Modular vehicle technology offers a flexible solution for on-demand transportation sytems and has recently gained attention as a promising approach to improve capacity adaptability and operational efficiency. However, planning such systems is computationally complex, as routing, scheduling and vehicle configuration decisions are tightly interconnected and must be optimized within a unified framework. The present work proposes an integrated optimization framework for modular bus fleets operating in on-demand mobility services. The problem is formulated as a Mixed-Integer Linear Programming model that jointly determines routing, vehicle configuration and passenger service decisions. To solve the problem efficiently on real-world-inspired networks, a two-phase heuristic algorithm is developed.