It has been common police practice for a long time to use the ratio between the number of interventions and the number of traffic accidents as an indicator of successful work, and as a standard it is usually 20. This "intervention/accident" indicator is justified if the traffic violations and traffic accidents in the relation are correlated, i.e. if they are stages of a common process of endangering traffic.
There is a "probability connection" between traffic accidents and serious traffic conflicts that precede these accidents, and therefore the violations that precede the conflicts are obviously in correlation with these traffic accidents.
The day to day task of the traffic police is to recognize the connection between traffic violations and traffic accidents, i.e. to recognize in general the process that endangers the safety and its individual stages.
In accomplishing this task using of the "traffic conflict techniques" and knowing their theoretical background can be very beneficial. Using these techniques in the recognition of various traffic violations - e.g. violating the right of way - is very important.
On the other hand, establishing the overall nature of endangering traffic safety and establishing the role of various offenders and their interaction, enables the police to establish when and where the traffic violations are "tied to space" and therefore for solving the problem seek the means other than police control and enforcement.
A traffic conflict can be generally defined as such a traffic situation involving two or more participants in the traffic approaching each other in space an time, and in such a way that they challenge a risk of causing a real collision, if they do not change their direction or speed.
Serious traffic conflict or almost accident is similar to this situation, but it implies a minimal period of time remained for undertaking an avoiding action, and only the high intensity action to avoid the risk can bring about the normal functioning of the system.
However, when it remains only a minimal period of time to perform avoidance, the question is, how much time it takes before the moment of possible collision, that is, what period of time to the very collision should be considered to be critical moment?
Unambiguous answer to this question is simply not possible, because, in order to undertake any form of avoiding action - braking, acceleration, swerving - various lengths of time are necessary, therefore, the critical moments for undertaking those actions are also different. However, because the most frequent practice in avoiding is braking, the time to collision is measured according to that type of maneuver.
Swedish definition of time remained to collision is rather widespread (although not universally) accepted so called, that is, a general definition of a serious traffic conflict, in the function of that time, in a following way:
A serious traffic conflict exists in a case when, if during the time since an avoiding action started till the moment of possible collision, both participants kept on moving along the same way and with the same speeds, but not longer than the time of stopping of a vehicle applying an intensive braking on a slight wet pavement, plus a half of second more.
Slightly different, although rather similar, but in any case more simple, is the situation when the time remained to collision could be defined as the time of 1,5 seconds. In line with it, a serious traffic conflict exists when the time remained to collision is less than 1,5 seconds.
A moment of the beginning of an avoiding maneuver is the moment of visible undertaking of that action, that is, visible for some objective by-observer. If it is about braking, it is the moment when the stop lights on a vehicle turn on.
How it can look like in practice?
If, for example, a driver of a vehicle approaching the crossroads, which also approaches another vehicle on the colliding way, starts braking at a distance of 20 meters from the point of possible collision, driving 30 km/h, then its time to collision is 2,4 seconds. When this time is larger than 1,5 seconds, which is a border time (according to more simple definition), a conflict, although it existed, was not a serious one. But, it the driver started avoiding by braking at a distance of 10 meters from the collision point, his time to collision would be 1,2 seconds, which is less than the border time for a serious conflict which is 1,5 seconds. In such a case, it would have been a serious conflict. In that case only an extreme braking - at least 7 meters per square second - could save this situation an avoid a threatening conflict (during available time of 1,2 seconds).
For a quick answer of existence of a serious traffic conflict, namely, for general gradation of an observed conflict, beforehand prepared tables and charts are necessary.
For practical work with traffic conflict techniques it is necessary to observe all traffic situations, and specially those which involve some avoiding actions. That moment should be registered as precise as possible, and particularly for the participant who undertook such an avoiding action. That means that his distance should be assessed from the point of a possible collision, and his speed at that moment as well. Further on, it should be identified which type of a maneuver was undertaken, and possibilities and similar circumstances for the participant who was undertaking such an avoiding action.
A risk of the traffic conflicts for a monitored location can be established by a continuous registration of all (serious) traffic conflicts in a certain period of time, together with registration of all traffic flows, and other characteristic behavior of the traffic participants (traffic violations regardless to their direct consequences).
Because it is sufficient to register conflicts in a certain location some ten hours (several times), it is clear that it is possible, even after a very short time, and again on the basis of comparison before-after situation, that is with an without, to provide a preliminary assessment of the undertaken safety measure, that is, generally, assessment of the risk level on a certain location.
Statistic tests, those identical to the ones used for traffic accidents, because here we also deal with frequencies, are the basis for making an appropriate conclusion.
When applying this technique of evaluation, it is important to take into account the fact that serious traffic conflicts present a good correlate of the accidents, and that these techniques are (mainly) evaluated in a satisfying way as a procedure used for establishing the risk in the road traffic.
In practice, registration of traffic accidents (if methods of automatic video-recording are used) is usually performed on previously prepared protocols.
Trained observers that are able to make sufficiently precise evaluation of speeds and distances, and then for noting all avoiding actions, that is, general characteristics of a traffic situation, are prerequisite condition for word with techniques of the traffic accidents.
The conflict technique has been used in the study of the influence of blinking green at the end of this phase in semaphore program. A video-tape that could be used for educational purposes was produced at the Police College in Zagreb, Croatia.