Waymo's head of lidar explains how the advanced laser tech went from science fiction to being a vital tool for self-driving cars
- Simon Verghese is head of lidar at Waymo, Alphabet's self-driving vehicle startup.
- Verghese has a doctorate from Berkeley and a background in physics, and he was working at an MIT national lab when Waymo convinced him to return to the Bay Area after more than 20 years on the East coast.
- Verghese spends his time developing new laser radars to compliment the evolving Waymo suite of hardware and software technologies.
- "The software problem is so difficult that we want to give our team the best hardware," he said.
For an ongoing series, Business Insider has been talking with Waymo employees from different parts of the company to learn more about their work. What we discovered were some of the coolest jobs at Alphabet, Waymo's parent company. This is the latest profile in the series. To read the others, click here. For a brief history of Waymo, click here.
When Simon Verghese moved east for a postdoctoral spot at MIT, the UC Berkeley-trained physicist figured on being back in California before too long.
"I thought it would be two years," he said in an interview with Business Insider.
Two decades and three children later, Verghese was still at MIT's prestigious Lincoln Laboratory, working with NASA, the Department of Defense, and the US Navy. Few offers could have lured him away from his work on various sensor systems for aerospace applications.
The one that did the trick came in 2016 from Waymo. The Alphabet autonomous mobility company that started life as Google's self-driving-car project wanted the professor to lead development of its lidar, the laser-scanning technology that enables vehicles to navigate complex environments as well, or better than, human drivers. The tech that just might be the biggest challenge in self-driving cars
Verghese already had the expertise; lidar had been his thing for almost a decade. The technology has been around for 50 years, but has taken on new importance in the past decade thanks not just to self-driving cars, but also for relief missions to survey damage after natural disasters, such as the 2010 7.0-magnitude earthquake that devastated Haiti. (Verghese and his MIT team supported flyovers of the region that monitored damage with lidar units, which have advanced from being enormous systems to becoming small enough to put on planes)
What convinced Verghese to make the move was Waymo itself.
"Half the reason I left the East Coast was to experience Waymo culture," he said. "It celebrates small teams taking initiative. It's not a top-down, command-and-control-type place. But everybody still manages to stay aligned and move quickly."
He cited Waymo's efforts to move from to the fourth generation of its lidar system from the third generation. "At my old job, that would have taken many years." Now, Waymo has already moved on to the fifth generation.
"I was pretty happy in Boston," he recalled. But Waymo tempted him with an obvious enticement: "Why don't you just come out and take a ride in one of our cars?"
Verghese was familiar with what Waymo had been up to. He knew about the DARPA Grand and Urban Challenges, which kicked off the competition to develop autonomous vehicles in the early 2000s. But by 2015, he was "amazed at how far it had come."
The decision to pull up stakes was easy. "In six months, I was there," he said.
Waymo has created what it calls a "driver" — a cluster of hardware and software that functions as a robot chauffeur. The company's goal is for this driver to be able to operate any form of road-going vehicle. Thus far, Waymo has developed a ride-hailing service, Waymo One, that launched in Arizona in 2018; and a freight service, Waymo Via. (Both continue to retain human drivers, for safety reasons.)
Over the past decade, Waymo has invested heavily in lidar, as have other self-driving startups, several of which have actually acquired lidar companies. This contrasts with Tesla, which has committed to cameras, attempting to avoid the expense of lidars.
"The intention is for this to be a core technology for Waymo," Verghese said. "The software problem is so difficult that we want to give our team the best hardware."
He added that he feels confident that between five and 10 lidars should be an important aspect of any safe, viable self-driving system. Visual systems, while useful, are "nowhere accurate enough for the extreme classification accuracy we need," he said. (Waymo also uses cameras and radars as part of its suite of hardware technologies.)
At a technical level, lidar units send pulses of light at targets and use the reflected results to create a model of the world. It adds an extra sense that a human doesn't have: the ability to see in the dark.
"I can walk from my bedroom to the bathroom," Verghese said, "and kind of feel my way around. But if I see a dark smudge, I'm not sure if it's my cat. If lidar is reaching out there, it knows that there's something that's four inches high and a foot long."
A science superstar who has embraced a new challenge
The longtime East Coaster didn't take too long to get back into a California lifestyle. Dating back to his grad-school days, he has been an enthusiastic ultimate Frisbee competitor; he now plays on a team that won a national championship for his age group. Before the pandemic sent him and his coworkers home, his typical day at Waymo started with an early, head-clearing bike ride. Then it was straight to work (he could shower at the Googleplex), a quick breakfast, and meetings.
"Lidar has different functional teams," he explained. "A big part of my job is making sure we have things prioritized."
A key concern is making sure that the Waymo driver has the lidars it actually required, based on its real-time experience with vehicles in the field. In effect, Waymo is inventing its own systems as it evolves. "We're developing sensors within a vertically-integrated company," he said. "We have to discover what kind of lidar we want." And then keep making it better and better, generation after generation.
Verghese has been a science superstar from a young age. He attended a special state boarding school while growing up in North Carolina, where he focused on physics, and his career has taken him to such professional heights that he really has nothing to prove to anybody.
But he reserves some exceptionally flattering comments for his fellow Waymnonauts, as well as the opportunity the company is chasing.
"I've been out there long enough that I'm most interested in tackling impactful problems with smart people," he said.
"This one is really up there. It's a chance to be at the beginning of a field that comes across many technologies at once, and that could deliver real human benefit."