Is This the Beginning of the End for Self-Driving Vehicles?

As 2023 draws to a close, it’s hard to make any kind of convincing argument that it’s been a good year for self-driving vehicles.

Long touted as the future, automated cars, in particular, have had a particularly bumpy 12 months. High-profile investigations into the Advanced Driver Assistance Systems offered by Tesla by American authorities and agencies have generated regular negative publicity, following on from a series of accusations in 2022 that the company was exaggerating the capabilities of ADAS tech with misleading advertising and marketing.

In addition, the year is finishing with serious questions being asked of the self-driving taxi sector in the United States, with one of the biggest players, Cruise, pausing all its activities after being stripped of its permits for driverless operations in California and suspending production of its eagerly anticipated, purpose-built Origin autonomous vehicle (AV). This dramatic chain of events followed an alarming incident in San Francisco in which a pedestrian was seriously injured after being dragged along the road by a self-driving taxi.

Credit: Getty Images

The resulting suggestion from the state’s Department of Motor Vehicles that Cruise posed an “unreasonable risk to public safety” and had not been transparent about the incident prompted accusations that the General Motors subsidiary had expanded too fast, too soon, and its autonomous tech simply wasn’t ready (and might never be).

Related:Ford, VW End Self-Driving Operations, Argo Closed

Little wonder, then, that a chastened GM, which has already lost $8 billion on Cruise since 2016, has subsequently pledged to reduce spending on the company by “hundreds of million dollars” in 2024, and initiate a reset with a much clearer focus on safety.

GM’s repositioning followed an even more extreme strategy change from fellow industry giants Ford and Volkswagen, who 13 months previously pulled the plug on their investment in autonomous driving company Argo AI, causing the Pittsburgh firm to close its doors permanently. The comments of Ford President and CEO Jim Farley at the time were quite telling. “We’re optimistic about a future for L4 [Level 4] ADAS, but profitable, fully autonomous vehicles at scale are a long way off,” he said.

All things considered, it’s a worrying landscape that has sparked much scrutiny and even soul-searching, causing many to question if the quest to deliver autonomous driving is on the right path – or a road to nowhere – and whether billions of dollars should continue to be spent on a future that, frankly, many never arrive. The memory of CES 2022 – when GM CEO Mary Barra suggested customers would be able to buy their own AV by the middle of the current decade – now seems a very distant one indeed. 

Related:Confidence in Self-Driving Vehicles Falling, J.D. Power Study

A more measured take, though, is that what has evolved is simply a dose of reality for an auto industry that had perhaps been a bit over-ambitious in its approach, and although there have been obvious difficulties, they should not cause automakers to lose sight of the bigger picture.

Steve Bell is a chief analyst (connectivity) at the Informa Tech Automotive Group. He points out: “There are hundreds of thousands of people being killed a year in road accidents. Addressing these safety issues would be better than focusing on the edge case [the Cruise robotaxi incident] that happened in San Francisco. Is it [the idea of autonomous driving] damaged dramatically in the long term? No, I think what it really means is that there’s got to be more evidence-based practice on the edge cases.”

What Bell does concede, though, is that there has been a realignment of focus. Wind back the clock a few years, and there was a very different attitude. 

“It was very gung-ho in the industry,” he says. “The attitude was ‘AI could conquer the world’. ‘AI could deliver autonomous vehicles’. ‘What can be done to move to Level 4 or Level 5 capability in terms of internal combustion engine vehicles?’.” 

Related:AAA Survey Finds Fear of Self-Driving Cars Is Increasing

In recent years, though, circumstances have changed. The pandemic, and its subsequent lockdowns, affected the global supply chain. There has been a very marked shift to electrification. The need to develop electric architectures and software-defined vehicles has become more pressing.

The result? “It has changed the priorities,” says Bell. “The core industry isn’t really as focused on autonomous driving, or autonomous vehicles or robotaxis. They’re more concerned about how they sustain their existing business. The redeployment of R&D resources is what is driving this industry at the moment. It’s not a simple thing to develop a self-driving vehicle and I think the industry realizes that.”

The Pace of Self-Driving Technology Advancements 

Although the rate of progress may not be matching the pace once envisaged by early self-driving cheerleaders like Tesla boss Elon Musk, advances are still being made. Development of architectures that facilitate continual updating means that – for some marques at least – the next level of assisted driving is only a software update away. 

And what’s undeniable is that some parts of the world are exploiting this potential more quickly than others. China, in particular, is showing what’s possible. 

“Look at what’s happening there with NOA [Navigation on Autopilot], a navigational system which is being developed on a city-by-city basis,” continues Bell. “I think China is where you’re going to start to see a level of advancement that you’re not going to get in Western Europe or in the United States.”

With armies of tech companies and EV manufacturers getting involved – both established names and new ones – the Chinese market is extremely competitive. The scale of how quickly it is evolving was underlined by electric automaker XPeng’s announcement in November that its City Navigation Guided Pilot tech – which delivers sophisticated driver assistance, including changing lanes and making left and right turns, in complex urban environments (without high-precision mapping, notably) – was being extended to 25 locations across the country, with 50 earmarked for operation by the end of the year.

XPeng's City Navigation Guided Pilot tech delivers sophisticated driver assistance, including changing lanes and making left and right turns, in complex urban environments (without high-precision mapping. Credit: XPeng

Hailed by XPeng as “the last step before full autonomous driving is realized,” the tech’s swift rollout is a dramatic example of what’s achievable.

“[In China], it will become the case where if you haven’t got the latest capability, you lose the potential sale in that particular city,” observes Bell. “But the interesting thing will be to what degree it gets brought elsewhere and what happens in terms of regulation if you’ve got Chinese vehicles coming in with that capability.”

It’s a pertinent point. The current geopolitical situation means that bringing Chinese auto brands to the U.S. with this kind of functionality is going to be extremely difficult, if not impossible. But Europe may be a different matter, with Bell predicting: “I think that what’s going to happen in Europe is that the Chinese brands will build their knowledge and capability, and you’ll start to see some of the over-the-air enhancements of advanced driving systems coming into the market.”

So, despite a stream of pessimistic opinion pieces in mainstream Western media outlets questioning the future of automated and autonomous vehicles, there are still reasons for optimism, even allowing for the missteps we’ve seen of late.

Public Confidence and Education 

One area that does unquestionably need to be addressed, though, is public confidence in the tech, which is seemingly on the wane. An American Automobile Association survey published in March revealed that fear of self-driving vehicles had risen significantly. This was backed up by a JD Power report in October that showed consumer readiness for automated vehicles was slipping.

On the current uncertainty facing self-driving taxis specifically, Bell is quite clear. “I don’t think that it’s a cataclysmic situation,” he says. “Track records have to shape this. If you can show that accidents per million miles being driven are improving, that edge cases are actually being addressed and that there’s more capability being put into the vehicles… it’s just a case of working through the process.”

And in the interests of fairness, it’s worth pointing out that Cruise’s problems in San Francisco have not been replicated elsewhere. 

In Phoenix, for example, where Waymo runs a fleet of around 200 self-driving taxis, the reception has largely been positive. As Mayor Kate Gallego noted: “Our residents have generally really appreciated this service.”

A Waymo self-driving taxi outside Phoenix’s Sky Harbor International Airport. Credit: Waymo

Bell feels there is more work to be done, though, on the education of consumers on the automated capabilities of the production cars they can buy. 

There has long been concern about how some systems are presented and marketed, and the potential for confusion is obvious, with an American Institute for Highway Safety study in 2022 showing that many drivers treated vehicles with partial automation as fully self-driving, with potentially disastrous consequences.

Accordingly, General Motors has made moves to increase public confidence in ADAS systems, such as its own Super Cruise, with its “Hands Free Eyes On” initiative, which aims to raise awareness of its benefits – and limitations.

Bell believes this kind of approach has merit. 

“If you look at the average person’s utilization of new technology in vehicles, it’s not that great,” he explains. “And as an industry, we’ve not been good at standardizing or reeducating drivers. 

“Is there an education in terms of what is the state of semi-autonomous driving? And is there consistency in terms of how you show indications on the dashboard or particular types of events like lane keeping or changing lanes? These are things we need to think about. The industry needs to take a lead. The broad issue is we need to advance the level of knowledge so that you can get discerning drivers and users out there.”

And that, perhaps, is the key to the future: a better understanding of what the tech is and what it can do. As 2023 has illustrated, self-driving – whether partially or highly automated or even fully autonomous – is an idea that is built on the most fragile of confidences. Despite the billions being spent on testing, highly sophisticated software and groundbreaking hardware, it only takes one or two high-profile incidents to have the whole concept called into question.

Like what you've read? For more stories like this on self-driving cars and other emerging technologies, sign up for our free daily email newsletter to stay updated!