A MIXED-INTEGER LINEAR PROGRAMMING MODEL FOR THE HETEROGENEOUS FLEET MULTI-TRIP DIAL-A-RIDE PROBLEM: A MILITARY TRANSPORTATION APPLICATION IN CASNAV (BRAZILIAN NAVY)

This paper proposes a mixed-integer linear programming (MILP) formulation for a Heterogeneous Fleet Dial-a-Ride Problem with Time Windows and Multi-Trip operations, motivated by military transportation planning in the Brazilian Navy context (CASNAV). The model extends classical Dial-a-Ride formulations by integrating heterogeneous fleets, multi-trip routing, soft time windows, and passenger ride time constraints into a unified framework, providing a more realistic representation of operational conditions in defense logistics. Computational experiments, implemented in AMPL and solved using Gurobi, show that the exact approach obtains global optimal solutions for small-scale instances while revealing scalability limitations as problem size increases. Results also indicate that the model can support fleet planning decisions, improve request consolidation, and generate solutions comparable to observed operational practices. The proposed formulation contributes as both an exact optimization model and a decision-support tool for tactical transportation planning in defense and maritime logistics contexts.