4D radar is a technology that uses echolocation and a concept called time-of-flight measurement to map objects in a 3D environment. It is currently being tested in the autonomous vehicle industry to map the locations of items in a vehicle's path. It differs from the older technologies like Lidar, standard radar, and cameras because 4D radar can tell when a vehicle is moving and how fast in all types of weather and environmental conditions.
Cameras can be affected by bright sunlight or darkness and can have problems with depth and contrast. Standard radar works well during harsh weather but can't detect objects in much detail. Lidar is great for detecting objects in space but is affected by bad weather conditions.
4D radar solves many of the issues these technologies have. It is a relatively new technology that uses a large Radio Frequency (RF) channel array to detect the relative speed, distance, and azimuth of items in the roadway, as well as the height of the objects above the road. The fourth dimension does refer to time, but 4D radar doesn't really map time. Instead, it uses time in its calculations to determine the elevation of objects and their relative speed.
To map the environment around the vehicle in high-resolution, a 4D imaging radar uses a Multiple Input Multiple Output (MIMO) antenna array. This can include dozens of antennas that transmit signals to targets in the device's surrounding environment and then receive the signals as they bounce off targets. The data that is received by the antennas is then used to generate a point cloud that represents the area surrounding the array. A large array enables accurate detection of both static and dynamic objects simultaneously in high detail. It can also capture doppler shifts in the environment and use them to indicate the direction an object is moving.
Because 4D radar sends dense signals out in all directions, it can work inside the cabin of a vehicle by classifying children and adults, monitoring vital signs, and detecting the position of occupants. This feature can be used to optimize the deployment of airbags, optimize the tensioner in seat belts, provide advance seat belt warnings, and detect intruders in or around the car.
4D radar also works for short, medium, and long-range applications because it has more range than current computer vision technologies. Another feature of 4D radar is the ability to see through objects because it does not rely on cameras or optics.
To summarize, here are the basics of 4D radar:
- 4D radar uses a large Multiple Input Multiple Output (MIMO) antenna array for echolocation. It sends signals that bounce off objects in the environment and captures the results to calculate the size, location, direction, speed, and elevation of objects in the environment.
- 4D radar replaces older technologies like cameras, radar, and Lidar that are used for computer vision and autonomous driving in vehicles.
- 4D radar has the advantage of being able to work in any weather and any level of lighting, detect elevation, speed, and direction accurately, and detect targets behind other objects in the environment.
Sources:
- https://www.forbes.com/sites/forbestechcouncil/2021/04/08/the-role-of-4d-imaging-radar-in-reaching-level-two-plus-autonomy/
- https://www.eetimes.com/4d-imaging-firm-raises-109m-for-see-through-walls-tech/
- https://f.hubspotusercontent10.net/hubfs/5865769/Automotive/Brochures/Vayyar%20Automotive%20White%20Paper%200315.pdf
- https://www.electronicdesign.com/markets/automotive/article/21151406/electronic-design-4d-imaging-radar-looks-to-advance-adas-and-level-5-automation