The thesis Towards Smart and Shared Intralogistics explores the transformation of manufacturing under Industry 4.0 and 5.0, focusing on high-mix, low-volume (HMLV) environments where flexibility and responsiveness are key. Central to this transformation is the use of Automated Guided Vehicles (AGVs) for dynamic material handling. The research addresses the scheduling, dispatching, and control of heterogeneous AGV fleets, inspired by the Brainport Industries Campus (BIC) in Eindhoven—a collaborative high-tech ecosystem.nnInitial studies focus on offline scheduling for single- and multi-load AGVs, proposing hybrid algorithms like Adaptive Large Neighborhood Search (ALNS) to optimize cost, fleet size, and efficiency. Later chapters introduce learning-based dispatching methods, including interpretable linear models and advanced Deep Reinforcement Learning (DRL) frameworks such as ND3QN and MAD3QN. These models significantly outperform traditional heuristics, improving cost-efficiency and adaptability.nnThe thesis culminates in the design of a modular digital platform, Intra-Logistics-as-a-Service (ILaaS), enabling shared AGV fleets across multiple tenants. ILaaS integrates planning algorithms with real-time data, promoting resource virtualization and intelligent coordination. The platform’s success hinges on cooperative governance and shared incentives, offering a scalable solution for resilient, efficient, and flexible intralogistics in modern manufacturing.