Piton detection systems do not work most of the time, Do not rely on technology to avoid collisions, according to a new AAA study


Vehicles equipped with automatic emergency brake systems (AEBs) that detect pitons are proliferating, but the annual number of deaths in traffic accidents over the last decade has increased, according to the American Automobile Association (AAA). . AAA researchers are concerned that too much dependence on new technologies can lead to lax behavior. Their fear is accentuated by the results of a new study that the organization has conducted, which shows that pedestrian detection systems do not work most of the time as they should, and that they are especially less effective at night.According to AAA, on average, a piton is every 88 minutes in traffic accidents in the United States. This represents more than 16 people per day, nearly 115 people a week. In 2017, nearly 6,000 pitons lost their lives, accounting for 16% of all deaths on the road, a rate that has been rising steadily since 2012, according to the report.

AEB and piton detection are two features that are part of the growing category of Driver Assistance Systems (ADAS), but according to the organization's study report, we may not be relying on these technologies. In its recent study, the association tested the ride detection behavior of four popular 2019 mid-size sedans, a Chevrolet Malibu, a Honda Accord, a Tesla Model 3 and a Toyota Camry in a variety of different scenarios. The vehicles were all equipped with AEB with piton detection system, the purpose of the tests both to detail the performance and limits of these functions. These systems typically use the on-board camera system and computer vision algorithms to detect pitons in the field of view. The tests were carried out on the dry asphalt, on a closed circuit with simulated pitonne targets, in day and night conditions, varying speeds and with different piton behavior, according to the AAA.

For each test, the longitudinal distance and the time to the collision were recorded when each vehicle gave a visual warning that a collision was imminent, as well as when the vehicle began to brake automatically. The impact velocity or separation distance was also recorded, depending on the results of the tests.

Very mixed results of different tests

For the easy test, the Chevy Malibu failed to brake enough to significantly reduce the speed before each collision, but it still detected the false adult peak while driving 32 km / h, averaging 2.1 seconds and 19.2 m before the impact, during the five tests. The Model 3 did not do better and also hit the piton dummy in each of her five tries. It slowed down by 4.5 km / h and alerted the driver an average of 1.4 seconds and 12.7 meters before the collision. In two tests, there was no braking at all, even though the system detected the piton manikin, according to the report.

The Honda Accord did better than the previous two cars. Although he alerted the driver much closer to the peak time before the 0.7 second collision, distance of 9.7m, he prevented the impact from occurring in three of the five tests and slowed the car 1km / h as fourth. The best performance of the four cars was raised with the Toyota Camry. According to the report, she gave a visual alert 1.2 seconds and 10.8 m before the impact and stopped completely before reaching the manikin in each of the five trials.

When AAA researchers tried to test each car 48 km / h, the results were worse. Only the Honda Accord was able to slow down more than 8 km / h by detecting the 20 km / h model. In repetitive tests, it has slowed down an average of 12.5 km / h, including in two cases where it has completely stopped.

AAA has also tested in more complex situations, including low-light testing one hour after sunset without ambient public lighting, but with the passing headlights of the cars running. Here is a summary of the results.

During the night tests, none of the four systems reacted to the peaks or even detected their presence, according to AAA. This is particularly troubling, as 75% of pitfall deaths occur after the nightfall, after AAA. The researchers also found that the systems were also ineffective at speeds above 48 km / h. Even at 32 km / h, a collision occurred 89% of the time when a car encountered a child (model) coming from between two station cars. With two adults standing along the road, back the traffic, and the car driving again 32 km / h, a collision occurred in 80% of the tests. A collision also occurred each test of a car meeting a piton immediately after a right turn.

The researchers found that the systems worked best when an adult crossed in front of cars traveling 32 km / h. But even in this scenario, collisions still occurred 40% of the time, according to AAA. This shows that technology is not (for now) even replacing an attentive human driver.

Greg Brannon, Director of Automotive Engineering and Industrial Relations at AAA, said the number of pitons is increasing, which proves how important the impact of these systems on safety could be when they are perfected. . But our research has shown that current systems are far from perfect and always require a committed driver to drive, he added.

General Motors, manufacturer of the Chevy Malibu used in AAA tests, published a report in response to the results of the study. GM said it has been proven that the technologies in question reduce the speed of impact or help prevent accidents altogether. The automaker cited a study by the University of Michigan Transportation Research Institute in which its self-contained emergency braking system reduced rear-end crashes by 44 percent, but did not specifically mention collisions with pitons. In the same vein as AAA, GM noted that the technical features do not replace the driver's primary responsibility.

The recommendations of the American Automobile Association to drivers

These results prove that your car may be equipped with the latest intelligent digital security systems, but it is far from perfect. LAAA recommends that drivers do not rely on pedestrian detection systems to avoid collisions. The organization also advises drivers to become familiar with vehicle safety techniques by asking dealer questions and reading the owner's manual.

Other AAA research has shown that drivers tend to over-trust assistive technology driving and can easily overestimate the capabilities of the technology. Drivers should also be very cautious at night, according to AAA. AAA also pointed out that, like drivers, pitons should remain alert at all times and avoid distractions such as sending text messages while walking.

Source: AAA

And you ?

What do you think ?
What do you think of the figures obtained in the study?
But our research has shown that current systems are far from perfect. What's your comment ?

Read also

The "Autopilot" system name may lead drivers to overestimate Tesla's actual capabilities, according to a survey
Americans lose confidence in self-driving vehicles following fatal accidents by Uber and Tesla, polls show
Tesla's Autopilot is safer than manual driving, according to an investigation commissioned by an insurance company

Source link


Please enter your comment!
Please enter your name here

8 + 1 =