728 banner ad

728 banner ad
220 banner ad

220 banner ad

Hyundai Invests in Metawave for Auto Radar Platforms

| May 15, 2018

Metawave for Auto Radar Platforms

Hyundai CRADLE, Hyundai Motor Company’s corporate venturing and open innovation business, announced it is investing in Metawave Corporation to build intelligent radars for autonomous vehicles.

Metawave is attempting to revolutionize radar sensing by leveraging adaptive metamaterials (conventional materials using engineered structures with unique electromagnetic properties not found in nature) and artificial intelligence (AI) to create smart radars. Hyundai is closely evaluating new technologies, such as Metawave’s high-performance radar capable of 3-D imaging, for future autonomous platforms.

“Next-generation radar technology can use advanced algorithms for object detection and classification,” said John Suh, vice president of Hyundai CRADLE. “A new radar system that can increase resolution and accuracy with an AI engine will be a disruptive technology.”

Three sensors are the fundamental components of the perception system for self-driving cars today: camera, LiDAR and radar. The camera is the highest-resolution sensor but cannot see objects beyond 50 meters. LiDAR extends the range to about 150 meters, with a fairly high-resolution imaging capability.

Both camera and LiDAR are affected by bad weather and dirty roads.

Radar, by its nature, operates at a lower frequency and “sees” objects at long ranges faster and in all weather and driving conditions. Today’s radar cannot cover wide angles at long ranges and lacks the resolution to differentiate between objects. It needs a large number of antennae and expensive chips to run the complex digital signals. This type of system takes time and energy, making it inefficient.

Metawave’s advanced radar platform, WARLORD uses just one antenna and pushes the complexity of the analog space, using ultra-fast and precise responses. With WARLORD, the antenna itself shapes and steers the beam in all directions using pencil beams, deep learning engines and AI algorithms to recognize objects quickly and send 4-D point cloud to the sensor fusion.

Category: Featured, General Update, News

Comments are closed.