Enlarge / An experimental self-driving car from Cruise Automation is equipped with lidar sensors.
The 2020 Consumer Electronics Show was full of companies selling Lidar. In short for light radar (yes, really), this powerful sensor type creates a three-dimensional point cloud of its surroundings. Experts and industry experts who are not called Elon Musk see this as a key technology for self-driving cars. There are dozens of companies developing lidar technology, and each company insists that its sensor be a cut above the rest.
But while every lidar in the CES halls is above average, things are starting to look different in the real world. At least one market segment – customized robots for warehouses, mines and other industrial locations – is beginning to buy lidar sensors on a large scale. Another segment – low-end lidars for driver assistance systems – will become a large market in the next few years.
For this piece, I asked both LIDAR company officials and independent experts to explain the status of the LIDAR market to me. They told me that Velodyne – the company that invented the modern three-dimensional lidar more than a decade ago – continues to dominate the industry.
But Velodyne faces growing competition from newer companies. One company in particular – Ouster – has started to supply affordable alternatives to Velodyne's flagship products. Although these products may not quite match Velodyne's performance, they are good enough and cheap enough to seriously jeopardize Velodyne's dominance.
The big fights in the lidar market are still in the future. So far, a lot of lidar has been sold outside of the automotive industry, but experts assume that automobile manufacturers are the biggest customers for lidar. Over the next few years, a number of automakers will be buying lidar in bulk for the first time – thousands of inexpensive lidar sensors to improve their driver assistance systems (ADAS). A number of LIDAR companies are positioning themselves to win these deals, and some are partnering with traditional tier 1 automotive suppliers to improve their opportunities.
The biggest price in the industry could be to provide more powerful lidar sensors for fully automated vehicles. Many companies want to serve this market, but these sales are still far away because the fully self-driving technology is not yet ready for prime time.
The rotating mechanical lidar still dominates in robotics
Enlarge / An Ouster sensor on a Postmates robot.
Velodyne invented the modern three-dimensional lidar industry over a decade ago and remains the industry leader. His original lidar had 64 vertically stacked lasers that were rotated to give a 360-degree view of the surroundings. Velodyne has since introduced variations on this basic design, including a high-end 128 laser design and cheaper models with 32 or 16 lasers.
Velodyne's premium lidar units traditionally cost $ 75,000 or more. Until recently, people who needed high-performance lidar had no choice but to pay. In recent years, a startup named Ouster has started to provide Velodyne with the much-needed competition.
The classic Velodyne Lidar contained 64 individually packaged lasers, each of which was combined with a single light sensor to detect flashes. This complex design contributed to the high price of the units. In contrast, Ouster uses semiconductor technology to package 64 lasers on a single chip, with 64 detectors packaged on a second chip. This lowers the cost of Ouster's lidar in much the same way that the microchip revolution in the 1970s made it possible to develop cheap PCs.
The result: in 2018, Ouster was able to offer its first lidar sensor, a 64-laser device called OS-1, for just $ 12,000. That was dramatically cheaper than Velodyne calculated for its 64 laser models. Since then, Ouster has expanded its product line. The company's products now range from a low-end device with 16 lasers for $ 3,500 to a long-range device with 128 lasers for $ 24,000.
"The fact that you can get a 64-channel spinning lidar for $ 12,000 was outrageous," said John Williams. "When we realized that there was, we were surprised."
Williams is the chief technology officer at Kudan, a company that develops software that allows robots to track their own location (a problem known in the robotics world as SLAM). Kudan customers build robotic products such as low-speed shuttle buses, storage trucks, autonomous forklifts, etc. Kudan's software must work with a variety of sensors, including cameras and lidar. So Williams has a good overview of which lidar sensors robotics companies use in the real world.
Williams told Ars that most of his customers are still buying Velodyne Lidars. "Ouster is the climber when it comes to disrupting the market," said Williams. "In terms of accuracy and noise, Ouster sensors are" not quite as good as Velodyne, "said Williams. But they are dramatically cheaper, and that is important to many Kudan customers. Customers who would not have thought of using Lidar to Velodyne- Buying prices considered after Ouster launched a 64-laser device for $ 12,000.
When I asked Velodyne's managing director, Rick Tewell, about Velodyne's prices, he insisted that they were "extremely competitive". But he didn't give me exact numbers and didn't seriously deny that the high-quality lidars from Ouster were cheaper than comparable offers from Velodyne.
Velodyne is also exposed to competition from some Chinese companies, which Velodyne rejects as a copycat company. Like Ouster, these companies offer sensors similar to Velodyne at low prices. According to Velodyne, in contrast to Ouster, these companies only copied the design of Velodyne. In fact, Velodyne sued two of these providers last August – RoboSense and Hesai – for patent infringement.
In recent years there has been no shortage of hype regarding the potential of solid-state lidar – lidar that is fixed and does not rotate 360 degrees. There are dozens of companies working on lidar sensors, and most of them are solid-state designs. Nevertheless, they seem to find little use in the real world – at least not in the commercial robot market. Williams said to Ars, "We have not yet encountered solid state in the wild."
This is due in part to the fact that some solid state companies have not yet started delivering products to the general public. This is partly due to the fact that some LIDAR companies concentrate on large-volume sales to automobile manufacturers and not on one-off sales to smaller robotics companies.
It's also because lidar spinning has some unique advantages. The most obvious is that a single rotating lidar sensor offers 360 degree coverage around a vehicle. To achieve a similar cover with a fixed lidar system, several sensors must be distributed in the vehicle. That means higher costs and higher power consumption.