Big data on wheels: Google vs. Uber in the driverless revolution

The excitement is palpable and the rumors are flying. Google looks set to challenge Uber for the position of the top tech company to do two things: advance the use of driverless car technology and offer a ride-sharing service. If the rumors are true, it’s not unreasonable to ask: is it even possible for a company to overtake Uber in this respect? If anyone could, it’s Google.

First, the buzz. Back in early 2016, Google partnered with Fiat Chrysler and began exploring ways to incorporate driverless tech with FC’s vehicles. Now, word on the street is that Google’s parent company, Alphabet, will collaborate with Chrysler to offer a ride-sharing service in 2017. The service will use Chrysler minivans outfitted with Google technology to render them semi-autonomous. Google’s fully autonomous rides are going to be released under a new Alphabet brand, Waymo. Google and Chrysler are two titans of their respective industries coming together to aim at toppling the number one ride-sharing service in the world: Uber.

But in terms of ride-sharing with driverless cars, Uber is already beating Google and Chrysler off the starting line. In Pittsburgh, you might see self-driving Uber cars picking up volunteers and dropping them off. In 2015, Uber set up a headquarters, the Advanced Technologies Center, to begin testing out driverless ride-sharing. The initial vehicles are from Ford, but Uber is unveiling its partnership with Volvo. In San Francisco, as I write this, the Volvo XC90 sport utility vehicle is picking up some people who hail an Uber X. The Uber app tells them (or warns them) their ride will be driverless, at which point they’re free to cancel. In Pittsburgh and San Francisco, the driverless cars are accompanied by an engineer, who makes sure everything goes well.

Uber’s quick moves are calculated to get out ahead of competitors, but up until very recently, there was no indication that Google would be in on ride-sharing. Driverless ride-sharing is a service completely dependent on the quality of technology. Since that’s the case, the ultimate winner here will be the company that masters the convergence of sensors, big data, and AI in the form of a driverless car.

The success of any driverless interface depends on how well the car’s computer processes the data that it feeds into a centralized system. In short, the Google vs. Uber battle would be a battle of Artificial Intelligences. Lidar sensors, GPS, altimeters, gyroscopes, and tachymeters mounted on the car all transmit massive amounts of data to the onboard computer’s Deep Learning program, which uses this input to alter algorithms and command the car’s movements in real-time. This has to happen with incredible speed and accuracy.

Uber’s experience in the AI field is nowhere near as extensive as Google’s–at least in terms of practical application. Google’s confidence with AI informs all of the company’s market-facing decisions now. RankBrain, which came from the Deep Learning program that initially defeated Go champion Lee Sedol, plays a part in all of Google’s search algorithms. Google’s hardware releases incorporate AI in some way, whether you’re talking the Pixel phone, which includes AI in the form of Assistant, or Google Home, a smart speaker that uses Assistant to help you find music and control your house’s smart devices, or Daydream View, which works in tandem with Pixel to navigate a variety of VR experiences, including Google’s Street View.

In general, when it comes to the world of tech and AI, compared to Google Uber is just a brash upstart. But the ride-sharing service’s speedy and sudden uptake with driverless tech is demonstrative of why the startup has such an immense valuation.

Uber is fiercely determined to be at the forefront of tech and transport. The Silicon Valley startup has found practical ways to apply innovations in APIs, navigation, and AI with an unprecedented level of agility. Uber’s partnerships with Ford and Volvo indicate just how savvy the startup is at impressing longstanding companies. These companies are aiding Uber in a Big Data coup d’etat. Google is all about big data and has been working on driverless cars since 2009. Yet Uber (with the help of automakers) is beating Google at putting big data on wheels.

There’s a caveat. Immediately after their pilot run, two of Uber’s driverless Volvos in San Francisco were caught running red lights. Uber says these cars were not part of the actual pilot program and didn’t have riders in them. But the cars were still driverless. It looks like Uber is trying to move too fast. If the AI in these vehicles can’t recognize a light is red, how is it intelligent enough to navigate the busy streets of any of our cities? Would a self-driving Google/Waymo car be this irresponsible?

Whatever the case, in the long term, the best self-driving cars will have the best AI, capable of processing navigational data quickly and accurately. In order for this to really work and be worth it, the AI has got to be smarter than the billions of humans who kill 1.2 million people a year with their cars.