Report numberRA-MOW-2009-010
TitleITS techniques for the improvement of traffic safety around intersections equipped with traffic lights
SubtitleStudy of the possibilities for dynamic speed advise from traffic control installations
AuthorsJohan De Mol
Wim Vandenberghe
Sven Vlassenroot
Koen De Baets
Published byPolicy Research Centre for Mobility and Public Works, track Traffic Safety 2007-2011
Number of pages81
Document languageDutch
Partner(s)VUB, PHL, UGent
Work packageOther: Innovation and technology for safer transportation

This report proceeds on the report “ITS and traffic safety, Intelligent Transport Systems” (RA-MOW-2008-007) and expands on the possibilities of ITS systems for the enhancement of traffic safety around intersections equipped with traffic lights. It seems quite obvious that there is a higher risk for traffic accidents at junctions then on other road segments, since road users cross at junctions. Based on accident statistics, this report proves that enhancing safety around intersections indeed has a strong positive impact on the global traffic safety. It also introduces the use of Intelligent Transport Systems (ITS) as a mean to enhance safety around crossroads equipped with traffic lights. A detailed technical description is given of the communication techniques supporting such systems, and an extensive overview is given of the state-of-the-art in the most relevant related research projects.

Accident statistics prove that an important portion of both fatal and non-fatal accidents occur at intersections. In the EU-13 this represents 5.476 casualties or 21.3% of all traffic casualties. In Belgium 19.3% of all traffic casualties are the consequence of incidents at intersections. When the absolute figures are divided per million inhabitants, Belgium has the fifth highest number of deaths at intersections in Europe (after Italy, Hungary, Estonia and Poland). This Belgian number, 20 lethal accidents per million inhabitants, is higher then the average EU-13 (18.1) and EU-16 (19) numbers. In the Intersafe project (performed within the PReVENT project) several accident scenarios where studied, and among others the percentage of intersection fatalities within the total of traffic casualties were determined for France, Great-Brittan and Germany. Junctions are responsible for 30 to 60% of the incidents with wounded, and 16 to 36% of the fatal incidents.

When examining the analyses of red light negation, it can be concluded that the critical situation can be avoided if the driver is informed and warned earlier. However, it has to be guaranteed that this information and warnings do not lead to an even more dangerous risky behaviour where they are interpreted as “if I drive faster I will be able to just catch the green light”. This would create an even higher risk of serious accidents, therefore enough attention should be given to this issue.

The above numbers and analyses prove that enhancing safety around intersections should be a priority in traffic policy. Extra attention should be given to preventing red light negation, since this is a major part of the dangerous traffic violations.

Research focusing on accidents on intersections proved that there is a connection between the type of intersection and the degree of traffic safety on that junction. A study was found that provides a quantitative evaluation of traffic safety around different types of intersections, expressed as the number of registered incidents with wounded per million passing vehicles. This number is the highest for intersections with traffic lights (0.11), followed by junctions on a main road (0.09), junctions with right of way (0.09), roundabouts (0.07) and junctions without right of way (0.06). The average number of wounded per accident decrease in the following order: traffic lights, main road, right of way, no right of way (respectively 1.22, 1.18, 1.11, 1.09). The seriousness of accidents, expressed as the number of hospitalizations rather increases in that order (respectively 17,16,18,18).

This means that rearranging intersections can be beneficial. An American study investigated what the effect would be if the junctions in North Virginia were rearranged. It was concluded that delays would be lowered with 62 to 74 % (according to junction type), meaning a reduction of 300.000 lost hours per year. The annual saving in fuel consumption would be 200.000 gallons (757.000 litres). Traffic safety would increase drastically: reforming junctions into roundabouts would result in 62 less accidents and 42 less wounded (comparison between 1993 en 2003 with five crossroads for which accident statistics were available).

Therefore a logical measure to enhance traffic safety is to rearrange intersections equipped with traffic lights or to transform them into roundabouts. Other possible measures can be related to traffic lights regulation, road layout (canalization, slopes, facilities for vulnerable road users, etc), improvement of visibility, driving education, speed management around intersections, enforcement (camera’s), road surface, etc. These measures are already applied in Flanders today. But a technique that is almost entirely neglected is the employment of Intelligent Transport Systems to increase traffic safety around intersections. This approach is further elaborated in this report.

When traffic control infrastructure at crossroads is extended with intelligent software and possibly sensors and means to communicate with neighbouring vehicles, applications can be developed with a positive impact on different domains. The three most important ones are traffic safety, traffic circulation and the environment.

Many studies focusing on intelligent intersections aim to enhance traffic circulation. Self-organizing traffic lights divide traffic into platoons by counting (e.g. using counter loops in the road surface) the number of vehicles waiting at the traffic lights, and adjusting the switching times accordingly. This technique was applied in a traffic simulator to the Wetstraat in Brussels, where it would lower the total travel times approximately 25%. In other research a system was developed where every vehicle can vote for switching of the lights. For this it communicates it’s identity, direction, position and place in the queue to the traffic light. Using this information of all the neighbouring vehicles, the light can calculate which light switch will result into the greatest total profit for all vehicles. Results showed an enhancement in average waiting times from 30 to 50%.

In the domain of positive effects on traffic safety, developments are taking place in a number of European research projects such as PReVENT, Safespot and VII. Frequent scenarios that are being tackled are avoiding of or warning for red light negation, avoiding accidents with vulnerable road users and coordination of turning left with oncoming traffic. Also, applications that aim for an enhancement in traffic circulation imply an enhancement in traffic safety.

Less research can be found aiming at environmental benefits, but again applications focusing on enhancement of traffic circulation imply positive effects on the environment.

A common aspect of many of these applications is the fact that they rely on communication technology. This can be divided into three major groups: local short-range communication, cellular data networks and digital broadcast technologies. When they are studied in the scope of intelligent intersection control, then both broadcast- and cellular technologies do not qualify as a possible supporting technology. This because of the one-way communication character of broadcasting, and the higher delays and end user cost of cellular data networks. This limits the choice to local communication media, more specific CEN DSRC, IEEE 802.11p, CALM-M5, CALM-IR and IEEE 802.15.4. When taking a closer look at their parameters, they indeed seem to be very suitable: they are interactive, free to use, can offer a high bandwidth and are not dependent of network operator coverage. An extensive technical description of these communication technologies is given in this report.

CEN DSRC is typically used for Electronic Toll Collect (such as Télépéage in France). However it is not suitable for the implementation of intelligent traffic lights because it only supports one-way communication. IEEE 802.11p is an amendment to the well-known IEEE 802.11 Wireless LAN technology (also known under the Wi-Fi hallmark) for use in vehicular environments. This technology is also not suitable for intelligent traffic lights in Flanders since it operates on the ITS frequency bands of the US. CALM-M5 however is the European derivate of IEEE 802.11p, and this technology indeed is greatly suitable. But just like IEEE 802.11p it will suffer van scalability issues, meaning that more research regarding scalable routing protocols has to be conducted before it can be used in a real rollout.

In the field of directional communication both European standards CALM-IR and CALM-MM qualify as an implementation candidate. CALM-IR is a communication standard based on infrared light, en is very good in sharply defining communication zones. CALM-MM operates on frequencies similar to radar, and can provide very high bandwidths. On short to medium term CALM-IR has the advantage that it is already much further developed then CALM-MM.

IEEE 802.15.4 is a communication technology used in wireless sensor networks. Its main characteristics are energy efficiency and scalability. This technology is most suited for application in mobile devices, thus for including vulnerable road users in the intelligent intersection. However this requires that the (typically SANET) routing protocols on top of this technology support mobility. This demands further research.

To conclude, on short to medium term three communication technologies qualify for the implementation of intelligent traffic lights: CALM-M5, CALM-IR and IEEE 802.15.4. CALM-M5 is suitable for omnidirectional communication with vehicles, CALM-IR for directional communication with vehicles and IEEE 802.15.4 for omnidirectional communication with vulnerable road users. Further research is needed regarding suitable routing protocols before CALM-M5 and IEEE 802.15.4 can be successfully applied in intelligent intersections.

Based on these technological developments, several research projects already investigate intelligent intersections.

INTERSAFE is a subproject of the PReVENT project. Goal is to inform and/or warn the driver about traffic lights; this information/warning contains information regarding the time and the proper speed to safely cross or exit the intersection. The idea is to avoid conflicts at junctions with these information/warnings. These conflicts can be caused by absent-mindedness (not noticing the traffic light or the state of the lights), by maladjusted driving behaviour in function of the expected red- or green cycle, or by an inadequate insight in the traffic lights installation. Finally, it is expected that this information/warning will stimulate the driver to adjust his driving behaviour, reducing the risk for conflicts. This adjustment of the driving behaviour mainly is related to a decrease in speed, this can be performed gradually, but can bend to severe braking if the information/warning is not taken into account on time.

Two demonstration vehicles were designed for testing this technique. A visual and auditive warning was used to give speed advice. Complying with this speed implies that the intersection can be safely crossed. Demonstrations at Versailles proved that this systems works well on the test roads from a technical point of view. However, the nature of the provided information and the voluntary character of the system imply that the safety on these intelligent intersections is highly dependent of they way the driver translates the information/warning into adjusted driving behaviour. From this research project, it appears that if some specific technical enhancements are conducted, supporting driving behaviour at intersections can improve traffic safety.

IRIS is a sub-project of the Safespot project. IRIS uses vehicle-infrastructure communication to analyze the movements of all individual vehicles, and laser scanners to identify vulnerable road users. Based on these inputs, the system can assess dangerous situations on time and take necessary measures to avoid accidents (such as adjusting switching times of the lights or sending warning messages to human-machine interfaces in the vehicles using wireless communication). The IRIS system focuses on three scenarios responsible for a major part of the accidents at intersections: red light negation, turning left (conflict with oncoming traffic) and turning right (conflict with vulnerable road users).

In the project experiments are performed in a driving simulator. Simulation makes it possible to perform a more systematic and extensive analysis of the applications and their possible variations. Early during development simulation can be useful to study the timing of the applications, derive optimal parameter settings and assess the potential impact of the applications if validation in the real world is not possible. In parallel with the simulation work, a real IRIS system is built and tested in the project. In May 2009 a first public demonstration of this system will be given Helmond, The Netherlands.

Aiming to improve traffic circulation and to reduce consumption and emissions, the project Tovergroen was set up in The Netherlands. Tovergroen is a system to detect trucks and give them priority if possible by prolonging their green phase. However, the detection system performed inadequate, not recognizing trucks or recognizing the wrong vehicles as trucks (campers, cars with trailers, etc). Despites these problems Tovergroen increases the chance that trucks do not have to stop with 5 to 10%. Tovergroen decreases red light negation of heavy traffic with approximately 30%. It is expected that the positive effect will be even greater using a more reliable detection system.

In cooperation with the Technical University of Munich, Inglostadt en GEVAS software, Audi has developed Travolution. Aim is to inform drivers regarding the appropriate speed to maintain to cross the intersection without stopping. Using wireless communication, the intelligent traffic light sends the duration of the red light to the in-vehicle system. The in-vehicle system then calculates the appropriate speed to catch the green light and informs the driver. In Inglostadt 46 intersections were equipped with this system, and two test vehicles were used. In the next phase, this experiment will be extended with 20 cars and an additional 50 intersections. Goal is to investigate how the optimization of traffic light controllers in urban areas can decrease pollution and travel times.

At the ITS World congress in New York, several other applications were demonstrated where information is communicated from traffic lights to vehicles. Although there is almost no information available regarding these demonstrated applications, a short description is included in this report.

DownloadPDF icon RA-MOW-2009-010.pdf


The Policy Research Centre for Traffic Safety carries out policy relevant scientific research under the authority of the Flemish Government. The Centre is the result of a

cooperation between Hasselt University, KU Leuven and VITO, the Flemish Institute for Technological Research.


Leuven vito