Sixteen years ago, Velodyne LiDAR founder David Hall faced a few challenges. His first company, Velodyne Acoustics was increasingly losing its domestic supply chain to China. Hall realized that, eventually, the entire manufacturing process might relocate. This would lave him with an empty factory space in Morgan Hill, CA and a 60-mile commute that might take 2 hours in traffic. As Hall explained to Robert Scoble of Rackspace Studios, the current efforts of Velodyne LiDAR evolved from addressing these challenges.ⁱ

Disturbed by the time he wasted commuting, Hall wanted to invent a way for his car to drive itself. Plus, he needed to find a new product he could build in his factory. “So, I looked around in the robot space and I got into the [DARPA] Grand Challenge,” says Hall. “I really honed in on LiDAR to fill the need for self-driving cars.”

The first Defense Advanced Research Projects Agency Grand Challenge was an autonomous vehicle race held in March 2004 in the Mojave Desert along a 150-mile route adjacent to Interstate 15. The goal was to encourage development of fully autonomous ground vehicles. But since no robot vehicle completed the course, no winner was declared nor was any money awarded. The $1 million prize was doubled to $2 million for a 2^nd race to be held in 2005.

This 2^nd Grand Challenge piqued the interest of Hall, who had the idea to enter in February 2005. By September, he’d installed his first prototype, which eventually became Velodyne LiDAR’s breakout sensor, the HDL-64E. “I was still writing software as we were driving to Las Vegas,” he recalls. Along with brother and co-founder Bruce Hall, team DAD made a strong showing among the original 40 entrants, but was not among the 5 who completed the entire course. Still, David Hall had found his next product, and the experience led him to forgo future competitions and instead commercialize the sensor. By the time the 2007 DARPA Urban Challenge arrived, Velodyne’s HDL-64E sensor was mounted atop five of the six vehicles that finished.ⁱⁱ

Only 5 years later, Google unveiled a secret project started in 2009 involving driver-less cars. In 2011, a fleet of robotic Toyota Priuses took openly to the streets with a human monitoring from behind the wheel. Soon, technology buffs around the San Francisco Bay Area were spotting the adapted vehicles, which included a roof rack assembly with a spinning HDL-64E.ⁱⁱⁱ By 2016, Google had renamed the project as Waymo (i.e. a new way forward in mobility) and claimed over 2 million miles driven on public roads across the United States, all with a Velodyne sensor mounted on top.

Meanwhile, Velodyne released their second sensor, the HDL-32E in 2010. Shaped like a coffee can, the 32E came in about a quarter the size and 1/12^th the weight of its predecessor. “We took a smaller direction,” said Hall, explaining the conceptual development behind the HDL-32E, “attempting to get into what we thought would be low speed robotics like robot lawn mowers.” Soon, the sensor had met and far exceeded these initial ambitions. By 2013, the HDL-32E was being mounted, four across the roof, on experimental Ford Fusion Hybrid research vehicles. In a 2013 demonstration video, Ford Research and Innovation Director Randy Visintainer touted Velodyne’s sensors for “providing complete situation awareness… with HD quality imagery.”^iv

In 2016, Velodyne took another significant step forward with the release of three new sensors, the Puck VLP-16, the Puck Lite, and the Puck Hi-Res. The goal of this third generation sensor was to not only shrink the form factor, but as Hall explained, move the cost into “the range of affordability—so it will be cheaper than your [DSLR] camera but more expensive than your iPhone.” Already, the many advantages of this smaller and cheaper yet powerful device are being seen in countless applications, as Pucks are being mounted on drone aircraft, industrial vehicles, and driver-less cars. While retaining key features of Velodyne’s LiDAR breakthroughs, such as 360° real-time 3D distance and calibrated reflectivity measurements, the Puck also relocates the rotating mechanisms inside a protective housing. This latest innovation allows the highly resilient Puck to operate in challenging environments.

Looking back, it becomes evident that each of the first three generations of Velodyne’s wildly successful sensors share one unique attribute. They rotate. “If they have the spinning one, that’s mine,” says Hall. “I’m the only spinning one.” This simple yet revolutionary feature is also intriguingly related to a race, but this time not involving cars.

“The whole idea of this is I used to race bicycles,” explains Hall. “And they had a camera that would film the finish line. The film would go across a slot and… form a linear picture of the bicyclists running through. So I thought, that’s a great idea. I gotta use that for something.” That something became Hall’s mechanical-electrical hybrid design, used in all three generations, where 64, 32, or 16 lasers obtain distance data linearly while being mechanically swept across the landscape to generate a 360° real-time image. This rotating design offered Hall a perfect alternative to a total array of lasers—which could be built but would be complicated and expensive. The result was a more cost-effective use of fewer lasers and detectors. “This was a way for recycling [the lasers],” says Hall. “So you get a lot of use out of them.”

Considering the vast potential for LiDAR, Hall references a variety of success stories beyond driver-less cars including countless 3D mapping projects and guiding Caterpillar self-driving mining vehicles in Australia. “These big three-story, thousand ton trucks are rumbling around with no drivers in them,” explains Hall. “And have been for a while.” Given the potential, the future of Velodyne LiDAR looms brightly while a massive expansion is underfoot, spurred by $75 million investments from both Ford and Baidu in 2016.

John Eggert, Senior Sales and Marketing Manager at Velodyne jokes that he’s an old sage, since he’s been around for a year and a half. “What we’ve seen in the last few years is a trend toward using LiDAR,” says Eggert. “The early adopters were correct. They’ve been validated.” Just in the past two years, the number of original equipment manufactures in the auto industry who are pursuing high definition LiDAR for driver-less vehicles has jumped from one to ten.^v

A glimpse of the future came in a significant announcement by Velodyne last month. Their latest breakthrough is a solid state LiDAR sensor which can be mass-produced for around $50.

With more innovations looming on the horizon, 2017 will surely be an exciting time for Velodyne LiDAR!

sources:

i “Velodyne: On Sound, LiDAR, and Marine” Rackspace Studios, SFO. YouTube. May 8, 2014

ii “Robots, start your engines” Tom Abate. SFGate. November 4, 2007

iii “Google Self Driving Car Spotted!” AdMauseum. YouTube. June 18, 2012

iv “Ford Fusion Hybrid Research Vehicle Overview With LiDAR Technology” @VegasBiLL@24k. YouTube. December 15, 2013

v “Reliance on LiDAR – 2 Year Snapshot” John Eggert. Velodyne LiDAR. December 5, 2016

Photo credits: “DARPA Urban Challenge Archive” Defense Advanced Research Projects Agency. Accessed December 28, 2016