Sun, Jun 03, 2018 - Page 15 News List

The race to autonomy might be won in China as Beijing pushes road testing

By Anjani Trivedi  /  Bloomberg

A self-driving vehicle operated by Tencent Holdings Ltd is displayed at a ceremony to receive a license for driving tests on public roads in Shenzhen, China, on May 14.

Photo: Southern Metropolis Daily via Reuters

Everybody wants autonomous vehicles now: It is the auto industry’s way forward and Silicon Valley’s latest preoccupation. China is no different.

Alibaba Group Holding Ltd (阿里巴巴) is testing self-driving cars in China and Baidu Inc (百度) last year started trials of autonomous technology.

BMW AG early last month became the first foreign automaker to get a license to test its offering in China.

In the middle of last month, Shenzhen-based Roadstar.Ai LLC raised a record amount from Chinese investors.

Meanwhile, Tencent Holdings Ltd (騰訊) and ride-hailing provider Didi Chuxing (滴滴出行) in the past couple of weeks said that they are testing autonomous cars in California.

They join a host of niche Chinese tech companies that account for about one-fifth of those with permits for self-driving trials in the US, alongside automakers and start-ups.

A lot of the developments follows a Beijing-mandated push. In April, the government laid out national road-testing guidelines for autonomous vehicles, adding to rules in place in Shanghai and Beijing.

That would make it possible for China to collect vast amounts of data on autonomous-vehicle testing on public roads, the way Alphabet Inc, Google’s parent, did with its sibling, Waymo.

If automakers see a future in China, the world’s largest car market, they will need to have data collected and created in China. Californian data will not fly.

China’s official entry into the race could address the biggest barrier to mass production of autonomous cars — cost — as it did with batteries for electric cars over the past year.

Getting rid of drivers cuts costs. Take a US electric car that runs 24 hours a day, seven days a week as part of a ride-sharing network. According to Nissan Motor Co chairman Carlos Ghosn, charging the car costs about US$250 a month and the lease takes US$300 to US$400 a month. Because the cars run around the clock, there are three drivers — together they draw US$15,000 a month. Eliminate that expense, and this is a very different proposition.

As much commercial sense as that makes for ride-hailers, such as Uber Technologies Inc and Didi, and automakers, such as Nissan, supplying fleets to these companies, the cost of the technology itself (and the associated safety testing) is still far too high — indeed, it is the biggest hurdle to the widespread use of autonomous vehicles.

Take one example: Laser-based sensors made by Velodyne Lidar Inc, the key element in a driverless car, cost between US$8,000 and US$80,000, and weigh anywhere from just under 1kg to as much as 13kg.

If these light-emitting radars were mass-produced, they might cost as little as US$1,000. Cheaper workarounds do not quite cut it yet, although Tesla Inc chief executive Elon Musk seems to believe they do.

This is where China can help. Look at Quanergy Systems Inc, a low-key Silicon Valley maker of lidars, which use light to create three-dimensional images from a 360-degree field of view.

Quanergy has already teamed up with the likes of Daimler AG’s Mercedes-Benz, Delphi Technologies PLC and Nissan, and is probably closest to making the ideal sensor.

The Californian firm also has a strategic partnership with Sensata Technologies PLC, a technology developer that says its content per vehicle in China is growing rapidly.

Quanergy’s lidars, which are solid-state and tiny, already cost less than the magic US$1,000. Gross margins are somewhere above 50 percent, while research and development costs are almost 70 percent of revenue.

This story has been viewed 3096 times.

Comments will be moderated. Remarks containing abusive and obscene language, personal attacks of any kind or promotion will be removed and the user banned.

TOP top