Autonomous driving requires multi-sensor data fusion
ON Semi-AI Motive, HW Platform Demo Development


On the 16th, ON Semiconductor announced that it would jointly develop a sensor fusion platform prototype for automotive applications with AImotive, a Hungarian autonomous vehicle company.

The two companies are jointly developing a series of hardware platform demos by combining ON Semiconductor's HD cameras, radar sensors, and preprocessor chipsets with AI Motive's advanced AI-based cognitive algorithms, hardware acceleration, and simulation technologies.

ON Semiconductor offers all four sensor modalities: imaging, radar, lidar, and ultrasonic sensing. By combining these sensor types with the technology of companies like AI Motive, it is possible to meet the requirements of a high-performance sensor fusion platform that supports ADAS and autonomous driving.

Through the platform to be developed, the two companies plan to demonstrate the accuracy, durability, and low latency of AI-based real-time sensor fusion using AI Motive's 'aiWare' hardware neural network (NN) acceleration IP, 'aiDrive' software portfolio modules, and 'aiSim' simulation environment on FPGA-based prototypes.

Davide Santo, Senior Director and General Manager of the Automotive Radar Sensing Solutions business unit at ON Semiconductor’s Intelligent Sensing Group, stated, “Customers expect to improve performance by combining various sensors,” adding, “AI and simulation are key technologies to meet these demands.”

AI Motive CEO Laszlo Kishonti said, “Together, the two companies will demonstrate that using the right chips and algorithms enhances the performance of integrated sensor fusion solutions while providing a cost-effective and practical solution for high-volume production.”