Study method for pedestrian behaviour in the area of pedestrian crossings located at tram stops

. The paper presents the problem of pedestrian safety at stops, with stops designed for opposite directions. Stop areas with a predominantly pedestrian traffic, where pedestrians cross tram tracks, are not protected by traffic signals. The article presents a method for conducting a study to identify types of pedestrian and tram driver behaviour, developed on the basis of a pilot study. This method helps to evaluate, inter alia, the tendency of pedestrians to risk going across the tram track before an oncoming tram, depending on how far it is from the pedestrian crossing. It also helps to determine the probability of the tram slowing down or emergency braking when crossing the pedestrian crossing, and the probability of difficulties with the departure of the tram, caused by pedestrians.

footways with pedestrian streams crossing stops when the trams are arriving even if they do not want to use the tram. The key role in this sense is played by the diversity of pedestrian behaviours and variable awareness of the traffic and braking capabilities of the trams.
The present study discusses the issues related to the comfort and safety of using transport infrastructure in the potential conflict zone between pedestrian and tram traffic. The original proposal for the study methodology includes road user behaviours (pedestrians and tram drivers) in the vicinity of pedestrian crossings located near tram stops. Conflicting interests of vulnerable road users (safety and comfort) and passengers of public transport (comfort and speed) must be reconciled in these areas.

Current approach to the study of conflicting situations between pedestrians and trams
The number of accidents and collisions that took place on individual tram stops is usually not relevant, as the accidents take place relatively rarely [7]. Moreover, studies of hazardous situations enable a description of atypical or temporary solutions -which should be tested prior their eventual dissemination. On the basis of the definition accepted in 1997 in Oslo, it can be stated that "pre-conflicting road user behaviour is a situation, when the users violate the established thresholds for road behaviour, yet which do not result in a road accident, and which do not force a change in the behaviour of other road users in order to avoid the accident" and "traffic conflict is a situation, when at least one of the users takes action to prevent an accident". The basic indexes evaluating traffic conflicts include: -CS (Conflicting Speed) -this is the speed of road user, who began evasive actions (just before starting the action) [8]; -TA (Time to Accident) -time which passes from the taking of an evasive action by at least one road user to the moment of a potential accident, if both road users continued movement at the same speed and the same direction; -TTC (Time To Collision) -time to the collision of two objects, if their trajectories and speeds remain unaltered. The lower value of TTC means a higher likelihood of the accident with more serious consequences [9]. However, the above indexes are more appropriate for the study of "vehicle-vehicle" conflicts than for "vehicle-pedestrian" conflicts [10], as they concern independent vehicles already in motion. Whereas in the case of "vehicle -pedestrian" conflict, particularly within stops, the conflicting situations often stem from the incorrect evaluation of the possibility to enter the track (pedestrian) or incorrect evaluation of the possibility to leave the stop (tram driver).
Therefore, techniques for the study of road user behaviour have been introduced and enhanced. They are typically conducted in a traditional form, with observers detecting conflicting situations and entering the types and numbers to the questionnaires, often as supplementary research to pedestrian and vehicle traffic measurements. Currently, a video technique is increasingly commonly used for this purpose enabling a lower number of participants to the observations and offering better accuracy of the conducted study. However, independently of the technical feasibilities of research, selection of the observed behaviours and correct, unambiguous definition of conflicting situations remain the most important issue. It largely depends on the approach of the researchers. A major step in the development of traffic conflicts was the research conducted within the project CiViTAS CARAVEL [11], implemented in order to enhance the attractiveness, comfort and safety of city transport. In the study, conflict measurements included three types of stops: with passenger exchange in the lane, with elevated lane (so called Vienna-like tram stop) and with elevated lane and island platform. The study did not analyse opposite stops located in areas with shared tram-pedestrian function.

Proposed method for the study of conflict situations between pedestrians and trams within the area of opposite tram stops
Conflicting situations between pedestrians and trams within tram stops typically take place while pedestrians are crossing tracks -and the proposed study will be most relevant in this field. Moreover, only crossings without traffic lights were included. In the proposed method, the level of safety is assessed using the analysis of frequency and potential outcomes of hazardous situations (traffic conflicts). For obvious reasons, the method must comply with the current legislation including the rules on priority. These rules are discussed in Article 13 of the Road Traffic Act [12]; they demonstrate the obligation to cross the track separated from the traffic lane only in the designated area, and where a passenger island situated on a public transport stop is connected to a pedestrian crossing, reaching and leaving the stop is allowed solely on the crossing. However, the most important issue is that the pedestrian is obliged to take special care prior to entering the traffic lane. While on the crossing, the pedestrian has priority over all vehicles, including trams. However, the pedestrian does not have priority over a tram while waiting to enter the pedestrian crossing, which is regulated by Article 14 of the above Act, prohibiting (among others) stepping onto the lane directly before an approaching vehicle, including pedestrian crossings. An issue to be addressed is evaluating the safe distance from the tram while entering the pedestrian crossing. The pedestrian is not allowed to step onto the tracks before an approaching tram, whereas the tram driver approaching a pedestrian crossing must maintain special attention and travel at a speed enabling him to stop before the crossing if there is a pedestrian on the crossing.
The proposed method includes the general behaviour of pedestrians who cross tracks as well as the behaviour of tram drivers. In principle, the method is to be used for determining the use of infrastructure dedicated to pedestrian traffic, as well as for determining the frequency of traffic conflicts to decrease their prevalence. It is also intended to help with determining the causes of hazardous behaviour of pedestrians and tram drivers. In the proposed method, pedestrian traffic volumes are recorded on the entire width of pedestrian crossings and outside them.

Example of the method in practice: study of conflicting situations within the tram stop "Dworzec Główny Zachód" in Krakow
The study of conflicting situations was performed following the presented methodology in several stop areas in Krakow. Figure 2 shows an example of the shares of passengers crossing the track in one of the analysed pedestrian crossings in relation to approaching trams.

Fig. 2. Share of individual pedestrian behaviours on the analysed pedestrian crossing.
As can be seen, about 1% of pedestrians behave very riskily, in addition, to nearly 8% trying to pass when the tram is 10-20 m from the pedestrian crossing, which can also cause conflicts. And this "distance dilemma" will be particularly closely studied in the future. In this case, the tram track is located in the pedestrian area. Despite being separated, pedestrian and cycling traffic does occur there sporadically. It should be noted, however, that in reality almost the entire space between the above pedestrian crossings does not constitute a spatial barrier -and it is used by pedestrians to cross the tracks. It can also be noted that the share of passages outside of designated places is significant, amounting to 20% and that it varies depending on the direction and time of day. The drivers respond by using sound signals (7-16%), decelerating significantly (4-11%) or even emergency stopping (0-4%). This means that the drivers are prepared for unexpected behaviour of pedestrians and drive at safe speeds.

Conclusions
Measurement of hazardous situations occurring at stops and crossings requires further discussion. At this stage of analysis, other features were not investigated such as weather conditions, pedestrian age, visibility and parameters of pedestrian crossings. The absence of such wider studies means that the presented methodology and results should be considered as a pilot. It is important to define the safety and hazardous situations. Pedestrian behaviour is influenced by the location of sources and destinations of traffic and footways. The presented method can be -with some adaptation -used for the study of conflicting situations between trams and bicycles.