

Post-Disaster Inventory Management and Relief
A decision support platform using simulation and Markov Decision Processes (MDPs) to optimize post-disaster inventory prepositioning, allocation, and distribution under uncertainty.
Summary
Post-Disaster Relief & Inventory Supply Management (PRISM) is an advanced decision support system designed to address one of the most pressing challenges in humanitarian logistics: managing inventories and supply flows in the aftermath of natural disasters.
At its core, PRISM integrates a comprehensive simulation framework that captures the dynamics of disrupted supply chains, fluctuating demand, infrastructure failures, and uncertain recovery timelines. Building on this foundation, it employs a Markov Decision Process (MDP) model to enable dynamic allocation and replenishment strategies that continuously adapt as new information emerges in disaster environments. PRISM equips policymakers, NGOs, and emergency response agencies with the ability to determine:
- Where to preposition relief items
- How to allocate scarce resources during crises
- How to adapt operations over time as conditions evolve
By combining stochastic modeling, simulation, and MDP-based optimization, the system delivers not only near-optimal solutions but also robust, real-time decision guidance.
Beyond its academic contributions, PRISM offers substantial societal impact. It provides actionable tools for organizations such as AFAD in Turkey, the Red Crescent, and international humanitarian agencies, helping them strengthen preparedness, accelerate response times, reduce shortages, and ultimately save lives. By transforming how relief logistics are planned and executed, PRISM directly supports the development of more resilient communities in the face of future disasters.