Facility layout planning is a core discipline in production management, directly shaping operational efficiency, material flow, and cost structures. Despite its criticality, facility layout planning presents a complex combinatorial problem, often approached through heuristics or metaheuristics that lack scalability and adaptability. This book investigates the use of (Deep) Reinforcement Learning (DRL) to automate and enhance layout planning by conceptualising facility layout planning as a Markov Decision Process (MDP). The author found that DRL agents trained solely through interaction feedback without domain-specific input can autonomously generate layout configurations that significantly reduce material handling costs and generalise across varying problem instances, thus demonstrating DRL's viability as a scalable and adaptive resolution technique for facility layout planning. Building on the conceptual parallel between human iterative layout adjustment and Reinforcement Learning processes, this research follows a Design Science Research paradigm of experimental artefact design. It unfolds over four peer-reviewed publications. Beyond the experimental contributions, this work opens a path toward AI-driven factory planning tools that can potentially reduce planning effort, improve layout quality, and ultimately enable more responsive and data-driven production system design in dynamic industrial environments.