How platforms are going to shape the future of autonomous vehicles

China’s Baidu has unveiled the second version of its software for autonomous vehicles, Apollo, at CES 2018, in Las Vegas. It was originally launched in April last year as an open and free platform available for the entire industry, in what was rightly interpreted as a challenge to Google by emulating the Mountain View company’s Android strategy against Apple in the smartphone sector.

Google’s Waymo is the only self-driving project with a fleet of vehicles without a safety driver, transporting people in Phoenix, Arizona. Launched in November, by next month it is due to complete its testing phase and will become a fully autonomous taxi service. Google was always very clear that it did not intend to build vehicles, but instead provide a driver. The decision to use the Chrysler Pacífica was simply based on convenience: Fiat Chrysler was way behind the competition and saw the possibility of working with Google as a low-risk project that could boost its image.

Meanwhile, the rest of the motor industry’s players, unable to get to building an autonomous vehicle alone, have been looking for technology partners, resulting in a kind of corporate musical chairs: the most recent deal has been between Volkswagen and Hyundai, who have signed an agreement with Aurora, the platform created by none other than Chris Urmson, the former director of the Google Autonomous Vehicle Project, as well as Sterling Anderson, formerly part of the Tesla Autopilot project, and Drew Bagnell, a professor at the Carnegie Mellon Robotics Institute and founder of the Uber Advanced Technologies Group.

Meanwhile, Toyota has just announced at CES its e-Palette together with Uber and Amazon, an autonomous platform that can be used to create anything from a supermarket on wheels to delivering piping hot pizzas (and vindicating the visionary imagination of Charlie Brooker in one of the episodes of last season’s Black Mirror :-)

Other initiatives include GM’s purchase of Cruise (which aims to employ 1,100 people on average salaries of $116,000 in the coming years), Ford’s $1billion investment in ArgoAI, the MIT startup nuTonomy. All these, and other agreements between car or components manufacturers, plus technology companies highlighting the importance of the issue with new innovations: a fiercely competitive scenario that ensures the autonomous car it’s going to be a reality much sooner than what many people believed.

Against this backdrop, Baidu has made its spectacular move with Apollo. From the outset, its status as an open platform has given it unmatched dynamism and flexibility: its website features no less than 88 partners. What’s more, many of them are Chinese companies, part of a particularly cohesive ecosystem that may not be known much outside the country, but that’s driving innovation. Protected by the absence of foreign competition, Chinese companies are able to test their products and services then launch them throughout the world. If Apollo becomes the chosen platform for leading China’s vehicle makers, their brands could soon be playing a big role on in the autonomous vehicle sector.

Waymo’s progress has been spectacular, but so far it’s only working with Fiat-Chrysler, in what seems to be a temporary arrangement rather than a strategic alliance. Apollo might seem to be behind the game and not expecting to have vehicles on the roads until December 2020, nevertheless has created a broad platform of car and component manufacturers, universities, research institutes and technology companies. As with so many other areas, the development of the autonomous vehicle looks set to depend fundamentally on such platform strategies.


(En español, aquí)