit geographically dispersed patients while satisfying clinical requirements, priority rules and route-duration limits. This dissertation addresses a multi-depot HaH setting decomposed into depot-level routing and scheduling subproblems under operational constraints. Two complementary solution approaches are developed: a Mixed Integer Linear Programming (MILP) model, used as an exact benchmark, and an Adaptive Large Neighbourhood Search (ALNS) metaheuristic, designed to obtain high-quality solutions within short computational times. The objective combines total travel time with the use of doctor teams, while feasibility is enforced through team-patient compatibility, maximum shift duration constraints and prioritisation rules for urgent patients. The computational study considers different demand scenarios and instance sizes, allowing the metaheuristic to be compared with the exact method across a broad set of instances. Results show that the ALNS obtains solutions very close to those produced by the exact method, while requiring substantially less computational effort. This benefit becomes more pronounced in larger instances, where the exact method frequently cannot prove optimality within the available time and the ALNS matches or improves upon the incumbent solutions obtained by the exact solver. Finally, a decision-support prototype is developed to translate the resulting solutions into route visualisations, illustrating the practical applicability of the proposed approach for the daily planning of HaH visits.