Source: Applus IDIADA
ADELANTO, Calif. – This is the third of three posts by Alvaro Esquer, Project Manager, ADAS in Applus IDIADA, about a study to determine the means of improving autonomous-emergency braking (AEB) performance in terms of efficiency and driver acceptance.
The first two posts:
TBR Technical Corner: Braking Requirements for Optimizing Autonomous Emergency Braking (AEB) Performance (Part 1 out of 3)
TBR Technical Corner: Braking Requirements for Optimizing Autonomous Emergency Braking (AEB) Performance (Part 2 out of 3)
The aim of this study is to investigate the means of improving AEB performance in terms of efficiency and driver acceptance. The first part of this article explained the behavior of average and professional drivers in emergency situations, which have been then compared with the limits of vehicle’s braking capabilities. The second part, on the other hand, focused on the brake jerk as a method to improve AEB systems performance.
Cars with the highest performance in Euro NCAP active safety tests can avoid crashes in most of scenarios. In other scenarios, the most difficult ones, some of the best cars are only able to lessen the impact speed. Car-to-case scenario was already exposed in the previous (second) part of the article. Other challenging scenarios are introduced in the following section, which could be the target scenarios where to use an improved brake profile application.
Performance of AEB systems in 2018 Euro NCAP tests protocols
Car-to-pedestrian tests were introduced in Euro NCAP test protocols the 2016 with four pedestrian crossing scenarios. Later, in 2018 the protocol was updated with longitudinal and night tests.
In pedestrian crossing tests the test dummy comes from the side and consequently the detection time becomes crucial here. CPNC50 is the test where most of the vehicles are having issues to get the highest score. This is a scenario where the dummy is a running child that appears from behind two parked cars as shown in figure 11.
This is assessed up to 60 km/h and the child is always running at 5 km/h. The AEB performance normally decreases for most of cars from 40 km/h and at 60 km/h shows almost no performance. Figure 12 displays a collection of impact speeds found on various vehicles in the market in the scenario CPNC50 at velocities between 45 and 60.
With the same hazard detection time, just improving the braking profile the impact speed could be decreased drastically.
Next updates on the test protocols includes crossing scenarios where the car is turning to the right or left and the pedestrian is coming from the front or from behind. These are very demanding situations where only an early detection or a quick intervention may be necessary. In such situations the rate of false positive may be problematic and thus braking at the last moment to brake may be the desirable scenario.
Car-to-bicycle tests were integrated in Euro NCAP protocol in 2018 and are to be updated in 2020. One of the most challenging scenarios that take part of 2020 test protocol is the so-called Car-to- Bicyclist Nearside Adult Obstructed scenario and is displayed in figure 13.
The cyclist appears from behind an obstacle at a velocity of 15 km/h and the obstacle’s edge at a distance of 3.55 m from the vehicle’s path. The detection time is very tight and it becomes more difficult as the vehicle velocity is increased as figure 14 shows. There it is compared the time that the bicycle is in the field of view with the time needed to stop the car from its initial velocity, which is referred as the last point to brake. Again two different braking profiles are compared: baseline AEB and the professional test driver brake application.
This chart is a simplification used to highlight the importance of a quick and strong brake response. The truth is that, for testing steady conditions the test car does not need to brake to stop to avoid the crash, probably the accident can be prevented if only reduces the velocity at the correct time due to the bicycle’s high speed.
The braking systems are an essential part on AEB system. Despite the sensorial network is the responsible of detecting hazard situations, without a quick an effective braking intervention the accident may be only mitigated.
An automatic full braking in emergency situation is not the only way that braking systems can support the driver. The same system can also support the driver when it detects a panic but insufficient brake application in proximate collision situations. This support consists of an additional brake pressure to ensure maximum braking capabilities. Collision warnings based on brake impulses are proved to be the most effective driver’s warning strategy. Furthermore, some other ADAS systems like the lane centering system uses brakes impulses to wake up drivers after a while without making any correction on the steering wheel.
The conclusion from this study is that the current AEB systems are not fully using vehicle’s braking capabilities. The reasons for that are not explored in this research as it was not the aim of this study to define the hardware or the components needed to reach the maximum performance. Nevertheless, this room for improvement should not be understood as a criticism to current AEB system, but as an open window for new developments. Euro NCAP test protocols are truly demanding and the performance found in best AEB systems is remarkable. Even in those mentioned scenarios where some impacts occur the level of safety added is positive.
A continuation of this study will cover the investigation of the braking system components that could improve braking performance in AEB systems.
About Applus IDIADA
With more than 25 years’ experience and 2,450 engineers specializing in vehicle development, Applus IDIADA is a leading engineering company providing design, testing, engineering, and homologation services to the automotive industry worldwide.
Applus IDIADA has locations in California and Michigan, with further presence in 25 other countries, mainly in Europe and Asia.