This paper presents a game-simulated environment that mimics real-world conditions, with a focus on autonomous vehicle navigation. Despite significant advances in the field of games and simulations, there are still a number of challenges to overcome, in particular, the ability to accurately transfer what has been learned in virtual environments to the real world. This project recreates an agent (a motorcycle), modeled with complex physics, navigating autonomously on a detailed map based on the urban geography of Vila Real, Portugal, recreated from real data, implemented in the Unity game engine. In this paper, we provide a detailed overview of the environment and agent creation processes, highlighting the integration of realistic road networks, obstacles, and interaction mechanics that enhance the fidelity of the simulation. The experimental phase demonstrates the motorcycle’s ability to navigate efficiently, adapting to road layouts, avoiding obstacles, and adjusting to dynamic conditions. The insights from this study can be applied and transferred to real-world application scenarios, particularly in optimizing route planning and driving behaviour for electric motorcycles.