Timing the vaccine flow: An intelligent navigator for strategic delivery and coverage

The unprecedented impacts of the COVID-19 outbreak have increased global interest in secure and efficient vaccine distribution. This study investigates the challenges of planning vaccine distribution, including demand fluctuations, limited data availability, and inequitable access to vaccines. It focuses on the allocation of vaccines to mobile and stationary Public Health Units (PHUs) and the management of demand through online registration and real-time updates. To address these challenges, we develop a data-driven Decision Support Framework (DSF) comprising a data-based and a model-based framework. The data-based framework introduces an augmented robust optimization approach that uses dynamic uncertainty sets to adapt to changing conditions. Two data generation algorithms, integrated with predictive analytics, enhance limited data, improve out-of-sample performance, and increase the explainability of robust solutions. These uncertainty sets are fed into the model-based framework, which integrates three mathematical models to dynamically configure PHUs and optimize the vaccination distribution plan. As a conceptual contribution, a new policy is proposed to maximize coverage through a Dynamic PHU Allocation (DPA) algorithm that updates mobile PHU locations. Additionally, an assignment model reallocates surplus inventory to address shortages and ensures proper storage for future use. The numerical results demonstrate that the proposed DSF enables adaptive decision-making, effectively coping with fluctuations in vaccination demand and reducing vaccination shortages by 80% compared to other approaches. The adaptability of the DSF, combined with the role of mobile PHUs in managing demand fluctuations and ensuring equitable access, offers insights for resource allocation and logistical planning in uncertain scenarios.