Uber sells robo-taxi business to FCA partner Aurora

The U.S. ride-sharing company Uber has unexpectedly parted ways with its autonomous driving division and sold it to the young U.S. company Aurora. In turn, Aurora is cooperating with the Fiat-Chrysler Group to develop autonomous commercial vehicles and shuttle vans. Additionally, around the same time, in June 2019, the Hyundai Group announced an investment in the company Aurora through the brands Hyundai and Kia to develop autonomous driving technologies, which is set to culminate in a pilot test with robo-taxis by 2021. Shortly after Aurora partnered with FCA, VW ended a collaboration with Aurora and focused on cooperating with Ford and its partner Argo AI. This also primarily concerns ride-hailing models and delivery vehicles.

The company Aurora develops software for autonomous vehicles and was founded in 2016 as a start-up by former chief developer of Google robot cars Chris Urmson and former Tesla autopilot expert Sterling Anderson. On the other hand, Uber had acquired former Google autonomous driving specialist Anthony Levandowski and his small company in 2016. Later, he was sued by the tech company for data theft and had to pay Google $179 million. Uber's activities in this area were also overshadowed by a fatal accident in which an autonomous SUV failed to detect and ran over a woman crossing the street with a bicycle at night.

Uber takes a step back: Finally becoming profitable

Uber now wants to focus on its platform business, according to the company. Recently, there was speculation about an acquisition of the relevant BMW-Daimler division of the joint venture, for which Uber has apparently made an offer. In turn, they plan to source robo-taxi technology from Aurora. For years, the company's leading idea had been to quickly replace the largest cost factor in operations, personnel, with autonomous technologies. Unlike previous long-term announcements, Uber CEO Dara Khosrowshahi now sees the deal as short-term: "This significantly advances us on the path to profitability," he said of the sale of what used to be a core component in the value chain. The plan is to be profitable for the first time next year.

VW CEO Herbert Diess recently reiterated his commitment to the technology in an interview with Wirtschaftswoche, stating that he believes the high market valuation of companies involved in autonomous driving is justified.

"If all taxis and trucks eventually drive autonomously, then this justifies a high valuation, especially in countries with high wages, such as the USA or Europe, where significant savings are conceivable," Diess continued.

He sees great significance and opportunities especially for Google's subsidiary Waymo and Tesla with the next generation of the so-called "Autopilot." The Californians are coming out next year with a system that will triple the performance compared to the current one, he noted.

"One of Tesla's strengths is that with their now very large fleet of vehicles, they continuously collect driving data and use artificial intelligence to make the system better. If you like, Tesla is not just a car company, but a neural network that is constantly learning to drive better," acknowledged the VW CEO.

He expects the first autonomous models to be production-ready between 2025 and 2030. He cites as a reason that the performance of computer chips doubles every 18 months. Combined artificial intelligence is developing even faster. It is foreseeable that the systems will soon be able to handle the complex situations involved in autonomous driving, the VW Group CEO believes.

What does this mean?

It almost makes one's head spin with the game of musical chairs currently taking place in the second big megatheme of the industry, besides the far more significant electrification in terms of CO2 emissions. However, the crucial proof is still pending or questionable: Whether it will ever contribute to the efficiency of traffic and therefore to climate protection. Perhaps sometime around the 2030s, and only then if the systems are scaled and integrated into public transport, thereby replacing private trips, saving space, and avoiding "deadheading". Only under this condition is the use of the extraordinarily expensive and complex technology likely to be economically viable.

In the meantime, for private users, level 3 semi-automated driving will suffice, with Daimler forging ahead next year - as an optional package for the new S-Class, legally limited to 60 km/h, all within the scope of "motorized individual transport," of course. In the meantime, users are grateful for the assistance in stop-and-go traffic, which increasingly works reliably, or on highways. Or even more importantly: turning assistants. For now, that's all that's needed. Whether a fully autonomous car will ever become affordable enough for private ownership remains entirely uncertain. It's highly unlikely.