SLYY150D October   2018  – April 2025 OPA855 , OPA857 , OPA858 , OPA859

 

  1.   1
  2.   Overview
  3.   Detection and imaging in autonomous cars
  4.   Lidar types
  5.   The Lidar subsystem
  6.   Lidar system integration
  7.   Conclusion
  8.   Additional resources

Detection and imaging in autonomous cars

Manufacturers are outfitting modern cars with a wide array of advanced control and sensing functions. Collision warning and avoidance systems, blind-spot monitors, lane-keeping assistance, lane-departure warning and adaptive cruise control are established features that assist drivers and automate certain driving tasks, making driving a safer and easier experience.

Lidar, radar, ultrasonic sensors and cameras have their own niche sets of benefits and disadvantages. Highly or fully autonomous vehicles typically use multiple sensor technologies to create an accurate long- and short-range map of a vehicle’s surroundings under a range of weather and lighting conditions. In addition to the technologies complementing each other, it is also important to have sufficient overlap in order to increase redundancy and improve safety. Sensor fusion is the concept of using multiple sensor technologies to generate an accurate and reliable map of the environment around a vehicle.

Ultrasonic waves suffer from strong attenuation in air beyond a few meters; therefore, ultrasonic sensors are primarily used for short-range object detection.

Cameras are a cost-efficient and easily available sensor; however, they require significant processing to extract useful information and depend strongly on ambient light conditions. Cameras are unique in that they are the only technology that can “see color.” Cars that have the lane-keep assist feature use cameras to achieve this feat.

Lidar and imaging radar share a broad array of common and complementary features that can map surroundings as well as measure object velocity. Let’s compare the two technologies in several categories:

  • Range. Lidar and imaging radar systems can detect objects at distances ranging from a few meters to more than 200 m. Imaging lidar has difficulty detecting objects at close distances. Radar can detect objects from less than a meter to more than 200 m; however, its range depends on the type of system: short-, medium- or long-range radar.
  • Spatial resolution. This is where lidar truly shines. Because of its ability to collimate laser light and its short 905- to 1,550-nm wavelength, infrared (IR) light spatial resolution of approximately 0.1 degrees is possible with lidar. This resolution enables high-resolution 3D characterization of objects in a scene without significant back-end processing. On the other hand, radar’s wavelength (4 mm for 77 GHz) has challenges resolving small features at long distances.
  • Field of view (FOV). Solid-state lidar and radar both have excellent horizontal FOV (azimuth), while mechanical lidar systems, with their 360 degrees rotation, possess the widest FOV of all advanced driver assistance systems (ADAS) technologies. Historically, lidar has better vertical FOV (elevation) than radar. Lidar provides angular resolution (for both azimuth and elevation), which is one primary feature necessary for improved object classification.
  • Weather conditions. One of the biggest benefits of radar systems is their reliability in rain, fog and snow. The performance of lidar generally degrades under such weather conditions. Using IR wavelengths of 1,550 nm helps lidar achieve improved performance under adverse weather conditions.
  • Ambient light. Lidar and cameras are both susceptible to ambient light conditions. At night, however, lidar and imaging radar systems offer very high performance because they provide their own illumination. Radar and modulated lidar techniques are resistant to interference from other sensors.
  • Cost and size. Radar systems have become mainstream in recent years, making them highly compact and affordable. As lidar has become more popular, its cost has dropped precipitously, with prices dropping from approximately US$50,000 to below US$10,000. The mainstream use of radar in modern-day vehicles is made possible by increased integration, which reduces system size and cost. The mechanical scanning lidar system from a few years ago – commonly seen mounted on various autonomous self-driving robotaxis – is bulky, but advances in technology have shrunk lidar over the years. The industry shift to solid-state lidar will further shrink system size and lower costs.