The self-designed Early-Fusion technology preserves the information integrity of the data collected by the sensors and can process high-dimensional sensor data efficiently.
Plans the optimal route between two destinations with the given the semantic map, road network and the traffic information.
Plans the optimal navigate plan through traffic, such as which lanes to choose or what decision to make at intersections etc.
Plans the optimal trajectory which is feasible, safe and within the speed limits.
By constructing simulation scenarios in a virtual environment, the overall performance of the autonomous driving system is tested, which greatly improves the efficiency of the development validation and the quality of the system.
We have self-developed autonomous vehicle roof box, camera and chips for sensors. They are close to mass production from both design and functional point of view..
We generate the HD map by fusing the data (collected from the LiDAR, camera and radar) with the traditional maps and labeled the lane automatically. The map can be applied to a variety of complex environments, with deviation rate of relative distance<0.2%.