The timeframe for deploying fully autonomous vehicles has been stretched, as OEMs and other businesses consider different areas of focus within the automotive industry.
It’s 2020 and as recently as five years ago, and even more recently, it was widely predicted that autonomous vehicles would be a common reality by now. It hasn’t happened.
Not only has it not happened, the forecasts for achieving full autonomy — the ability for self-driving vehicles to drive themselves anywhere, anytime, and in any condition — seems to be extending rather than contracting.
“I think full autonomy is still a long way off,” said Magna CTO Swami Kotargiri, in an interview for SAE Automotive Engineering. He’s not alone in that opinion.
“A degree of realism has set in regarding the commercial introduction of the autonomous vehicle,” states a report on a worldwide survey of “stakeholders in the automotive ecosystem”, conducted by TU-Automotive. While the tone of that report was positive towards AVs (autonomous vehicles), 22 per cent of all respondents believed it will take until 2035 before all vehicles will be autonomous. A further 25 per cent believe it will be 2040 or beyond.
That same report noted that it is the technology suppliers, not the vehicle OEMs, that are driving the development of AV technology. While some might consider Silicon Valley tech companies to be those most-likely suppliers, more traditional automotive sources are also front and centre in AV development.
One such traditional supplier is ZF — best known in the past for its transmissions, but now widely divergent in its product lines. It’s also the fifth largest automotive supplier in the world in terms of sales to OEMs.
ZF is heavily involved in the development of AV systems, as Aaron Jefferson, Vice President of Product Strategy for the company’s Global Electronics division recently explained. Jefferson was addressing a group of international automotive journalists, who are jurors for the 2020 World Car Awards at that organization’s annual L.A. Test Drives event in Pasadena, California.
According to Jefferson, ZF is now following a two-pronged approach towards autonomy: Level 2+ systems for privately-owned passenger vehicles, and Levels 3 and 4 for commercial people and cargo-movers.
Level 2+ systems include many of those Advanced Driver Assist Systems (ADAS) with which we are already familiar — things like adaptive cruise control, automatic emergency braking, lane keeping assistance and lane centering, as well as traffic jam assist and traffic sign recognition. ZF calls them “feet-free, hands-on” systems, which aptly describes them.
Many of them are already available — not just on premium vehicles, but on several vehicles destined for the mainstream market. Progressing towards more encompassing levels of autonomy, however, will entail dramatic, perhaps near-asymptotic increases in both sophistication and costs, Jefferson suggests.
In very rough numbers, while current Level 2 and near-term Level 2+ systems may cost in the low thousands of dollars, or less at the manufacturer level, Level 3 and beyond systems could cost $10,000 or more. That is why, when we do get there, it is likely to be with vehicles geared towards the commercial market, including cargo-delivery and ride-hailing services. And even then, it’s likely to be within strictly defined areas.
Magna’s Kotargiri concurs. “I think (full autonomy) will start in more geofenced areas,” he said. “I think that’s where you take the first step.”
For ZF’s part, it has mapped out an evolutionary progression of systems leading from where we are now to fully feet-off, hands-off driving, if not yet truly driverless vehicles. “As we continue developing technologies for higher levels of automation we will use Level 2+ systems both to learn, train and validate the systems, and to build consumer confidence in the benefit and effectiveness of semi-autonomous functions,”said Jefferson.
In the near term, an integrated system called ZF coASSIST is scheduled to start production in the fourth quarter of this year. Relatively low in cost, its features include traffic jam support and highway driving support, with driver-initiated lane change and highway exit capabilities.
Looking further ahead to the 2023-2024 timeframe, ZF coDRIVE will build on ZF coASSIST’s features to achieve truly hands-free, feet-free capabilities, including lane change and highway entry and exit. Adding onto that base, ZF coPILOT will offer such features as voice control and HMI (Human Machine Interface) displaying vehicle surroundings in real time. Plus, it will accommodate open platform third-party software.
It’s an ambitious plan for just five years from now. But it’s typical of the pace of development that is pervasive throughout the industry. Undoubtedly, ZF’s competitors are equally committed to making similar progress.
So, even though forecasts of full autonomy by 2020 may have been far-fetched, the reality of what is possible now and in the immediate future is still the stuff of dreams.