Submarines use radar to navigate the deep seas. An autonomous vehicle, on the other hand, would use LiDAR. While they both have very similar names and are based on sensors, radar and LiDAR aren’t quite the same. Often, they are pitted against one another.
Yet, both are also necessary in the future of automated vehicles. These depend on advanced sensor fusion technology to perceive their surrounding environments and keep occupants safe. This need led to a richer development of two systems used to underpin autonomous vehicle stacks: LiDAR vs. radar. Let’s break down the pros and cons associated with each system, starting with explaining what a LiDAR is.
What is LiDAR?
LiDAR, or Light Detection and Ranging, is a remote sensing tool that uses light to detect how far away objects are from the sensor. By shooting out a pulse of light waves that bounce off surrounding objects, it can capture data that is refracted back to create a three-dimensional, 360° map of the surrounding area.
LiDAR sensors are best known for capturing their environment in extreme detail, even better than the human eye (depending on weather conditions and time of day). Here’s an example.
Radar, or Radio Detection and Ranging, is a type of sensor that uses electromagnetic radio waves to determine the distance, angle, and speed of objects related to the source. These sensors can capture data from much further distances than LiDAR systems, but the resolution of these data is less precise. In fact, their results aren’t detailed as LiDARs, whose level of detail enables building exact 3D models of objects.
LiDAR vs. Radar for Autonomous Driving: 5 Key Differences
As the similarity of these two acronyms suggest, LiDAR and radar share a nearly identical function in detecting signals and determining ranges based on the information collected. However, the differences between light waves and radio waves provide pros and cons to automated vehicle systems based on:
- Wavelength Reach
How precise is LiDAR vs radar? LiDAR tracks details with remarkable accuracy in three-dimensional space by capturing the position, size, and shape of objects relative to the sensor. When combined with advanced perception software, this LiDAR data can be analyzed from the “point cloud” and classified as objects and obstacles. By scanning the environment thousands of times every second, LiDAR helps AI make complex decisions around the intent of pedestrians, vehicles, and hazards.
Radar is better suited for capturing information related to velocity and range. Stuck in a two-dimensional world, it cannot capture the breadth of information that LiDAR systems perceive. This means that in some cases, objects may be falsely identified or fail to be detected.
One of the biggest problems previously facing LiDAR systems was their performance in direct sunlight or inclement weather. Because they rely on light waves to capture data, older LiDAR systems could become distorted by raindrops, snow, and fog. Innoviz’s LiDAR systems are resistant to these conditions.
Radar does not rely on visual data, and thus performs optimally in all conditions.
3. Wavelength Reach
Radio waves have much larger wavelengths than light waves—while they detect signals through the same principles, the wavelength frequency of radar vs. LiDAR gives each system different capabilities.
The large wavelength of radio waves allows them to be transmitted at great distances. However, radars in passenger vehicles are limited by the size of the antenna. They can detect signals much further away, but the detail that they capture has low resolution.
Light wavelengths are significantly smaller—LiDAR systems can capture details at a much smaller level from distances camera sensors cannot track. However, they do not have the same wavelength reach as radar systems.
While LiDAR has clear advantages in terms of safety and performance, companies like Tesla have shied away from the technology completely. This is primarily due to one reason: LiDAR’s price point.
Radar may be more affordable to everyday consumers, but as LiDAR technology has evolved, the cost gap has narrowed dramatically. Solid-state LiDAR sensors are significantly more affordable and reliable than their predecessors as they have no moving parts. They’re costing hundreds, not thousands of dollars. As innovation continues and manufacturing occurs at scale, LiDAR will continue to grow less expensive.
Radar is excellent for adaptive cruise control and monitoring cross traffic, blind spots, and collisions. However, radar cannot capture the breadth of information that LiDAR systems perceive. This means that objects can be falsely identified, or not appear if they are too small. These errors have led to crashes leading to several high-profile lawsuits that have resulted in agencies like the National Highway Traffic Safety Association stressing the need for increased federal regulation over these systems.
LiDAR’s ability to precisely capture data makes it the superior choice for features like emergency brake assist, pedestrian detection, and collision avoidance. The granularity of detail vastly outperforms radar- and camera-based technologies.
Why LiDAR Fills the Safety Gap for Autonomous Driving
With one exception, the majority of autonomous vehicle manufacturers agree that LiDAR systems are the future of the industry. With higher accuracy and resolution than radar, LiDAR achieves the promise of autonomous vehicles: a safer world without automotive crashes.
Yet even though LiDAR carries significant advantages, radar still has a place in the self-driving cars of tomorrow through sensor fusion.
Sensor fusion uses LiDAR, radar, ultrasonic sensors, and cameras in unison to give a complete picture of the environment around an autonomous vehicle. By leveraging multiple types of signals and “fusing” them together, the individual weaknesses of each sensor are negated. Simplified, radar may be used for long-distance hazard detection, LiDAR can detect pedestrians at night, and cameras can read traffic signs, all as part of a unified system.
When it comes to autonomous vehicles, radar- and camera-based systems are not sufficient on their own. LiDAR and radar sensors paired together can help overcome what one cannot do on its own.
Take Your Vehicle Further: LiDAR Technology by Innoviz
There are over six million car crashes each year in the United States. The vast majority of these are caused by human error. With LiDAR technology powering autonomous vehicles, needless tragedies like these could soon be a thing of the past.
At Innoviz, we are working tirelessly on creating affordable and safe LiDAR systems for vehicles to make a crash-free future a reality for all. Contact our team to learn more about how we are blazing the trail for the safe roads of tomorrow.