Our sensing solution, DeepRoute-Sense, includes a slim vehicle roof box and advanced sensor fusion calibration service. The box contains 8 vehicle cameras, 3 LiDARs, GNSS, among other sensors, and the corresponding telecommunication and data synchronization controllers.
Our perception technology assists with accurate interpretations of the vehicle’s surroundings. DeepRoute.ai’s perception method was ranked #1 for bird’s eye view cyclist detection on the KITTI dataset. For bird’s eye view vehicle detection, it is ranked #6 and has proven more successful than competitors.
Through semantic mapping and a real-time positioning system, DeepRoute.ai’s planning and control algorithm can plan the optimal path under complex road environments, and provide a safe, comfortable and effective driving experience.
By creating simulations of urban roads, vehicles, and pedestrians, we can conduct comprehensive performance testing of our self-driving technology in a virtual environment. This greatly improves the efficiency of our development and provides cost-effective solutions.
DeepRoute.ai utilizes a 5G remote control self-driving system.
