key: cord-0059043-o4tmc03p authors: Campisi, Tiziana; Basbas, Socrates; Tesoriere, Giovanni; Canale, Antonino; Vaitsis, Panagiotis; Zeglis, Dimitris; Andronis, Charilaos title: Evaluation of Pedestrians’ Behavior and Walking Infrastructure Based on Simulation date: 2020-08-19 journal: Computational Science and Its Applications - ICCSA 2020 DOI: 10.1007/978-3-030-58802-1_53 sha: 47b92778f5ed70e743d24ba6f00d033652e4c5d8 doc_id: 59043 cord_uid: o4tmc03p Sustainable mobility mainly refers to cycling, walking, e-scooters, public transport etc. The increase of the usage of these means of transport against the use of motorised vehicles can improve the overall quality of the urban environment. Pedestrian streets play a significant role towards the direction of sustainable mobility and the facilitation of environment friendly daily trips. This research concerns one of the most important pedestrian streets in the centre of the city of Thessaloniki, Greece. The evaluation of the existing situation in the pedestrian street together with the evaluation of various scenarios concerning changes in the behaviour and direction of movement of pedestrians due to incidents is examined in the framework of this paper. The evaluation took place by using the pedestrian simulation software PTV Viswalk and four different scenarios were tested. The first scenario is the base scenario referring to the existing situation. The second scenario deals with an increase in the pedestrian flow due to unexpected events. The other two scenarios refer to evacuation phenomenon, thus preventing access to part of the pedestrian street. The examination concerns the pedestrian Level of Service (LOS) and the identification of critical segments of the pedestrian street in terms of pedestrians’ flow and composition. The simulation results show that even in the case of doubling of pedestrian flow, this will not cause an overall significant drop in LOS, except for specific sections of the pedestrian street. These results can assist the authorities so to minimize walking difficulties in the pedestrian street and at the same time to take care for waiting areas and the provision of optimum evacuation routes. Transportation systems can be assessed in a variety of ways. Evaluation methods take into consideration various parameters including users, transport modes, the land use system, transportation problems and solutions, mobility patterns, performance indicators etc. There are many methods to examine the performance of the transport system, each of which reflects particular perspectives regarding "who", "what", "where", "how", "when" and "why". Different methods favor different types of users and transport modes, different land use models and different solutions to transport problems. Motorized traffic is easier to be examined, but this approach only considers a narrow range of transport problems and solutions. Mobility is more difficult to be examined, since it requires monitoring of people's travel behavior [1] . Pedestrian mobility is still not understood by everyone as a possible solution to the problem of traffic congestion and therefore as an effective and efficient mode of transport for short and medium distance travel. Accessibility is more difficult to be examined, since it requires taking the land use system into account, but it reflects the final goal of transport more precisely and allows the widest range of problems and solutions to be considered [2] . Evaluation can take place in existing infrastructures or it can refer to hypothetical scenarios in order to mitigate the critical issues, especially those which are associated to incident management. This paper focuses on the evaluation of the Level of Service (LOS) of a pedestrian street which is located in the center of the city of Thessaloniki, Greece. There are several studies which investigate the LOS in pedestrian streets using microsimulation techniques in the city of Thessaloniki [3] [4] [5] [6] . In other studies the views of the pedestrians in relation to the infrastructure that they use is examined. Such views are examined in the cities of Trikala and Thessaloniki, Greece [7, 8] . The examination of the environmental impacts due to pedestrianization schemes is also an important aspect which must be considered. Similar research was carried out in Greek cities, including the city of Thessaloniki, Greece [9, 10] . The pedestrian street which is examined in the framework of the present paper attracts a large number of visitors on a daily basis since it includes an important archaeological site and a large number of commercial activities. The LOS was calculated considering four different scenarios in order to identify the potential critical points as far as the movement of pedestrians during an incident is concerned [11] . Collected data include pedestrian flow, pedestrian speed and pedestrians' personal and trip characteristics. The evaluation took place through the use of the pedestrian simulation software PTV Viswalk [12] . Several studies in the literature discuss various collective phenomena observed in pedestrian crowded situations. This issue has recently attracted the interest of a rapidly increasing number of scientists. In this aspect, the present research focuses on pedestrian dynamics in normal and in the presence of an incident situation. The acquired data were necessary to construct the Origin-Destination (O-D) matrices which together with the cartographic background are perquisites for the implementation of the pedestrian simulation software PTV Viswalk [12] . The models were calibrated using PTV Viswalk to reflect the "real" movement and behavior of pedestrians by adjusting different possible parameters in an open and at the same time confined area [13] . PTV Viswalk was implemented considering the social force model that governs the interaction between individuals [14] . Vehicles and pedestrians are simulated as single objects within a larger system [15] . The movement of pedestrians cannot be predictable, therefore compared to vehicle movement, it is more complex to simulate these movements. The PTV Viswalk software was developed in addition to PTV Vissim software [16] , with the aim of simulating and analyzing pedestrian flows, more realistic than it is possible using only PTV Vissim software. When PTV Viswalk is used separately, it is not possible to simulate vehicular traffic. This paper aims to determine the LOS of walking facilities, considering the perceptions of pedestrians. For this work, an investigation of the pedestrian flows was carried out in the Gounari-Navarinou pedestrian street, Thessaloniki, Greece. Monitoring was conducted between the 15th and the 30th of May 2019, during the morning and afternoon period when the market and the surrounding activities are in function. In addition, a questionnaire-based survey was conducted in September 2019 in order to identify the preferences and intentions of the pedestrians (Fig. 1) . Following the Helbing model and pedestrians' LOS definition it is possible to obtain the pedestrian densities and the respective LOS for the various subareas of the pedestrian street. Many parameters can be adjusted in PTV Viswalk to calibrate a model to reflect reality as possible as it can be. These parameters include the speed of different users, the composition and the variation of the pedestrian flow. This research shows a comparison of different scenarios where the composition of the flow remains constant while the flow is increased, and the movement conditions changed from normal conditions [18, 19] to evacuation conditions due to an incident, with restriction in the infrastructure imposed by the presence of an obstacle [20] . Through these comparisons it can be investigated what will happen in the event of an incident. More [17] specifically how bottlenecks in the movement of pedestrians can be avoided and how people can move away. As already mentioned, this work focused on the assessment of the pedestrian level of service (PLOS) relating to a section dedicated to exclusive use by pedestrians. The section runs along some areas characterized by monuments and archaeological excavations; therefore, the section is in use by both tourists and residents. Four different scenarios were considered in the framework of this research. Scenarios 1 and 2 considered the presence of the two origin and destination nodes (Fig. 2 , nodes A and B) together with five secondary nodes (Fig. 2 , nodes C, D, E, F, G) that correspond to the accesses of the secondary roads that intersect with the studied stretch. During the evaluation of the current flows (scenario 1) and the hypothesis of doubling the flow (scenario 2), pedestrian composition was considered unchanged. In both cases, the standard travel speeds related to the gender and age groups of pedestrians were considered. Furthermore, a hypothesis has been made to analyze two non-standard conditions and to evaluate the level of service of the infrastructure in the event of an evacuation. This evaluation is important since it allows to mitigate the impacts on people passing through in the event of a hypothetical closure of some parts of the infrastructure in question due to an unexpected incident. Therefore, it was hypothesized to consider access and exit from the under-study section, excluding the possibility of access/exit from node E (3 rd scenario) and access/exit from node F (4 th scenario). Furthermore, in the 4th scenario, it has been hypothesized that the stretch of the road bordering the archaeological area had suffered a narrowing due to a possible maintenance of the infrastructure. These nodes were respectively closed to the pedestrian flow as their position is strategic with respect to the areas with the highest number of tourists and residents near the archaeological area. In the case of an evacuation (scenarios 3 and 4), a higher travel speed was hypothesized caused by the instinctive response of people to move from the examined area to the exit nodes. Table 1 presents the four scenarios. Evacuation scenario with restricted area in the final part of the pedestrian street (node F on secondary direction). Increased pace/rate for evacuation Fruin (1971) has defined different levels of comfort for pedestrian movements based on macroscopic magnitudes [21] . The concept of Level of Service in terms of comfort has been defined as a criterion for public safety in infrastructures. LOS connects different flow qualities with ratios of the maximum capacity of a structure. The capacity of a facility is defined as the maximum sustainable range at which people can reasonably be expected to cross a uniform point or segment of a lane during a specified period, usually in individuals per hour. Each service level represents a range of operating conditions in which level A represents the best operating conditions and level F the worst. Fruin applied his calculations to urban environments like city streets in normal conditions. As mentioned above, an investigation of the pedestrian flows was carried out in the Gounari-Navarinou pedestrian street, Thessaloniki, Greece. Thessaloniki is the second biggest city in the country with a population of around one million inhabitants. The specific pedestrian street is considered as on one of the most important pedestrian streets in the city center. The under-study area is characterized by the presence of an important archaeological site and various commercial activities. This area of the pedestrian street extends from Tsimiski street to Egnatia street. These two streets serve large traffic volumes on a daily basis. The main nodes on the investigated pedestrian street are presented in Fig. 2 . The total width of the pedestrian street is around 20 m for a representative cross-section. Street furniture, green areas and shy distances are around 14.5 m in total. Therefore, the active width used by the pedestrians is around 5.5 m. There are four cases along the pedestrian street, where there are conflicts between the vehicular traffic and the pedestrians (please see nodes A, C, D and B in Fig. 2 ). It must be mentioned at this point that there are traffic lights at conflict points (nodes) A, D and B, so pedestrian movement is characterized by acceptable level of safety and comfort. Low traffic volumes together with traffic calming measures characterize conflict point (node) C, thus providing a rather safe environment for the pedestrians. Parking is not allowed along the pedestrian street except for emergency and service vehicles. In addition to the evaluation of the functional characteristics of the pedestrian street, an estimate of pedestrian flows was made based on field counts. Estimations were also made concerning the age of the pedestrians. Such monitoring is necessary in order to be able to calibrate, as close as possible to real conditions, the characteristics of the composition of pedestrian flows. Pedestrian actual flows are presented in Table 3 while pedestrian actual flows in terms of pedestrians' estimated age are presented in Table 4 (counts made between 15 and 30 of May 2019). Under study area: Gounari-Navarinou pdestrian street and respective nodes Source of the map: [17] It must be mentioned at this point that data were collected for two weeks and thus it became possible to identify that the peak flow corresponds to 12:00-13:00. The data presented in Tables 3 and 4 as disaggregated by gender and age groups and as totals (as shown in the different columns of Tables 3 and 4 ). The pedestrian parameters related to each scenario are presented in the following Table 5 . Pedestrian dynamics modeling and their implementation in a computer is a demanding process which requires knowledge of transport and computer simulation. The purpose of this paper is to carry out an evaluation through a microsimulation tool by comparing different scenarios. In particular, the fourth scenario involved a reduction in the walkable area due to a hypothetical neighboring maintenance; this area is also characterized by the presence of an archaeological site that attracts tourists. Therefore, this scenario requires simulating an evacuation condition in a reduced infrastructure condition. The area subject to this hypothesis is shown in Fig. 3 hereinafter. For the simplicity of calculations, the infrastructure is divided into subareas and in particular the Gounari-Navarinou pedestrian street has been divided into 63 sections or subareas (Fig. 4) . The software for the estimation of the pedestrian LOS requires the acquisition of data related to the composition of pedestrian flows (age, gender and presence of wheelchairs or disabled people) and their respective speeds. The survey results showed that over 75% of the pedestrians are not accompanied. 12% walk together with other pedestrians (as a group), while the remaining part is constituted by women with children or prams. Almost 60% of the observed pedestrians move in the Gournari-Navarinou pedestrian street for leisure activities, while the rest walks for working purposes or to return home. Over 20% of them visited the under-study pedestrian street at least once a day. These results were concluded by the questionnairebase survey in the pedestrian streets of the under-study area. The pedestrian flow has been evaluated during the peak hour period, where the flow was equal to 2166 ped/h (at 12:00-13:00). The LOS evaluation of the under-study pedestrian street was based on the analysis of four different scenarios, considering the current situation (base scenario) and an increasing percentage of flux linked to the study area. The speed of pedestrians is based on a hundred counts and it is equal to 1.35 m/s. A reduced speed has been considered for the elderly as well as for women with children and for people with wheelchairs. The beginning and the end of the pedestrian area are marked respectively by the letter A and B (please see Fig. 2 ). The model was calibrated in the framework of the simulation process by considering the different parameters relating to the presence of different categories of pedestrians and their mutual attraction/disturbance. More specifically, different speeds assumed or presence of area narrowings or movement priorities were considered. It is considered as appropriate step not to use the default values but to calibrate the model considering the counts made and the lifestyles of the pedestrians as recorded during the questionnaire-based survey. Table 6 shows the variability of the LOS in accordance with aforesaid HCM2010 standards, describing with chromatic scale evaluation useful for easier reading of the subsequent results shown on the map. Minimum and maximum values of LOS used in the microsimulation process are presented in Table 6 . The microsimulation has allowed comparing the scenarios assessing pedestrian speed and density. The results concerning LOS are presented in Fig. 5 . The study was conducted considering a series of hypotheses for the calibration of the scenarios to first study the problems connected with both the variation of the flow and a possible geometric variation of the pedestrian area. It is clear that the scenarios do not consider all the combinations of admissible solutions but those most frequently expected, therefore the work presented is by way of example but not exhaustive. In particular it is the variation of service level shown with scenarios 1 and 2 doubling the pedestrian flow, and also through scenarios 3 and 4 proposing a variation, or rather a narrowing, and analyzing the critical issues that have appeared. The simulation layouts show a reduction in the level of service near sub-areas 14 and 17 in correspondence with the intersection with the street during scenario 2 of doubling the pedestrian flow, corresponding to vehicular traffic in Ioannou Michail street (Node C). The third scenario involves a reduction in the level of service in the smaller areas due to the formation of a bottleneck that slows down pedestrians (sub areas 40-43-46-49). At the same time, there is a reduction in LOS from A to C in correspondence with scenario 4 along the section which is geometrically reduced (brown area) and adjacent to the archaeological area (considering areas 34-37 and 38). Generally, the work reveals a consistent variation of LOS only near the restricted areas or with reduced geometry and this is strictly connected also to the variation of speed and the variation of flow. The comparison of the scenarios has been included in Table 7 in terms of perceived density, average speed and speed variance. In particular, the results show that in scenario 2 with a doubling of the flow there is an increase that [17] goes beyond the doubling for density but an increase of less than 10% for the speed variation has increased pedestrian outflow (due to the dangerous situation). There are roughly similar variations in scenarios 3 and 4. Sustainability and accessibility in the field of transportation can be an incentive to construct pedestrian-type areas, allowing vulnerable road users to use them taking into account comfort and safety criteria. It is well-known that mode choice is strongly related to the subjective well-being [23] and this is the reason why pedestrians' perception of comfort and safety aspects plays an important role. The study area has been selected for its high historical and cultural value but also for its value as a reconstructed pedestrian street and also for the presence of numerous commercial activities. Through an in-situ investigation which allowed monitoring of pedestrian flows and through the microsimulation it was possible to compare different scenarios by observing the changes in the pedestrian Level of Service. The selected scenarios aim to highlight the variation in the behavior of pedestrians in terms of speed and the formation of bottlenecks as the infrastructural conditions of the standard type or evacuation or even narrowing of an area to be evacuated vary. The simulation of several case studies with a variation of the pedestrian composition allows to preventively knowing the areas where critical issues can arise. In addition, the simulation allows preventing this by implementing a series of actions in advance such as infrastructure improvement or reduction in the presence of obstacles or the control and regulation of flows in some areas. Further research with more extensive data collection is needed in order to examine more complicated scenarios and to assess the behavior of pedestrians under critical situations. The present research was based on counts made by observers in the field. Future research must take advantage of video recording for the pedestrians' trajectories. Pedestrian Facilities Users Guide: Providing Safety and Mobility Accessibility, urban design and the whole journey environment Sensitivity analysis and the alternative optimization of the pedestrian level of service: some considerations applied to a pedestrian street in Greece Evaluation of LOS at pedestrian streets and qualitative factors. A pedestrians' perception approach Pedestrian level of service evaluation in an area close to an under-construction metro line in Thessaloniki Comparing the microsimulated pedestrian level of service with the users' perception: the case of Thessaloniki Pedestrians' needs in the urban environment: the case of the city of Trikala How walking conditions affect pedestrian preferences: the case of Thessaloniki The impact of pedestrianization schemes on the environmental quality at central areas Environmental impact evaluation of major pedestrianization schemes through the use of modelling techniques Pedestrian level-of-service model for urban arterial facilities with sidewalks The pedestrian microsimulation applied to the river Neretva: the case study of the Mostar "old bridge Social force model for pedestrian dynamics Passenger and pedestrian modelling at transport facilities Sovershenstvovanie obrazovanija v oblasti gorodskogo i transportnogo planirovanija Verification and validation of Viswalk for building evacuation modelling The management of pedestrian emergency through dynamic assignment: some consideration about the "Refugee Hellenism" Square of Kalamaria (Greece) Pedestrian planning and design How eudaimonic aspect of subjective well-being affect transport mode choice? The case of Thessaloniki