Safety critical automated driving applications require reliable sensor suite to create and ensure the safety case of the vehicle. Creating this reliable world view in the different environment conditions while keeping the system cost base controlled is a key challenge. Linking the sensor to a robust and industrialized platform is needed to take the development of the automated driving applications from prototype to serial product.
By leveraging the extensive knowledge inside Sioux available from other domains, Sioux’s smart sensing solutions incorporate sensor fusion that optimize the sensor set and allow to create a reliable world view. Sioux applies knowledge on machine learning and Artificial Intelligence for enhanced sensor processing, both from within the automotive as outside the automotive domain. This includes software for environmental perception, object detection, path prediction, positioning and classification while considering account vehicle motions and the dynamic environment conditions.
Sioux supports with sensor and application proof of concepts on system and vehicle level, integrating with the ADAS system, vehicle network and user interfaces. In the implementation of the sensor module into a platform, increasing power and processing resource restraints put additional challenging on the computing hardware. Sioux supports with selecting the right hardware, made possible by understanding the performance conditions of both the software as the hardware side of the embedded system.
By applying edge computing and AI, Sioux can optimize environmental perception algorithms and define the most cost-efficient hardware platform. Sioux can also apply their own developed functional safety compliant ECU which can be used as basis for development, benchmarking and the serial product for customer specific control systems.