Ford thinks driver assist is a safer bet than driverless cars, but it’s fooling itself

When Ford announced yesterday that it was pulling its support for Argo AI, the autonomous driving startup it had financed since 2017, it cited as one of its reasons a belief that driver-assist technology will have more near-term payoffs.

The automaker said it will be shifting its resources from the Level 4 autonomous technology that was under development at Argo to “internally developed L2+/L3 technology.”

“So it’s taking that investment and putting it towards a business where we think we will have a sizable return in the near term relative to one that’s going to have a long arc,” said John Lawler, Ford’s chief financial officer, during a call with investors Wednesday.

But driver-assist technology, especially so-called Level 3 systems, carry its own risks. There is a question as to whether Level 2 and Level 3 systems make driving safer or simply more convenient. Efforts by other automakers to commercialize Level 3 systems have run into roadblocks, while some AV developers have dismissed Level 3 as simply too dangerous to even consider.

The Levels of Automation as defined by the Society of Automotive Engineers are themselves controversial and open to interpretation. Level 2 is used to describe advanced driver-assistance systems (ADAS) in most modern vehicles today, which includes features such as adaptive cruise control and lane-keep assist. (L2 Plus is likely a reference to features that allow a driver to take their hands off the steering wheel under certain conditions, like on a highway, as long as they remain vigilant and ready to take over.)

Once you go past Level 2, it starts to get tricky. Level 3 refers to highly automated driving, where the driver still needs to be able to take over the vehicle upon request. Level 4 automation means that the car is capable of handling most driving situations itself without human intervention, whereas Level 5 is largely theoretical and covers complete automation in any condition.

But most AV operators, including Waymo and Cruise, have said they are skipping Level 3 and working exclusively on Level 4 technology. The reason is that Level 3 is seen as being potentially dangerous, given the need for drivers to stay attentive despite the vehicle performing most of the driving tasks.

There have been studies that show that hand-off between automated system and human driver can be especially fraught. When people have been disconnected from driving for a longer period of time, they may overreact when suddenly taking control in an emergency situation. They may overcorrect steering, brake too hard, or be unable to respond correctly because they hadn’t been paying attention. And those actions can create a domino effect that has the potential to be dangerous — perhaps even fatal.

“Drivers who think their cars are more capable than they are may be more susceptible to increased states of distractions, and thus at higher risk of crashes,” robotics expert Missy Cummings wrote in a paper from 2020 evaluating the risks of Tesla’s Autopilot system (which is just Level 2). “When an operator’s task load is low while supervising an autonomous system, he can experience increased boredom which results in reduced alertness and vigilance.”

Other automakers have been tripped up by the promise of Level 3 driving. In 2018, Audi said its A8 sedan would come with a feature called Traffic Jam Assist that, when active, would relieve human drivers of the need to pay attention during stop-and-go traffic. But the feature was contingent on approval from local authorities, and Traffic Jam Assist remains dormant in most markets around the world. Audi has no plans to activate the feature, and Level 3 automation remains a morass of legal, regulatory, and business-related challenges.

BMW had said its forthcoming electric vehicle would also include Level 3 capabilities, but when the iX SUV was unveiled in 2020, the feature was noticeably absent. Volvo says it’s negotiating with California regulators over the release of a Level 3 highway driving system with its upcoming EX90 SUV.

To date, the only automaker to publicly release a Level 3-equipped vehicle has been Honda. Earlier this year, the company began leasing its Legend Hybrid EX in Japan, equipped with the top-shelf Honda Sensing Elite suite of driver-assist technologies. But according to Automotive News, which tested the vehicle, the system “introduces a whole new level of stress.” A test drive in Japan mostly showed off the system’s limitations rather than its capabilities.

The Autopian’s Jason Torchinsky quotes an unnamed AV engineer who works for a major automaker calling Level 3 “hot garbage, and potentially incredibly dangerous.”

For its part, Ford sounds jazzed about the prospects of Level 2 Plus and Level 3 technology, noting that 83,000 customers are already using the company’s hands-free highway system, BlueCruise. Ford said it expects to have more than 30 million connected vehicles on the road around the globe by 2038, all of which have the potential to be revenue-generating machines.

In Ford’s view, it would rather collect monthly ADAS subscription fees from its millions of customers than wait years — or potentially decades — for an urban robotaxi service to begin generating profits.

But the question isn’t whether Level 2 or 3 systems can make money; clearly, they can — just look at Tesla, which has 160,000 customers who have paid at most $15,000 for the company’s Level 2 Full Self-Driving feature. But the real question is whether these systems are safe enough to deploy. Tesla is under multiple investigations for its self-driving claims, with a particular focus on dozens of crashes that have occurred involving FSD and Autopilot — the latest being a criminal investigation into whether Tesla overstated its vehicles’ capabilities.

Ford is a much more conservative company than Tesla. It seems extremely unlikely that Jim Farley would be as flippant about rolling out a new technology to his customers as Elon Musk. But Level 3 is uncharted waters. And the jury’s still out as to whether there are any real benefits to the technology beyond a near-term cash influx and the bragging rights to beating your competitors to introduce a brand-new technology.