In the sixth grade, Austin Russell turned a Nintendo gaming handset into a mobile phone. At 15, he built a holographic keyboard. By 17, he'd filed for a patent.
Now at 22, he's running a startup at the heart of Silicon Valley's latest technology mania.
As founder and chief executive officer of Luminar Technologies, Russell and his team are building lidar -Â a hyper-accurate laser sensing technology crucial for self-driving cars.
Google parent Alphabet is suing Uber Technologies for allegedly stealing lidar designs, while startups Velodyne Lidar and Quanergy Systems have raised at least $US150 million ($196 million) apiece from giants like Ford Motor, Baidu, Daimler and Samsung Electronics.
Russell has raised a similar amount, according to people familiar with Luminar's finances. The company, founded in 2012, had sought a valuation above $US1 billion when it was raising money last year, one of the people said.
It's unclear who invested - Luminar is in "stealth" mode, meaning it hasn't announced itself to the world yet. A spokeswoman declined to comment, as did Russell's father Michael, a commercial real estate veteran who serves as chief financial officer. A message sent to Austin Russell through his LinkedIn profile was answered by his assistant, who declined to comment.
Peter Thiel awarded Russell a $US100,000 Thiel Fellowship when he was 17, letting him quit Stanford University.
The billionaire venture capitalist is a regular visitor to a sprawling Portola Valley ranch, 65 kilometres south of San Francisco, where Luminar tinkers and tests lidar systems while employees and guests crash on the couch, according to someone who has been there.
Car companies, including BMW and General Motors, have also dropped by.
That a relatively unknown college dropout of barely drinking age can raise millions of dollars shows the appetite for lidar.
"It's a gold rush and we're selling pickaxes," said Velodyne President Mike Jellen, who graduated college years before Russell was born.
Several car companies want autonomous vehicles on the road by 2020 or 2021, which means they're starting to order lots of lidar systems. Velodyne expects to ship 12,000 units this year, 80,000 in 2018 and 1.7 million by 2022.
New Silicon Valley era
Luminar's rise also says a lot about Silicon Valley's past and present. It's still the place where prodigies can find generous backers for audacious plans. The ideas used to be mobile apps or web software.
Now, it's increasingly technology that interacts with the physical world - cars, robots, drones and software for automation. Russell is part of this new era.
In January, in the up-scale Nob Hill section of San Francisco, a gangling Russell attended a party for 1517 Fund, a VC firm partly backed by Thiel. Towering above the crowd, he lingered in the corner near the entrance, speaking in a booming voice, and avoiding eye contact with a reporter.
He was mostly immersed in his phone, which he showed occasionally to a small group gathered close to him, while more than 100 up-and-coming entrepreneurs and older mentor types chomped pizza and sipped beer.
Some of Luminar's money has been used to buy a small fleet of Tesla Model S electric cars, which it uses for testing, said one of the people who has visited. It's also funding research and development to solve challenges that have plagued the nascent lidar market.
A top-of-the-range lidar from Velodyne sells for more than $US50,000. It offers cheaper lidar, which generates lower-definition 3-D images, for about $US8000, while Quanergy has a product that sells for some $US4000. Autonomous cars often require two or more lidar sensors, so having a capable system can get expensive.
Russell is trying to develop a lidar priced significantly less than $US1000, according to people with knowledge of Luminar's planning. Quanergy aims to have one that sells below $US100 in three to four years.
Whereas radar uses radio waves to detect objects, lidar uses laser beams, helping it produce more accurate 3-D images. It's an essential ingredient for autonomous driving because it generates a real-time image of passing and surrounding objects and helps a vehicle accurately locate itself. Satellite navigation systems are only accurate to within about 5 metres - not enough for a driverless future.
In a recent demonstration, the images generated by Luminar's lidar system were higher-definition than those produced by competing equipment made by Velodyne or Quanergy, according to someone who saw the equipment first-hand, but was not allowed to discuss it publicly.
Another version generated even sharper images, but the information was processed with a slight delay -- because of a lack of computing power to crunch all the data rather than a problem with the core technology, the person said.
Luminar may have bigger plans. A trademark filing from June 30 described a "vehicle collision avoidance system" with ultrasound sensors and radar apparatus, not just the optical technology used in lidar.
In recent weeks, it posted 19 jobs online, seeking engineers, attorneys, a "Fiber Laser Production Manager" and a vehicle integration specialist.
Russell has the ability and drive to make Luminar successful -- as long as he focuses his prodigious brain, according to Tony Jordan, his physics teacher at St. Margaret's Episcopal School in San Juan Capistrano, about 80 kilometres south of Los Angeles.
"The kid's mind is so broad that he literally always had 50 ideas going at one time," Jordan said. "My only thought was if he ever slows down enough to see one idea all the way to completion."
Russell showed early promise, working as a consultant software engineer at age 10, according to a 2014 interview posted on YouTube. A year later, when his parents refused to buy him a mobile phone, he hacked a Nintendo DS portable gaming console into a handset. In 2013, he filed a patent for a "three-dimensional imager and projection device."
His parents' garage in Newport Beach, California, was the hub for his early inventions, said Jordan. "They had a hard time ever parking their cars in the garage because he had absconded with the Ping-Pong table and made that his lab table," said Jordan, "That's where some of his best thinking was done."
A 15-year-old Russell came to a school staff meeting one morning to demonstrate a system that beamed a holographic keyboard onto a table top and typed words. "That blew most people's minds," Jordan said. Other early inventions included a laser that could detect whether a mole was cancerous, and a theoretical framework to charge electrical devices by beaming energy down from satellites.
'Push yourself to the limit'
In June 2013, the Orange County Register named the then 17-year-old Russell a top graduate, noting his interest in photonics -- the use of electromagnetic energy -- and his plans to attend Stanford that autumn.
"You should push yourself to the limit at least, ultimately," Russell told the newspaper. "That's what you have to do if you want to make an impact on the world."
In 2014, Russell led Luminar in a digital health care competition sponsored by Qualcomm. In a video about their entry, the young, mop-haired CEO described his lidar system and another technology called a "real-time hyperspectral camera system" that measures the molecular structure of objects well beyond what a human eye can see.
Such cameras cost tens of thousands of dollars and were the size of a small fridge at the time. Russell said Luminar had built one as small as "a few pennies."
He exhibited a competitive streak in high school, where he led a robotics team to the national championships, Jordan recalled.
"You've never seen someone more upset when they lost, or the robot went down, or the person handling the robot mishandled one of the feats that they had to do to win," said Jordan. "I really enjoyed watching someone who was usually so cool and so on nine different planes that you were never sure he was with you, he also had this single-minded focus and competitiveness."
Russell will need those qualities because the window to make the most of the lidar business may be closing quickly. Current systems are too expensive for production cars, yet bringing prices below $US100 may make it hard to generate significant profit.
"It's going to go the same way as radar has and become commoditised," said Sebastian Thrun, CEO of online learning specialist Udacity and the former head of Google's driverless car project.
"Radar used to be $US60,000 apiece and now it's like $US80 apiece. There's no reason lidar can't cost $US80 apiece."
Bloomberg
0 comments
New User? Sign up