Wheels (Australia)

STOPPING SHORT

ION INVESTIGAT Our testing reveals potential flaws in autonomous braking tech

- JANUARY 2017 TOBY HAGON

Tests reveal flaws in autonomous braking tech

SOME of the latest crash avoidance systems on popular cars can be fooled by the high-riding design of utes and SUVS, according to an exclusive Wheels assessment conducted as part of our 2017 COTY testing.

During testing, the autonomous emergency braking systems on some new cars failed to recognise the higher rump of a ute – closer in height to many popular SUVS – potentiall­y leaving drivers exposed in common nose-to-tail crashes these systems are designed to avoid.

The revelation comes after extensive independen­t testing at the Ford Proving Ground in conjunctio­n with Insurance Australia Group.

Cars from Jaguar, BMW and Volkswagen worked almost flawlessly when driven at a regular sedan test target at between 10 and 25km/h, reinforcin­g the usefulness of AEB systems.

However, each failed to consistent­ly identify a ute target that simulates the rear-end style

of cars that account for more than half of all new vehicle sales in Oz.

The head of IAG research, Robert Mcdonald, said it appeared some carmakers were optimising their AEB systems to pass the Euro NCAP test rather than trying to detect a wide variety of real-world scenarios.

The Euro NCAP testing of AEB employs a target ‘vehicle’ that extends to ground level, made of a soft material designed not to damage test vehicles at impact speeds up to 50km/h. But the target also includes various reflectors and other materials to ensure it works with sensors typically used in AEB systems. According to Euro NCAP, the target “is developed to have the radar signature, reflectivi­ty and visual signature comparable to that of a ‘C’ segment vehicle”.

IAG developed its tests to learn more about the potential benefits of AEB with the view to offering insurance discounts.

“We know the European target vehicle doesn’t represent a ute very well,” Mcdonald told Wheels. “The height and the shape is a challenge for some AEB systems.”

Mcdonald said IAG had just added a ‘ute’ target vehicle because “we think they’re more representa­tive of the average vehicles that are in Australia”.

In an era of increased crash avoidance technology – aligned with the developmen­t of autonomous technology – he said it was important to ensure cars could respond to a wide range of often challengin­g real-world scenarios. “We welcome the adoption of AEB by Australian carmakers and want to encourage the further sophistica­tion of the technology.”

Half of the 28-strong COTY 2017 field had at least one model fitted with AEB. All 14 of those vehicles were tested against the sedan and ute target vehicles, with most accurately detecting and reacting to the sedan. But at least seven of the 14 tested failed to detect the ute at least once.

In defending the AEB performanc­e of the F-pace and XF, Jaguar questioned the validity of IAG’S target vehicles, suggesting they may not represent real vehicles. “In conjunctio­n with our supplier we have conducted a series of 15 ‘classifica­tion’ runs in a Jaguar F-pace at various speeds against a Nissan Navara pick-up truck and in all cases the system successful­ly classified the vehicle as a ‘target’ and would have initiated braking,” Jaguar said in a statement.

Mcdonald, meanwhile, said cars that struggled to pick multiple objects often used only one method of detection, such as radar or cameras. By contrast, the Mercedes-benz E-class uses five multi-range radars, stereo cameras and 12 ultrasonic sensors, while the Volvo S90 pairs radar and camera technology to detect objects as diverse as caravans, boats, buses, trucks, bikes, pedestrian­s and large animals. Full COTY results and AEB testing will be published in the February issue of Wheels and on Wheelsmag.com.au on February 2.

 ??  ?? IAG targets are constructe­d from actual cars with the aim of accurately representi­ng a vehicle in the real world
IAG targets are constructe­d from actual cars with the aim of accurately representi­ng a vehicle in the real world
 ??  ??
 ??  ?? Some cars with only one method of detection failed to detect the ute test rig Tests were conducted at speeds of 10km/h, 20km/h and 25km/h
Some cars with only one method of detection failed to detect the ute test rig Tests were conducted at speeds of 10km/h, 20km/h and 25km/h
 ??  ??

Newspapers in English

Newspapers from Australia