Are Modern Driving Assistance Features Unreliable? AAA Researchers Say Yes

Matt Posky
by Matt Posky
are modern driving assistance features unreliable aaa researchers say yes

The American Automobile Association (AAA) is recommending automakers limit the use of advanced driving aids after concluding they’re not really up to the challenge of providing reliable safety.

Over the past two years, AAA has focused on testing crash prevention systems to see if they’re all manufacturers claim — deciding that while many are useful in some instances, they’re far too inconsistent to be considered reliable safety nets. Like us, the group worries that making these features commonplace has created a false sense of security among drivers. While one might assume advanced driving aids have to be halfway decent to be put into vehicles, AAA’s pedestrian detection test from 2019 showed they’re anything but consistent.

On Thursday, America’s favorite motor club returned to report on its latest findings on five systems currently offered by the industry. For the test, AAA selected a 2019 BMW X7 with Active Driving Assistant Professional, 2019 Cadillac CT6 with Super Cruise, 2019 Ford Edge with Co-Pilot360, 2020 Kia Telluride with Highway Driving Assist, and a 2020 Subaru Outback with EyeSight. The group was sent to numerous testing sites in California, Utah, and Nevada, and given a 4,000-mile shakedown on public roads — where the outfit found the systems averaged a misstep or disengagement roughly every 8 miles.

Lane keeping seemed to be the system giving testers the most trouble, as it occasionally became befuddled by faded markings, automatically steering vehicles off road in some instances. AAA said these incidents would have been exceedingly dangerous if the driver wasn’t paying attention or had placed too much faith in the onboard features. This was made worse by cars tracking true and steering themselves through several shallow corners with aplomb, only to totally come apart and disengage in a nearly identical curve later.

Problems were not isolated to turns, however. Vehicles sometimes drifted out of their lane on a straight trajectory as well, only to over-correct and find themselves in the wrong lane.

Obstacle detection was another serious issue. Testing in a closed environment, AAA found that test cars typically hit stationary obstacles ( video here) representing a disabled vehicle upon approach, rather than avoiding them. When encountering this test scenario, in aggregate, collisions occurred 66 percent of the time with an average impact speed of 25 mph.

“Active driving assistance systems are designed to assist the driver and help make the roads safer, but the fact is, these systems are in the early stages of their development,” said Greg Brannon, AAA’s director of automotive engineering and industry relations. “With the number of issues we experienced in testing, it is unclear how these systems enhance the driving experience in their current form. In the long run, a bad experience with current technology may set back public acceptance of more fully automated vehicles in the future.”

It also might turn you into a grease spot on the pavement if you’re not ready to handle their occasional hiccups. However, automakers have grown more vocal about how these assistance packages aren’t meant to replace an attentive driver (even though many literally steer, brake, and accelerate the vehicle for you). This has prompted loads of debate among consumer advocacy groups, which claim the language being used to sell these systems is misleading and that the systems ultimately dull a driver’s senses by lulling them into feeling overly confident in a car’s limited capabilities.

Even the way the cars were tested showcased some of the issues confronted by consumers. A few systems couldn’t be tested in a closed environment. For example, Cadillac’s Super Cruise requires a divided highway to work at all. But they all have their own personalities, with different levels of engagement and thresholds for disengagement. Ultimately, AAA said it didn’t see much (if any) improvement from the last batch of vehicles tested. The group also noted that legal experts and regulatory groups have started to caution that it may be decades before fully autonomous cars are ready for widespread deployment, and that the systems that foreshadow their arrival could be doing more harm than good.

“AAA has repeatedly found that active driving assistance systems do not perform consistently, especially in real-word scenarios,” Brannon said. “Manufacturers need to work toward more dependable technology, including improving lane keeping assistance and providing more adequate alerts.”

Researchers have released the study in full if you’d like to see the metrics used and how the vehicles stacked up against each other. AAA wants the takeaway to be that these systems are still lacking across the board. It’s recommending the auto industry cool its jets until serious improvements can be made, and to stop trying to put these advanced driving aids into every vehicle they manufacture until they’re substantially improved.

[Images: AAA]

Comments
Join the conversation
2 of 39 comments
  • Fastlight1955 Fastlight1955 on Aug 07, 2020

    This seems like a good time to post my thoughts on the whole autonomous driving issue. The standard disclaimer with all of these so-called "safety systems" or "driver assistance systems" is that the human driver is ultimately responsible for the car, which means that he has to be able to take over control when the system runs out of talent. For him to be able to do that, he has to pay even closer attention to the situation than if he were doing all the driving - constantly analyzing and second-guessing the system for errors. For me (with 50 years of driving experience and no at-fault accidents), that's WAY more mental effort than just taking complete control of the car. And as long as that is the case, I have no interest in having ANY of those systems on my car, and I SURE don't want to have to pay for them and maintain them. In short, until and unless these systems are perfect, they are nothing more than a high tech, but still imperfect, "backseat driver". Either I'm driving, in which case I don't want or need the systems' assistance - or the car is driving, in which case I expect it to be at least as good of a driver as I am (such that I can read the paper if I want to). And if anything does go wrong, I expect the manufacturer of the car to be financially responsible for my losses - even if my name is on the title - because after all, it was their software that was driving. And in today's world, that ain't likely to happen. So I'll take a pass on ALL of these "driver assistance" technologies.

  • Ponchoman49 Ponchoman49 on Aug 07, 2020

    No kidding! As a long time computer tech guy I have been cautioning tons of people to not trust these advanced safety features and completely rely on them. We have become a tech obsessed world the past 20 plus years but there is no better and reliable tech than our own eyes, ears and senses.

  • SCE to AUX Toyota the follower, as usual. It will be 5 years before such a vehicle is available.I can't think of anything innovative from them since the Gen 1 Prius. Even their mythical solid state battery remains vaporware.They look like pre-2009 General Motors. They could fall hard.
  • Chris P Bacon I've always liked the looks of the Clubman, especially the original model. But like a few others here, I've had the Countryman as a rental, and for the price point, I couldn't see spending my own money on one. Maybe with a stick it would be a little more fun, but that 3 cylinder engine just couldn't provide the kick I expected.
  • EBFlex Recall number 13 for the 2020 Explorer and the 2020 MKExplorer.
  • CEastwood Every time something like this is mentioned it almost never happens because the auto maker is afraid of it taking sales away from an existing model - the Tacoma in this instance . It's why VW never brought the Scirrocco and Polo stateside fearful of losing Golf sales .
  • Bca65698966 V6 Accord owner here. The VTEC crossover is definitely a thing, especially after I got a performance tune for the car. The loss of VTEC will probably result in a slower vehicle overall for one reason: power under the curve. While the peak horsepower may remain the same, the amount of horsepower and torque up to that peak may be less overall. The beauty of variable cam lift is not only the ability to gain more power at upper rpm’s on the “big cam”, but the ability to gain torque down low on the “small cam”. Low rpm torque gets the vehicle moving and then big horsepower at upper rpm’s gains speed. Having only one cam profile is now introducing a compromise versus the VTEC setup. I guess it’s possible that with direct injection they are able to keep the low rpm torque there (I’ve read that DI helps with low rpm torque) but I’m skeptical it will match a well tuned variable lift setup.
Next