AirSim is an open-source platform designed to bridge the gap between simulation and reality for autonomous vehicle development. It offers high-fidelity visual and physical simulations, enabling the creation of large amounts of annotated training data in various conditions and environments. The simulator is built on Unreal Engine, featuring a physics engine capable of real-time hardware-in-the-loop (HITL) simulations with support for protocols like MavLink. Key components include a modular design for extensibility, a detailed vehicle model, accurate environment models, and realistic sensor models. Experiments with a quadrotor demonstrate the simulator's effectiveness, showing close agreement between simulated and real-world flight characteristics. The platform aims to facilitate the development and testing of machine learning algorithms for autonomous vehicles, with future improvements focusing on advanced collision response, camera sensor noise, and more realistic wind and thermal simulations.AirSim is an open-source platform designed to bridge the gap between simulation and reality for autonomous vehicle development. It offers high-fidelity visual and physical simulations, enabling the creation of large amounts of annotated training data in various conditions and environments. The simulator is built on Unreal Engine, featuring a physics engine capable of real-time hardware-in-the-loop (HITL) simulations with support for protocols like MavLink. Key components include a modular design for extensibility, a detailed vehicle model, accurate environment models, and realistic sensor models. Experiments with a quadrotor demonstrate the simulator's effectiveness, showing close agreement between simulated and real-world flight characteristics. The platform aims to facilitate the development and testing of machine learning algorithms for autonomous vehicles, with future improvements focusing on advanced collision response, camera sensor noise, and more realistic wind and thermal simulations.