Text/Sam Abuelsamid
As part of its DreamDrive assistance package, the recently launched Lucid Air boasts 32 sensors, including 14 cameras, 5 radars and a forward-facing lidar. Image credit: Lucid
2021 has been a tough year for most automakers, and no one really foresees it coming. One of the many secondary effects of COVID-19 has been severe disruptions in supply chains, especially semiconductor supply chains. This challenge is likely to continue into the second half of 2022 and may even worsen for some time as automakers roll out dozens of new electric vehicles and new driver assistance features and sensors.
In the coming year, more vehicles will be equipped with hands-free, partially automated highway driver assistance systems. Back in 2017, General Motors introduced the first such system with its first-generation Super Cruise system. However, it only works for one vehicle, the Cadillac CT6, and it won't start expanding supply until the spring of 2021. Unfortunately, production was suspended due to battery problems with the Chevrolet Bolt EUV, and the chip shortage interrupted Cadillac Escalade's EUV functionality. Still, GM has promised to launch more than 20 models with the Super Cruise system by 2023, including an upgraded full-size pickup.
Ford's BlueCruise driver assistance system debuted on the Mustang Mach-E and F-150 in 2021 and will land more Ford and Lincoln models in 2022. Image source: Ford
Ford unveiled the BlueCruise system on the F-150 and Mustang Mach-E in 2021, and while it's not as powerful as the Super Cruise, it will appear on more models in Ford and Lincoln in 2022. Lexus' flagship model, the LS500h, will also bring an updated version of it next year and will come with a hands-free Teammate system.
Later in 2022, Nissan will finally launch its long-awaited second electric car, the Ariya crossover. Features of ariya include the second-generation ProPilot assist system, which will now allow drivers to take their hands off the steering wheel while driving on the highway.
All of these systems rely on a combination of surround cameras and radar sensors to detect the vehicle's position on the road, as well as the location of other vehicles associated with it. But as the expected capabilities of these systems grow, the sensors that automakers have been using for the past decade or more are no longer sufficient.
From Tesla's AutoPilot/FSD to Toyota's Camry, most cars are equipped with 1.3 megapixel cameras. That's the equivalent of the first digital camera or early mobile camera you bought 20 years ago. Even the first iPhone in 2007 had a 2-megapixel camera, and no one thought it was any better.
The Fisker Ocean will be the first car to include a Magna/Uhnder digital imaging radar sensor with a long-range sensor in front and a mid-range sensor in every corner. Image credit: SAM ABUELSAMID
In the next year or two, we will see cars equipped with 8-megapixel imaging sensors on the market. These cameras with higher resolution and wider dynamic range will be able to identify targets at greater distances and in more challenging lighting conditions than today. In China, some of Xiaopeng Motors' new models, such as the ET5 sedan, will feature higher-resolution camera sensors, and GM's 2023 Ultra Cruise System will do the same. Some other models arriving in North America in 2022 are likely to follow this path as well.
Being able to classify objects at greater distances is undoubtedly important, but it is also important to measure their distance, speed, and direction. Passive cameras can do this, but they can only reliably do this if multiple cameras are pointing in the same direction and the distance between them is known. This allows the software to do some basic trigonometric functions to calculate the range and orientation of objects in the environment. Ideally, active sensors like lidar and radar will also provide this information, especially if the cameras are all pointing in different directions. Since the 1990s, radar-equipped cars have been on the road as part of adaptive cruise control. But these sensors, like those tesla will stop using in mid-2021, have low resolution and are difficult to distinguish between multiple targets.
In 2022, the so-called imaging radar will appear. These high-resolution sensors from companies such as Magna/Uhnder, Continental, Arbe, and Oculii use a variety of technologies to expand the typical 6 channels on current radar to 192 or more channels, and the results look a bit like lidar point clouds. With so many different feedback signals, these sensors can be used to distinguish between vehicles in front of them, overpasses, pedestrians, and cars parked on the side of the road. The first vehicles equipped with the Magna/Uhnder Icon imaging radar will be the Fisker Ocean. An unnamed automaker is also expected to have Continental sensors sometime in 2022.
The InnovizOne lidar sensor is expected to be available on at least one BMW model later in 2022. Image source: INNOVIZ
Lidar-based sensors are another type of active sensor that is about to emerge. Since the first rotating car lidar sensor was used at the 2005 DARPA Challenge, dozens of companies have struggled to make lidar perform better and last longer, at a cost well below $80,000. Lidar sensors that cost well less than $1,000 should be available in 2022. A number of vehicles equipped with lidar have been delivered, including the Lucid Air and the latest Mercedes-Benz S-Class and EQS. In the coming months, the Teammate-equipped Lexus LS500h will be equipped with lidar, and the Volvo XC90 and Polestar 3 will also be equipped with lidar. These vehicles with a common platform will use Luminar lidar as standard equipment. Ultra Cruise will also be equipped with a Cepton lidar and imaging radar next year.
The Volvo Concept Recharge offers a design preview of the upcoming XC90 replacement model for 2022, which will be equipped with standard Luminar lidar. Image source: Volvo
Of course, adding all these sensors and functions will require more computing power than the current generation of cars. Some of the models GM will launch in 2022, likely including the Cadillac Lyriq, will take advantage of Qualcomm's new Snapdragon Ride computing platform, which offers more than 700 trillion operations per second (TOPS) in a multi-chip configuration. Nvidia will also gain a larger market share in drive-assisted computing with its new Orin System-of-a-Chip (SoC). One Orin chip system can reach a computing speed of 250 TOPS, the aforementioned Volvo and Polestar will have a dual Orin system, and Xiaopeng Motors will have a four Orin chip system. Mobileye, the current leader in driver assistance computing, is not behind in launching its EyeQ5 on multiple brands, including BMW, in 2022.
Much of the content of driver assistance systems in 2022 and beyond was actually developed for highly automated driving,eg, such as robotic axes, but these vehicles have been commercialized for a longer period of time. Due to the popularity of AV technology, mainstream driver assistance systems will become more perfect.
The author of this article is a Forbes contributor, and the content of this article represents only the author's own views.