Modeling Disaster Response Logistics With Behavioral And Mobility Dynamics

Abstract

This talk presents a people-centric framework for disaster response logistics that integrates optimization modeling, human mobility, and empirical behavioral insights. The work proposes mixed-integer non-linear, discrete-time models to jointly determine the location of points of distribution (PODs) and allocation of relief supplies while accounting for beneficiaries’ movements and endogenous facility choice. Unlike standard formulations based on static, distance-based assignments, beneficiary decisions are captured by a flexible attraction function that incorporates factors such as congestion, perceived supply availability, and social considerations. The objective minimizes a social cost function combining logistical costs with human impacts, including deprivation and walking burdens. Ongoing research extends this framework to incorporate truck-and-drone delivery systems to improve access to vulnerable populations under dynamic demand. In parallel, ongoing empirical work investigates relief supply–seeking travel behavior during the response phase and directly informs the specification of the attraction function and the modeling of beneficiary movement, contributing to a new class of empirically grounded disaster logistics models.