You can enable subtitles (captions) in the video player
Is the driverless car industry hitting a roadblock? Apple's cancellation this year of Project Titan, its secretive attempt to develop an autonomous EV, is a stark reminder of the ongoing challenges the industry faces. The tech giant started work on the project back in 2014. But even for a company with Apple's resources, developing self-driving cars is expensive.
And car manufacturing is a low-margin business to begin with. McKinsey has made bullish predictions that a driverless private passenger car market could generate around $400bn by 2035. But so far, the industry's collective efforts have cost more than $160bn and revenues are paltry.
Safety concerns are a major problem. The industry prioritises fast development, but ensuring safety requires a more cautious approach. And deep learning systems, the brains behind self-driving cars, can struggle with unpredictable situations.
San Francisco's Municipal Transportation Agency recorded more than 600 self-driving vehicle incidents between June 2022 to June 2023. And autonomous car accidents involving companies like Tesla, Uber, Amazon Zoox, and Cruise have cast a long shadow. After a Cruise driverless car hit a woman at a San Francisco intersection last October, its parent company, GM, recalled its vehicles. GM has lost more than $8bn on Cruise since it acquired the start-up in 2016, and it cut spending on the unit.
As it continues to deal with safety and cost challenges, the industry may also be undergoing a structural change, with China rapidly emerging as a strong competitor to the west. China started commercial driverless development in 2013, about five years after groups in the US. But as of September last year, autonomous vehicles in China had driven a cumulative total of 70mn kilometres. That's on par with the US.
Their extensive network of cameras and data gives them a significant advantage in developing this technology and overtaking the west. A McKinsey global survey of autonomous vehicle execu,tives published in January, revealed expectations for commercially viable robotaxis have been extended by around two years to about 2030. However, the road ahead is likely to be long and winding, with public trust and technical challenges remaining crucial hurdles.