Velodyne re-asserted its dominance of the lidar market today by announcing a product with 128 laser beams, twice as many as its previous top-of-the-line model.
“The VLS-128 is the best LiDAR sensor on the planet, delivering the most advanced real-time 3D vision for safe driving,” Mike Jellen, the president of Velodyne LiDAR, said in a statement.
The announcement, which had been widely anticipated, leaves out the one detail that everyone most wants to know: the price. The company’s previous top-of-the-line product originally sold for more than US $70,000.
Today, though, a host of rival companies are on the scene, some of them promising solid-state products that cost less because they have no moving parts. Some of those lidars are priced in the hundreds of dollars, but as advocates of moving parts like to point out, solid-state technololgy offers only very low resolution.
Velodyne argues that its new lidar’s resolution is high enough to identify objects without any input from other sensors, a stratagem known as sensor fusion. If so, that would relegate the cameras, radars, and other sensors to the role of mere backup devices.
Lidar provides a three-dimensional view of the world by scanning laser beams back and forth or—as in this case—round and round, for 360 degrees of coverage. Each beam scans in two dimensions, and a stack of such beams can thus represent three. The more laser beams you have, the better the 3D resolution.
The company’s statement claims a “whopping” 10-fold improvement in resolution, which doesn’t quite square with what Anand Gopalan, chief technology officer for Velodyne LiDAR, told IEEE Spectrum last week. He said doubling the beam count roughly triples the resolution, thanks in part to the finer spacing of the beams. But the seeming discrepancy may simply reflect additional gains in resolution that Velodyne has achieved from faster scanning and better algorithms for processing the data.
Shown here are Velodyne's HDL-64E LiDAR sensor (left) and the company's VLS-128 sensor (right).Photo: Velodyne
How the product compares with the lidar that Waymo now builds for itself is not clear. Waymo had originally based its self-driving technology on the Velodyne 64-beam model, but it wasn’t satisfied with it and therefore built itself three in-house models—for long, medium, and short ranges.
“We think short, medium, and long range [lidars] are not necessary,” Gopalan said. But, he added, Velodyne had listened to customers’ suggestions for getting around key problems, called corner cases, that the old lidar couldn’t handle.
“One problem is seen in high-speed highway driving, where you want to see far out but with enough resolution to quickly classify things,” he said. “With a higher number of beams you get really higher effective resolution—without compromising the frame rate.”
Velodyne hasn’t released the names of its customers, but Gopalan said that “with one or two exceptions” it was working with most of the industry and that almost every customer had expressed an interest in the 128-beamer.
“You’ll see it in experimental cars by early 2018,” he said. The new lidar is being built by entirely automated processes in the company’s San Jose, Calif. factory, where production is ramping up. As the volume rises, unit costs will fall, but the price actually charged is of course entirely at Velodyne’s discretion.
How much higher can the laser beam count go? A similar question was asked back in 1975, when the television comedy show “Saturday Night Live” ran a mock commercial parodying the seemingly outlandish two-blade razor Gillette had recently introduced. The commercial was for the “Triple-Trac razor.”
Gillette’s engineers were miffed because they were already working on a three-blader, though it came out only in 1998. Today, the top-of-the-line multiblade razor boasts seven blades.