It’s not a secret that Google has ambitions to become a hardware company.
The tech giant unveiled a suite of new hardware products in the fall, like its first smartphone, the Google Pixel, and connected speaker, the Google Home.
And now the company is taking the same approach with self-driving cars. Waymo, the self-driving car company run under Google’s parent company Alphabet, announced this week that it will manufacture an entire suite of sensors in-house.
That means all the sensors that self-driving cars rely on — radar, lidar, and cameras — are all being designed and manufactured at Google where they will be integrated to work together as an entire self-driving hardware suite.
In many ways, it’s a brilliant move by Waymo to reclaim its position as a leader in the self-driving car space, though some lingering questions remain.
The long road to market
There’s never really been a question as to whether Waymo was a competitor when it came to self-driving car software. Its cars have driven nearly 2.5 million miles autonomously — the equivalent of 300 years of human driving experience, according to Waymo.
Where Waymo has run into problems is figuring out a way to turn its self-driving car moonshot into an actual, marketable product.
Waymo originally had a grandiose vision to produce a fully driverless car without a steering wheel or pedals. Although some engineers on the project pushed Waymo to pursue partial autonomy like Tesla to start bringing the tech to market, Google co-founder Larry Page has reportedly been opposed to taking that approach, Bloomberg reported.
But slow progress on the self-driving car project prompted several executive departures. Chris Urmson, the CTO of Google’s self-driving car unit before it became Waymo, left in August. He was one of several executive departures — Anthony Levandowski, a co-founder of the Google self-driving car project, left to start self-driving truck startup Otto, which is now owned by Uber.
During his time at Google, Urmson lobbied the Senate to create regulations that would allow self-driving cars without a steering wheel or pedals to drive on public roads.
While all of this was going on, competition continued to mount, most notably from Uber. Uber launched a pilot program for its self-driving cars in Pittsburgh in September, and has since launched its second one in Arizona.
But Waymo recently made a series of smart moves to avoid getting eclipsed by the competition. It decided to keep driver controls in its cars and expanded its partnership with Fiat Chrysler. Waymo will begin testing self-driving Chrysler minivans in Arizona and California by the end of January.
Waymo is also in talks with Honda about creating a fleet of self-driving cars.
Couple that with Waymo’s recent announcement about building self-driving tech entirely in house, you can start to see how Alphabet is becoming a serious competitor in the autonomous space again.
Waymo says designing all of its hardware in-house has seriously slashed costs because all of the sensors are designed to work together (as opposed to combining different off-the-shelf parts to work together). So far, Waymo has reduced the price of lidar by over 90% to roughly $7,500, and said the price will continue to drop as the technology scales.
Essentially, Waymo has created a self-driving ecosystem — hardware and software — that it can integrate into cars easily through partnerships with automakers. After a summer of executive departures, Waymo has made serious strides in putting together a clear path to market.
But we will still need to wait and see if Waymo’s approach will play out accordingly. Alphabet as a whole may be making moves to become more of a hardware company, but hardware has not typically been its strength.
We have yet to see how well Waymo’s hardware ecosystem performs with its software, as the Chrysler minivans will be the first vehicles outfitted with the system. Waymo also has yet to launch a public demo of its self-driving cars, something BMW, Volvo, and Uber will all do in 2017.
Still, it looks like Waymo has a much better plan to get to market than it did just two months ago.
This is an opinion column. The thoughts expressed are those of the author.