20Vehicles today are relative super heroes compared to older ones and are packed with all kinds of powers. Most cars use a variety of cameras, ultrasonic sensors and radar to enable features like adaptive cruise control, parking assistance, automatic emergency braking, and blind spot monitoring; these allow the cars to “see” but are limited in terms of range and depth. However, new technology is on the way that will allow for upgraded super powers, including autonomous driving.
A recent announcement by self-driving powerhouse Argo AI suggests that self-driving technology has progressed to an important milestone. Recent developments in their LiDAR technology allows them to detect items up to 400 metres away, and even improve its capabilities in low light conditions.
But all that leads us to the question: What exactly is LiDAR? Furthermore, how important is it to the future of the automotive industry?
LiDAR stands for “Light Detection and Ranging” and describes a sensor technology that can create a map of the environment around it. In early self-driving vehicles, you could spot these spinning cylindrical items on the car. They’ve become far sleeker looking since then. The maps created by these devices are critical for self-driving features. Outside of the automotive industry, LiDAR is used on mobile devices, where range isn’t so much of a concern, allowing for features like augmented reality, measuring distances, and blurring backgrounds in photos and videos.
Generally, LiDAR sensors send out infrared light, and measure the time it takes for the light to bounce off an object and back to the sensor, creating a three-dimensional map. However, there are two different types of LiDAR: Time of Flight (ToF) LiDAR and frequency-modulated continuous-wave (FMCW) LiDAR, and while they perform the same function, each one has its advantages and disadvantages.
The ToF type is the most common form of LiDAR on vehicles that map their surroundings by measuring pulses or photons of light that it sends out and bounce back. The other form of LiDAR, FMCW, sends out a continuous stream of light rather than pulses of light, to map its surroundings. This form of LiDAR has a limited field of view, so vehicles with multiple LiDAR sensors are typically using these, while ToF typically has a 360-degree range, allowing for a single device to do the job.
The map created by a LiDAR sensor is important for a self-driving vehicle, as it helps the car “see” the world around it. LiDAR technology provides more depth and detail than other solutions, such as radar and cameras. It can also work at night.
However, LiDAR sensors were originally pretty expensive, so rather than plopping a few of these devices on consumer vehicles, a supplier or third party company would go out with a limited number of vehicles with these sensors to create a map, supply that detailed information to automakers, who will use it with their driver assistance features. Similar to how Google sends out a vehicle with a ton of cameras to get street view photos, a supplier would send a car out with a bunch of LiDAR sensors, which would then get a super detailed map.
General Motors uses LiDAR maps with its hands-free driving technology Super Cruise. “We believe using LiDAR map data helps Cadillac’s Super Cruise outperform other driving assistance systems,” says Stephanie Lang, General Motors Assistant Manager of Advanced Technology Communications. “We engineered the system to leverage LiDAR map data from the start.”
They combine the map data with on-board sensors and other technology to enable Super Cruise. “The precision LiDAR map data provides specific details for upcoming events, like corners and exits, while the real-time cameras look at lane lines and understand the vehicle’s position in the lane,” she explained. “The layered approach of combining precision LiDAR map data with cameras, sensors and GPS enables a hands-free driving experience that provides drivers with confidence and convenience. The HD mapping data also provides high accuracy and high quality road information that ensures Super Cruise can operate on limited access highways, while also providing smooth control through curves, enhancing user comfort.“
While the vehicle itself isn’t outfitted with LiDAR sensors, it relies on a LiDAR-created map from supplier Ushr that is updated quarterly, which provides the rest of the vehicle’s sensors and computers with the data to drive confidently without much driver intervention.
A LiDAR sensor used to cost around $75,000, but prices have dropped significantly and production increased thanks to demand for expanded applications of the tech in items like cell phones. GM has shown it can provide self-driving features without an onboard LiDAR sensor, but is that ideal?
There are some ways that having an onboard LiDAR sensor makes sense for self-driving features. As Canadians know, roads can change drastically due to construction or snowy weather so relying on outdated maps can be a problem.
Also LiDAR fills in the gaps where other sensors struggle. For example, radar is used to detect objects that surround a car, and can determine how far away they are and how fast they’re moving. This is why automakers use radar for parking sensors, blind spot monitors, and adaptive cruise control, but these same sensors struggle when it comes to detecting the exact position, size and shape of an object, elements that are vital for self-driving features like pedestrian, cyclist, and animal detection. On the other hand, radar works well in fog and other inclement weather conditions, which is where LiDAR struggles.
Additionally, cameras are used for safety and driver assist systems, as they can recognize objects pretty well, but struggle in low light and with depth perception, where LiDAR fares better.
These applications and the limitations of each technology show that LiDAR does have a place onboard vehicles aiming to deliver advanced self-driving features, but LiDAR alone can’t get the job done. When it comes to creating a safe driving environment, redundancies and backup are key.
While costs have been a reason to keep sensors off current vehicles, LiDAR suppliers like Argo AI partner with automakers like Ford and Volkswagen to develop some commercial-grade self-driving vehicles that utilize onboard LiDAR. Using Argo AI’s LiDAR technology, Volkswagen is planning to test self-driving vans in Germany this summer, with the goal of launching a commercial delivery and micro-transit service in the country by 2025. Ford has an alliance with VW for its electric vans and autonomous vehicle technology, meaning it’s not far behind.
Volvo has announced its intentions to use LiDAR from supplier Luminar. Vehicles built in 2022, riding on the automaker’s next generation SPA 2 Modular Platform will be hardware ready for the LiDAR gear that will provide fully autonomous highway driving, but only when the safety of the system in a given location and condition has been verified. Volvo describes LiDAR as the key for creating safe autonomous vehicles.
“Autonomous drive has the potential to be one of the most lifesaving technologies in history, if introduced responsibly and safely,” said Henrik Green, chief technology officer at Volvo Cars. “Providing our future cars with the vision they require to make safe decisions is an important step in that direction.”
For now it’s likely that self-driving technology will be used for fleets of robo-taxis, or delivery vehicles, though there have been developments that suggest the tech could be used in the trucking industry. Thanks to the constant progression of LiDAR technology, the limitations and price of these capabilities are becoming more accessible.