In this session recorded at Unreal Fest Orlando 2025, Jim Piavis of Scaled Foundations shows how Open GRID utilizes Unreal Engine as the foundation for cloud-based AI model development, leveraging AirGen (formerly AirSim) for physics simulation and integrating with a large catalog of AI models.
By taking advantage of high-performance graphics, procedural rendering, and Nanite tesselation, Open GRID is able to provide a real-time, interactive simulation environment for robots and their sensors. The adaptation of AI models like object detection, segmentation, and visual navigation are then easily composed and integrated to perform macro-scale tasks that adapt to a changing environment.
Topics covered in this talk will include the perceived gap between simulation and real-world performance; how closing this sim-to-real gap requires a continuum of risk reduction techniques; and how Scaled Foundations has used highly realistic virtual environments to demonstrate the deployment of robust and safe robotic intelligence behaviors from sim to the real world across multiple robotic form factors.
To find out more about simulation using Unreal Engine, check out our website: www.unrealengine.com/uses/simulation
#Simulation, #Robotics, #AIModelDevelopment, #AirGen, #PhysicsSimulation, #ScaledFoundations, #OpenGrid, #Nanite, #UnrealEngine, #UE5, #UnrealFestOrlando