Jan 31, 2022

3 Minutes Read Zendar, a startup developing innovative high-resolution automotive radar technology, today announced that Hyundai Mobis, in 2021 ranked the world’s seventh largest automotive Tier 1 supplier, is buying a minority stake in the company as part of its Series B funding round. This investment from an automotive Tier 1 manufacturer provides strong evidence that Zendar’s software-defined radar technology is set to disrupt the market for automotive ranging and object detection technology, a key component of next-generation driver assistance and autonomous vehicle systems. Zendar fuses information from multiple radar front-ends distributed around the vehicle with proprietary Distributed Aperture Radar (DAR) technology to produce radar resolution comparable to that of LiDAR (optical) ranging technology, but at much lower cost, and without LiDAR’s vulnerability to adverse operating conditions such as rain, snow or dust. Since Zendar’s technology is implemented purely in software, the cost and size of the hardware is lower than that of today’s automotive radar implementations. Vinayak Nagpal, co-founder and CEO of Zendar, said: “Improving automotive radar systems with the conventional approach of adding more and more antenna channels is not very practical because of the high cost and complexity of the additional hardware required. This new investment from Hyundai Mobis strengthens our conviction that Zendar’s distributed aperture approach is the future of automotive radar, offering ten times better ability to detect and recognize objects without additional hardware cost.” Gregory Baratoff, Vice President and Head of Autonomous Vehicle System Development at Hyundai Mobis, said: “‘Hyundai Mobis’ leadership position in automotive driver assistance is the result of the relentless deployment of the most advanced technology. By investing now in Zendar’s unique technology, we are readying ourselves to achieve the next step change in autonomous driving capability with technology which is affordable and scalable.” Overcoming the limitations of conventional radar Today’s automotive radar systems are limited to simple functions such as highway autopilot. To progress to higher levels of autonomy, the vehicle requires much higher-resolution ranging information. Previous attempts to develop high-resolution radar have taken a hardware-intensive approach which substantially increases system cost and complexity – a strong deterrent to mass adoption in the automotive industry. By contrast, Zendar has developed a set of breakthrough software technologies for radar sensor configuration and radar signal processing. Together, they produce a 10x improvement in the vertical and horizontal resolution of today’s radar systems without increasing the overall cost of the solution. Zendar software works by fusing raw information from multiple radar front-ends with motion data from inertial measurement units, and uses machine learning methods instead of traditional radar signal processing algorithms to detect targets. Zendar uses simple and inexpensive radar front-ends combined with a central processor to scale up the resolution of the multiple existing radar systems in the vehicle. The Zendar software runs on the vehicle’s existing ADAS compute infrastructure. The software combines multiple radar front-end inputs to produce in real time a 4D (x, y, z and doppler) depth map which has LiDAR-like detail, but which does not have the high cost of LiDAR, and which works in all weather conditions and nearly all terrains. Demonstrations of the Zendar radar technology on public roads prove that it produces a sufficiently detailed depth map to support Level 5 autonomous operation. In highway mode, Zendar’s radar has a range of 350m, sufficient to support automated driving at highway speeds. TAGS