Project start
Less stop more go: Wuppertal researchers set their sights on driver assistance systems
The tail lights light up: constantly braking and accelerating in traffic is not only annoying, it also harms the environment. // Photo Colourbox
Set cruise control and then: clear road ahead! Adaptive cruise control (ACC) systems are now standard equipment in vehicles. It is hoped that their sensors will increase road safety by reducing the influence of human error such as distraction at the wheel or making rash decisions. Equally tempting is the idea that computerised systems will be the saviour against annoying traffic jams.
However, experiments carried out in the past with vehicles travelling behind each other in assistance mode show that the ACC systems currently in use fall short of these expectations. The problem: due to time delays in signal transmission, they are not sufficiently responsive to keep the gaps between them large enough to compensate for changes in speed in the flow. After a while, the traffic in the experiment on the circuit comes to a standstill and the collective of vehicles becomes unstable. The result: stop-and-go waves that eventually mature into traffic jams; in the experiment, these are even more massive than in normal road traffic. Better systems are needed on the road to fully autonomous technologies that protect us from accidents and standstills.
Getting a grip on unsafe factors
This is where the researchers from the Chair of Road Safety and Reliability at the University of Wuppertal come in: Prof Antoine Tordeux and colleague Dr Raphael Korbmacher are working on new, robust algorithms in the SmartACC project - clear rules, in other words, for the mathematical models on which ACC systems operate. "The basic idea behind the further development of the models," explains Tordeux, "is that maintaining a constant time gap between vehicles keeps the traffic system stable."
The strategy as such is not new (see infobox "Technology & Strategy"), but the Wuppertal researchers are now going one step further by developing models in the sense of an "adapting time gap" that are resistant to various relevant influences. In addition to the reaction time factor, the consideration of more or less random, not always predictable events, such as in the case of the weather when heavy snowfall disrupts the sensor technology, is considered particularly challenging. Scientists refer to this as stochastic noise.
Stop-and-go traffic waves often lead to major traffic jams. Modern assistance systems for cars should help to avoid this in the future. // Photo Colourbox
Tests with 40 mini-cars
In order to gain a better understanding of the interaction of these influences, the researchers are working with computer simulations. This allows them to adjust the algorithms with increasing precision. In addition, they also want to start experiments to test the new models - with up to 40 mini-cars controlled by the algorithm. "In the best-case scenario, these will then drive much longer without malfunctions than was the case in previous experiments," summarises Korbmacher. Finally, the scientists want to put themselves and other test subjects in the driving simulator and, with the support of a virtual reality application, test whether the algorithms developed can also keep up with the not entirely unimportant issue of driving comfort - in other words, whether they do not cause the vehicle to accelerate too much and then brake abruptly again.
If the research turns out to be successful, this would be an indication of how autonomous driving can contribute to greater safety and well-being in the future. "On the road, optimised systems could make it easier to avoid rear-end collisions. By avoiding traffic jams, everyone involved would have more time and would also benefit the environment. Because fewer braking and acceleration cycles in slow-moving traffic mean less fuel consumption," says Korbmacher.
Technology & strategy behind the ACC systems
ACC systems are based on radar distance measurements to the next vehicle, vehicle following models from traffic engineering and control technologies from electrical engineering.
The first market-ready ACC systems were introduced by leading car manufacturers at the end of the 1990s. Initially they were considered luxury equipment, but today they are standard. In scientific research, ACC systems are viewed from different perspectives. A central topic is the question of which traffic control strategies are most effective in maximising safety and comfort. The most promising is the constant time-distance strategy, in which the distance between vehicles varies depending on the speed. It ensures that greater distances are maintained at higher speeds, resulting in a smoother flow of traffic and a lower risk of accidents. This is also the strategy recommended by safety standards such as ISO 15622. This assumes a time interval between two vehicles of between 0.8 and 2.2 seconds. The ADAC recommends 1.8 seconds in its guidelines.
The project "Intelligent control for safe roads: Development and testing of advanced adaptive cruise control systems" (SmartACC for short) is being funded by the German Research Foundation for two years with around 285,000 euros.
