Russell's parents were “a little skeptical" of the decision, he says. But he had built a research lab in his parents’ garage in Southern California by the time he was 11, so they knew they had a prodigy on their hands.
And Russell himself has no regrets.
“Dropping out is definitely right for some people and can work really well if you're very clear on what you want to do and how you want to go about doing it. And clear about just life goals in general,” says Russell.
“A lot of it is just about timing,” adds Russell. With the speed at which things like hardware and photonics advance, if he had waited until he graduated to build his company, it might have been too late.
During the fellowship is when Russell's work with light went from “theoretical research” to a “real-world business,” the young entrepreneur says.
The technology Luminar is building, lidar sensors, are the industry standard for autonomous cars. There are many other companies working to build lidar sensors, including Velodyne, Quanergy and Innoviz. Companies using Velodyne technology include Ford, Volvo and Mercedes-Benz. Innoviz's technology is being used by Beijing-based automotive electronic products company HiRain and BMW, among others.
But Luminar says it is aiming to make better quality lidar sensors than the rest of the competition. To do this, Luminar builds all of its own components — lasers, receivers, scanning mechanisms and processing electronics — as opposed to using “off-the-shelf” parts, according to the company. Luminar reports that its sensors “see” at 50 times better resolution and have a tenfold range compared with industry competitors (thanks to the use of different materials and light frequencies).
So far, Luminar — which is headquartered in Palo Alto, California, with manufacturing in Orlando, Florida — has raised more than $36 million in funding at an undisclosed valuation and has more than 400 employees. The start-up was in stealth mode for the first five years of its existence to keep details of the work it was doing private. It publicly launched in April 2017.
That September, Luminar announced a partnership with the Toyota Research Institute, the research arm of its namesake automaker parent company.
“We moved swiftly and early to adopt the Luminar platform into our fleet, and as a result we’re rapidly advancing our program,” James Kuffner, chief technology officer of the Toyota Research Institute, said in a written statement at the time. “The level of data fidelity and range is unlike anything we’ve seen and is essential to be able to develop and deliver the most advanced automated driving systems.”
In June, Luminar announced investment from and a collaboration with Volvo.
“LiDAR is a key technology for enabling autonomous cars to navigate safely in complex traffic environments and at higher speeds,” said Henrik Green, senior vice president of research and development at Volvo Cars, in a statement at the time. “Our collaboration with Luminar allows us to learn more about its promising technologies and takes Volvo Cars one step further to the highly autonomous cars of the future.” (Indeed, Volvo said it will offer a true self-driving car by 2021.)
A spokesperson for Luminar says the company has two other high-caliber partners, but declines to name them.
Luminar started shipping early iterations of its lidar sensors to its partners over a year ago for the companies to use in self-driving test vehicles. The latest version of its sensor was announced and shipped in April. Luminar declines to disclose the cost of its sensors, as well as company revenue. According to a spokesperson, the “deals with the four major partners are not insignificant.”
However, there is a one very famous critic of the kind of tech that Luminar makes — Elon Musk. Rather than including lidar sensors for navigation, Teslas (which are built with "full self-driving hardware") have a combination of cameras, radar and ultrasonics.
“In my view it is a crutch,” Musk said of lidar sensors during Tesla’s February investor conference call. "Now, perhaps I am wrong, in which case I will look like a fool, but I am quite certain that I am not.”
Russell is aware of Musk's thoughts on lidar technology but is not swayed.
“Well, I would just say, in general, that there are two different types of systems. So if you want a car that works great as an assisted driving system, that requires constant human input and maybe follows a couple lanes on a road, you can actually — you don't need lidar at all for that. You can do something very well with with just cameras and radar. But the challenge comes down to solving that last 1 percent of all these different edge cases,” says Russell, referring to irregular incidents like a child suddenly running into the street.
“Not only do you need lidar, we need way better lidar, if we want to be able to solve this problem at all,” says Russell.
Musk’s notion of building autonomous vehicles without lidar sensors is unrealistic, says Russell, because the alternative — computer vision — is not close to being ready for implementation.
“I mean in theory you could actually have an autonomous car that doesn't need any other type of sensors or just a couple cameras like human eyes looking ahead, you know, but the reality is is that we are nowhere close to that from just a computer vision standpoint.”
According to its website, all Tesla vehicles produced in its factory "have the hardware needed for full self-driving capability at a safety level substantially greater than that of a human driver," but additional software (and regulatory approval) is required before the cars will be able to drive autonomously, Tesla says on its website. The electric car company did not respond to an email from CNBC Make It seeking comment on Russell's remarks.
Much surrounding the future of self-driving cars still remains to be seen.
Indeed, launching a company in a nascent industry was hard, says Russell. In the early days, it could be a “big challenge” to get people to take him seriously — not just because of his young age, but because he was pitching a product for an industry that wouldn't fully exist for several years.
But Russell earned respect by aligning himself with the right people and proving his idea worked, he says.
“The reality is that you just have to find the right supporters, the right people that ultimately believe in your vision and run with that,” Russell tells CNBC Make It.
“You have to start somewhere with things, and it's just about gaining momentum on that. You have to build the right proof of concept, prototypes, other things — in hardware or software or whatever type of technology you’re creating — and just go from there and ultimately build up that sense of credibility.”
Being a Thiel fellow gave Russell credibility, he says, as does his co-founder, Jason Eichenholz, who has worked in the photonics industry for 25 years and is the inventor of 10 different patents in the space.
Having purpose is also important, says Russell. His is a desire increase safety on the road by implementing his lidar sensor technology at scale.
"That's what inspires me to be able to see this through,” Russell tells CNBC Make it. “It's important to have some type of higher driver or passion or something to ensure that you keep going, because it's always going to be a little bit of a roller coaster.”
In fact, Russell turns to the words of his critic: “To quote Elon, he's said that being an entrepreneur is like chewing glass while staring into an abyss,” says Russell, “which I'd agree with in a lot of respects, but that's at the difficult times. At the great times it can be rewarding like nothing else.
“It's a lot of blood, sweat and tears, a lot of all-nighters," says Russell, who often works 100 hours a week. "But, yeah, it's worked out for us in the end.”
— Video by