Autonomous Vehicle Hazard Perception Clips

Five.AI is developing the world's most reliable autonomous software stack for driverless cars in complex urban environments. Their AI-based approach — using multiple sensors, computer vision, neural networks and motion planning — required compelling visual proof of concept. They needed clips they could show to prospective investors that highlighted all the potential hazards a driverless car would encounter on a real urban journey, demonstrating the extraordinary complexity of what they were building.

Five.AI approached Jellylearn because of the quality of the work we had done for the DVSA in developing the UK Hazard Perception Test. Drawing on our deep expertise in producing validated CGI hazard perception content, we created a series of high-quality photorealistic 3D clips set in complex urban environments. Each clip depicted the full range of hazards an autonomous vehicle must identify and respond to — from pedestrians and cyclists to unusual lighting, reflections in shop windows, and challenging weather conditions.

Watch the clip