The slam algorithm based on multi-sensor fusion is used to construct high precision 3D reconstruction point cloud and semantic map. Based on the high-precision map, real-time laser positioning, visual positioning and semantic positioning which are integrated at the vehicle end to achieve an all-climate and multi-scenario centimeter-level positioning system.
Multi-sensor fusion such as vision and laser radar are adopted to provide accurate full-time sensory information of the surrounding environment in an all-round way.
Driven by the scenario data of multiple application fields, it provides the prediction, decision, planning, and control algorithm that supports full-scenario and balances safety, smooth and precision at the same time.
Based on automatic collection, storage, transmission and analysis of the whole process of vehicle-to-road data, the safety and user experience of the driverless algorithm system are continuously optimized, and real-time and predictive operation and maintenance of system components are performed to support the operation and management of autonomous driving applications
Designed to provide high performance, high reliability and highly integrated intelligent driving function controller for L3-L4 autonomous driving The controller can be flexibly deployed in various models to meet the mass production application demand for automotive-grade application