- Research from CLSA suggests the number of autonomous vehicles on US roads will rise from 800 this year to 796,000 in 2022.
- CLSA defined levels of vehicle autonomy in five different grades, with different companies focused on different segments.
- As the technology matures, regulation and insurance were cited as the two most challenging areas for the industry to navigate.
The race to commercialise autonomous vehicles is one of the key battlegrounds in big tech.
And aside from building out the technology, the nascent sector still faces a number of regulatory hurdles.
Safety concerns came into focus when a self-driving Uber vehicle struck and killed a pedestrian in March.
But as an example of the commercial opportunities in the sector, Morgan Stanley pegged the potential value of Waymo — the autonomous vehicle company owned by Google — at a staggering $US175 billion.
And according to research from CLSA, the number of autonomous vehicles on US roads is about to rise — sharply:
Based on that forecast, market dynamics are about to shift from testing to commercialisation.
But the details of how the shift will play out — and the many regulatory questions still unanswered — mean the practical rollout will probably be more complicated.
There’s also the question of how one actually defines autonomy. As a broad guideline, CLSA referred to this table:
“There are already a lot of level-1 cars on the road”, with features such as “lane-departure warnings and adaptive cruise control”, said Michael Barr, a portfolio manager on the Autonomous Vehicle Fund at Neuberger Berman.
“These features are building blocks. Think of full automation as a culmination of multiple building blocks. We’re already in the process of creating level 5 through some of these building blocks that are level 1 and level 2.”
But in terms of pushing towards Level 5, different companies are going about it in different ways.
For example, Waymo has skipped Level 3 — a vehicle that drives itself with some passenger instructions — entirely.
But Barr thinks Level 3 cars will provide a hybrid alternative as the market moves towards fully autonomous vehicles.
“In dense urban areas, we expect to see some form of driverless ‘robotaxi’ ridesharing play out in the next five to 10 years in geo-fenced settings,” he said. Geo-fencing is using technology to build a virtual border around an area which can warn or trigger a shutdown if mobile devices leave or enter it.
“Then you’re going to have the general public probably having level-3 cars. For instance, you’ll get the car on the highway, push a button, the car drives you and then you re-engage just in time to get off the highway.”
Just last week, Volvo showed how it has partnered with LiDAR developer Luminar to enable its cars to reach Level 4 autonomy – the one where you get to sleep behind the wheel, if you’re brave enough.
Before that happens though, appropriate regulatory solutions will have to be found. A recent report by global law firm Herbert Smith Freehills claimed there were 716 legislative barriers to the adoption of automated vehicles.
As Neuberger analyst Yan Taw Boon noted, “the reality is that technology is way ahead of regulation”.
In view of that, both Moon and Barr agreed that the US is best placed to develop regulatory guidelines around self-driving cars.
US lawmakers have already tabled two bills in the House and Senate — the Self Drive Act and the AV Start Act — aimed at providing a framework for cyber-security and safety, respectively.
“Overall, the government seems supportive of the technology,” Barr said. But whether insurers are on board is still up for debate.
“We haven’t yet had much commentary from the insurers on the question of liability when it comes to potential accident,” Barr said.
So while the technology races ahead, regulation and insurance were cited as the main factors which will delay a mass rollout.
However, “we find it hard to envision a scenario where regulations completely prevent this technology from ultimately coming to fruition”, Barr said.