To help autonomous driving enter the fast lane, Continental invests in chip company Recogni

Continental has acquired a minority stake in the German-American startup Recogni, which is developing a new chip architecture based on artificial intelligence (AI) for real-time object recognition. The future processor will be used in Continental's high-performance vehicle computers and other applications to process sensor data more quickly during automated and autonomous driving.

As a strategic investor, Continental will provide not only financial investment but also expertise in AI, vehicle sensors and advanced driver assistance systems to promote the development of chip design. Continental experts expect that this new chip may go into mass production in 2026.

In addition, this highly specialized new processor can serve as an ultra-economical data booster - enabling vehicle computers to quickly perceive their surroundings with minimal energy consumption, thus laying the foundation for autonomous and unmanned driving. The two companies have agreed not to disclose the acquisition amount.

“Only faster chips can enable connected, automated and driverless technology,” said Frank Petznick, head of Continental’s Advanced Driver Assistance Systems division. “With our research into new chip designs, our strategic partnerships with large chip manufacturers such as NVIDIA and our investment in Recogni, we are taking the first step towards the highly specialized sensor modules and control units required for the high-performance computers of future vehicles.”

"Continental provides strong support to Recogni in the field of autonomous driving to conquer the biggest technical challenges," said R KAnand, CEO of Recogni. "To this end, Recogni will focus on creating the world's highest performance and lowest energy consumption AI and developing ASICs (application-specific integrated circuits) with new inference algorithms."

The market demand for processors that can be optimized for the specific needs of AI is growing. A few years ago, automotive computers needed to analyze several megabytes of sensor data per second, and this data is expected to grow to several gigabytes per second in the next few years. In the past, lane departure warning systems only required a simple black and white camera, but in the future, autonomous vehicles will enable more than 20 high-resolution body environment sensors such as radar, cameras, and lidar at the same time in some cases.

The cooperation between Continental and Recogni was initiated by the Continental startup project co-pace platform. This team of technical experts from all over the world paid attention to Recogni very early, and they have been working closely together in the past 18 months to seek the maximum feasibility of technical cooperation, especially the theoretical efficiency of new chip design.

“It’s all theoretical at the moment, but simulations show that our sensor data analysis will run exponentially faster once we start using the Recogni chip,” said Annika Ratte-Front, head of AI at Continental’s Advanced Driver Assistance Systems division.