Pony.ai raises $112 million Series A for autonomous cars

Baidu and Google self-driving alum team up to transform transportation

We've solved for our ability to work remotely - thanks to computers and smartphones. Now we're solving for our ability to shorten our commute from home to work -- thanks to autonomous cars. As you can imagine, this will have a significant effect on the migration trend outside of urban areas, which is already underway. 

While this is a massive trend in urban and suburban development I'm fascinated with, we're not there yet as today we're still seeing the formation of self-driving, autonomous cars and the players in the race to create this new world for us.

Pony.ai is one such newcomer. The Fremont-based startup, with offices in China, opened its doors in December 2016. Since then it's raised a $15 million seed round, and 13 months later, it's secured $112 million in a Series A from Morningside Ventures and Legend Capital. Other investors in the round include Sequoia China, IDG Capital, Hongtai Capital, Legend Star, Puhua Capital, Polaris Capital, DCM Ventures, Comcast Ventures and Silicon Valley Future Capital. 

The startup was founded by two veterans in the autonomous space [if there's such a thing in this very nascent industry] - James Peng, CEO, who was previously Chief Architect at Baidu, where he led the self-driving division. The other is Tiancheng Lou, who worked on autonomous vehicles at both Baidu and Google X.  

In the last year, the startup used its initial funds to create a prototype and test its software when it received its California testing permit in June 2017. Some accomplishments its self-driving vehicle has been able to perform to date are: driving through a four-way stop and driving through an unprotected left turn, meaning there is oncoming traffic and the car may have to yield before making the left turn, said Peng, in an interview with me.

Now after miles of testing on the road and some successful proof points, Peng is using the new funds to broaden its team of 60 and to create 50 cars this year and 100 next year to test some more and also put vehicles, such as shuttles at universities or corporate campuses. Pony.ai also plans to operate a fleet in the Southern Chinese city of Guangzhou. "We need to understand how to manage that fleet," said Peng. 

The company is currently in talks with automakers, though Peng couldn't disclose. 

The cost to produce one of these autonomous vehicles is between $100,000 and $200,000, he said, depending on the sensors in the car. At the moment, Pony.ai will not likely license its software but collaborate with the automakers. In other words, both will contribute to building these next-generation vehicles.

Peng doesn't see people owning self-driving cars, partly due to cost, though he sees costs for parts coming down rapidly, referring to Velodyne cutting its sensor in half this year. He believes, as many do, that self-driving vehicles will be used for rental or on-demand driving. The company is considering making its own on-demand service to compete with Uber and Lyft, but is also considering partnering with these services. This decision will come after the team learns how to operate and manage their 50-100 car fleet. 

Meanwhile, automakers are already anticipating to have their fleets out on the road next year.

Crowded market

GM is already anticipating that it will have its Cruise AV commercialized in 2019. The lack of steering wheel and pedals reminds me of the lack of phone jacks and phones on our walls. My youngest child doesn't know what a wired phone looks like. He may forget what a steering wheel looks like altogether.

As capital pours heavily into this space, the question is: How do you determine which company to back?

For Michael Yang, VC at Comcast Ventures, the reason to back Pony was clear.

The startup's focus on the China market, China's favorable regulatory environment [allowing more autonomous vehicles on the road than in the US], Guanzhou's financial sponsorship [the city gave Pony.ai a grant and is encouraging the startup to get 100 vehicles on the road, yesterday], and the founders' pedigree at Baidu and Google, were reasons behind Comcast's investment.

"We think of this investment as a big China opportunity," said Yang, explaining how Comcast's assets could also be valuable to Pony.ai. "NBC/Universal [owned by Comcast] is building a theme park in Beijing. From a corporate perspective, we can help Pony.ai with marketing direct to consumer when they launch their ride-sharing service in Guanzhou."

But the addressable market for ride-sharing autonomous vehicle providers isn't the only big opportunity in this space, Yang said. The reason why startups are piling in to create LIDAR sensors is because every autonomous vehicles needs them and it's not a secret that those produced by Velodyne are too expensive. There's also opportunities to design the cars. There's no standard in how the interior of an autonomous car looks today. "Does it look like a living room or a space pod?" Yang asked. There's opportunity everywhere. But his bets would largely be in the software and ride-sharing services buckets. 

As for what Yang would like to see from his investment in two years? "We want to know the cars works safely, and consistently. Then we want to better understand consumer behavior. How much they were willing to go in autonomous vehicles, by themselves or fully autonomous, with their families, etc. We'll know this more in China before we see it in the US."