Single Pedestrian Research Articles

Automated analysis of pedestrian walking behaviour at a signalised intersection in China

An analysis of pedestrian walking behavior using gait parameters is presented; it uses automated video analysis to collect pedestrian data at a signalized intersection in Nanjing, China. Two aspects of microscopic pedestrian behavior are considered: First, the walking mechanism represented by pedestrian gait parameters. Second, the non-conforming crossing behavior of pedestrians. The effect of various pedestrian related attributes are investigated. The results show high accuracy in automatically detecting pedestrian violations, with an 85.2% correct detection rate. The walking speed and gait parameters for spatial violators are found to be significantly higher compared to non-violators. It is also found that pedestrians who enter the crosswalk during the late stage of the green pedestrian phase often adopt higher walking speed. The gait analysis shows that males tend to have a higher walking speed, walk ratio and longer step length than females. Single pedestrians are found to have higher speed and step frequency compared to pedestrians in groups. The presence of bikes on crosswalks significantly decreases the pedestrian walking speed, step length and frequency, leading to more gait variability. Such results are useful for many future applications such as calibration of simulation models and violation detection.

Utilization of Biodynamic Models of Human Body in Dynamic Analysis of Footbridges

The submitted paper deals with a dynamic analysis of the footbridge, which has been exposed to the effects of vandalism and crossing pedestrians. The loading produced by pedestrians was considered for two typical states. Firstly, the analysis was aimed at the ordinary traffic where pedestrians were discretized by the SDOF damped biodynamic models and secondly, the movement of single pedestrian, which was based on the inverted pendulum and biodynamic properties of human body, was evaluated. The jumping vandals were replaced by the motion of SDOF undamped biodynamic model. The calculations have been done at the computational model of the real footbridge across Opatovska Street in Prague.

Formulation of human–structure interaction system models for vertical vibration

In this paper, human–structure interaction system models for vibration in the vertical direction are considered. This work assembles various moving load models from the literature and proposes extension of the single pedestrian to a crowd of pedestrians for the FE formulation for crowd–structure interaction systems. The walking pedestrian vertical force is represented as a general time-dependent force, and the pedestrian is in turn modelled as moving force, moving mass, and moving spring–mass–damper. The arbitrary beam structure is modelled using either a formulation in modal coordinates or finite elements. In each case, the human–structure interaction (HSI) system is first formulated for a single walking pedestrian and then extended to consider a crowd of pedestrians. Finally, example applications for single pedestrian and crowd loading scenarios are examined. It is shown how the models can be used to quantify the interaction between the crowd and bridge structure. This work should find use for the evaluation of existing and new footbridges.

Panic evacuation of single pedestrians and couples

Understanding the timing requirements for evacuation of people has focused primarily on independent pedestrians rather than pedestrians emotionally connected. However, the main statistical effects observed in crowds, the so-called “faster is slower”, “clever is not always better” and the “low visibility enhancement”, cannot explain the overall behavior of a crowd during an evacuation process when correlated pedestrians due to, for example feelings, are present. Our research addresses this issue and examines the statistical behavior of a mixture of individuals and couples during a (panic) escaping process. We found that the attractive feeling among couples plays an important role in the time delays during the evacuation of a single exit room.

A framework for experimental determination of localised vertical pedestrian forces on full-scale structures using wireless attitude and heading reference systems

A major weakness among loading models for pedestrians walking on flexible structures proposed in recent years is the various uncorroborated assumptions made in their development. This applies to spatio-temporal characteristics of pedestrian loading and the nature of multi-object interactions. To alleviate this problem, a framework for the determination of localised pedestrian forces on full-scale structures is presented using a wireless attitude and heading reference systems (AHRS). An AHRS comprises a triad of tri-axial accelerometers, gyroscopes and magnetometers managed by a dedicated data processing unit, allowing motion in three-dimensional space to be reconstructed. A pedestrian loading model based on a single point inertial measurement from an AHRS is derived and shown to perform well against benchmark data collected on an instrumented treadmill. Unlike other models, the current model does not take any predefined form nor does it require any extrapolations as to the timing and amplitude of pedestrian loading. In order to assess correctly the influence of the moving pedestrian on behaviour of a structure, an algorithm for tracking the point of application of pedestrian force is developed based on data from a single AHRS attached to a foot. A set of controlled walking tests with a single pedestrian is conducted on a real footbridge for validation purposes. A remarkably good match between the measured and simulated bridge response is found, indeed confirming applicability of the proposed framework.

A novel crowd random loads model for pedestrians walking on footbridge

With modern footbridges becoming lighter and slenderer, the serviceability has gained more and more concern. For analyzing the vibration and evaluating the serviceability of footbridge, it is essential to give a proper model of loads induced by man, especially by crowd. At present, there are some models for crowd loads (e.g., Matsumoto’s model). In these models, a factor relating to the number of pedestrians is always introduced to multiply the amplitude calculated for single person. However, not only the amplitude but also the distribution, which is not considered in the present models, of the loads will effect on the dynamic response of structure. In addition, for crowd loads, it has strong stochastic nature including the distribution and the difference between pedestrians in the crowd. The walking parameters for single person of Chinese people are obtained by a walkway measure system, and the random characteristics of these parameters are analyzed. Then, a crowd random loads model is set up, in which the randomness of single pedestrian and the distribution of pedestrians on structure are involved. Finally, based on the model, the dynamic response of a footbridge under random pedestrians flow is analyzed by finite element method. According to the results, it is more near to real dynamic response situation of footbridge under crowd loads with the model and method introduced in this paper.

Correlation dimension of collective versus individual pedestrian movement patterns in crowd-quakes: A case-study

Video recording right before the Love Parade, Duisburg (Germany) crowd-quake, occurred on 24 July 2010, has been analysed in order to investigate the spatial properties of the crowd (collective case) and those of the single pedestrians in the crowd (individual case). The Grassberger–Procaccia correlation dimension, well known to be able to distinguish patterns in spatial point processes, was used. Our results for this case-study reveal that crowd and single pedestrians are characterized by different spatial behaviour: the whole crowd behaves as a quasi-homogeneous spatial point process through time, with an averaged correlation dimension of about 1.92; while the single pedestrians show a quite large variation of correlation dimensions indicating different spatial patterns, ranging from clustered to quasi-homogeneous.

Fire safety in Italian-style historical theatres: How photoluminescent wayfinding can improve occupants’ evacuation with no architecture modifications

Abstract Architectural Heritage is often prone to fire risk especially when many significant wooden structures with a particular historic and artistic value are present. This is the case of the Italian style historical theatres. Increasing fire safety of this architectural heritage generally clashes with preserving the original building features: massive and irreversible interventions are often needed so as to respect current severe regulations. Moreover, upgrading interventions can be insufficient so as to effectively improve occupants’ safety level, especially in overcrowded spaces and when people do not know much of the building itself. Occupants’ safety depends on their behaviours and their possibility to rapidly evacuate to a safe place. One of the most effective ways to help them to achieve this aim seems to be the adoption of a good emergency evacuation wayfinding system, especially in smoke or black-out conditions. This paper analyses the effectiveness of a reversible, easy-to-remove and low-impact system for evacuation guidance based on photoluminescent materials (PLM). The proposed continuous wayfinding system (CWS) is composed by PLM tiles along evacuation paths (both corridors and stairs). The application to a case study, the Italian style historical theatre “Gentile da Fabriano”, is then provided. Tests involve more than 100 individuals in smoke and black out conditions. CWS effectiveness was compared to a traditional punctual system in terms of motion speeds (for single pedestrians) and total evacuation time (for the whole building evacuation drill). Questionnaires filled in by involved pedestrians qualitatively evaluated the individuals’ acceptance of CWS. Tests with CWS show that individual's motion speed rises up to 50% and the total evacuation time is reduced down to 25% in respect to the traditional system. Comparisons with previous studies on PLM signs are provided. CWS can be easily introduced in this kind of historical theatres so as to increase the occupants’ safety level.

Group and Single Pedestrian Behavior in Crowd Dynamics

An analytical study was proposed in this paper to explore various behavioral aspects of group and single pedestrians. The results were obtained from reviewing the literature and studying primary data. Data collected from field surveys, direct observations, and video recordings were analyzed to produce the following findings. For crowds with densities between 0.4 and 0.8 persons per square meter, individuals walked faster than groups, on average. There was also significant evidence to suggest that individuals were likely to have more trajectory changes than groups. As the crowd density got higher, the likelihood of pedestrians splitting from groups of two decreased. With higher crowd densities, the rate at which split members came back to the group decreased. More individual pedestrians chose spacious stairs; groups of two were more likely to go for escalators.

Dynamic Analysis of the Structure Exposed to the Moving Periodic Force and Viscoelastic Models of the Human Body

Submitted paper is focused on dynamic analysis of the footbridge structure, which is represented by the simply supported beam. The continuous beam model was discretized by the FEM to MDOF system and loaded by pedestrians, walking across the structure. Three states of loading are considered in this paper. Firstly, it is the mathematical description of pedestrian's ordinary traffic, where each pedestrian is randomly generated by the Monte Carlo method. Secondly, the movement of single pedestrian, simplified by viscoelastic models, is introduced. Thirdly, the DLF model of stochastic vandals is investigated as well.

Parameters affecting the reduction factor in pedestrian load model based on pulsating stationary force

All dynamic pedestrian load models for serviceability control on pedestrian bridges which can be found in Eurocodes (EC) or guides to EC are based on dynamic load model of single pedestrian. Though the model of moving pulsating force is similar to the real character of pedestrian loading (walking along the bridge) the load model based on stationary pulsating force acting at the most adverse position of the bridge deck may be concurrent in bridge design for the reason of simplicity of acceleration assessment. Pulsating force moving along the bridge span do not produce the same dynamic response as a same pulsating stationary force. Therefore, the amplitude of stationary pulsating force have to be factorised by the reduction factor in a way that both load models (stationary and moving) cause the same maximum structural dynamic response. In this paper the structural parameters (structural system, span length, structural frequency and structural damping) affecting the reduction factor in pedestrian load model based on stationary pulsating force are researched. From conducted analyses it can be concluded changes in structural frequency does not affect the reduction factor, while changes in structural damping, structural system and span length affect the reduction factor, i.e. reduction factor is not a constant value as it is defined in load models proposed in some codes or guides.

Automated Pedestrian Safety Analysis at a Signalized Intersection in New York City: Automated Data Extraction for Safety Diagnosis and Behavioral Study

Automated computer vision techniques were used to analyze 2 h of video data collected at a major signalized intersection in New York City. The main objectives of this study were to diagnose pedestrian safety issues and identify contributing factors at the intersection and to demonstrate the feasibility of the automatic extraction of pedestrian data required for pedestrian behavior analysis—mainly pedestrian speed and gait parameters. The safety study was conducted with traffic conflict techniques. The main factor that contributed to the high number of pedestrian and vehicle conflicts was found to be pedestrian violations, mainly temporal violations in which pedestrians crossed the street during the “Don't Walk” or flashing “Don't Walk” phase. During the 2 h analyzed, about one-third of pedestrians were noncompliant with the signal timing or crosswalk boundary (17.9% spatial violations and 15.3% temporal violations). Pedestrian speed, step frequency, and step length were automatically extracted for 333 pedestrians and were found to follow the normal distribution with 95% confidence (mean and standard deviation of 1.47 ± 0.27 m/s, 1.96 ± 0.17 Hz, and 0.75 ± 0.14 m, respectively). Gait analysis showed that the walking speed for single pedestrians was 9% higher than for those who walked in groups. Males tended to be slightly faster than females, with higher step length but lower step frequency. Violators tended to have higher walking speeds compared with non-violators, and the difference in speed was dependent on step length but not on step frequency.

Modeling the Driver's Recognitive Characteristics to Crossing Pedestrians at Night for the Maximum Speed Limit and Lighting Index in China

Aiming to reduce the number of crashes caused by speeding at night on road section with a crosswalk, a study was conducted on the maximum speed limit and safe average luminance at night. In order to investigate the potential relationship between drivers’ recognitive characteristics and driving speed under different road lighting features, data of remaining driving time (period from the time that crossing pedestrian is recognized to the time that vehicle arrives at crosswalk with an uniform speed) were recorded. The results of the data analysis show that it is more difficult for divers to recognize crossing pedestrian at night when a single pedestrian is statistic and wears dark clothes. The remaining driving time decreases with the increase of driving speed and the decrease of road luminance. With the collected data, several multivariate nonlinear regression models were established to capture the relationship among the variables of remaining driving time at night, the driving speed, and the average luminance. Then the modeling results were used to develop the reasonable speed limit and safe average luminance by physical equations. The case studies are also introduced at the end of the paper.

Experiment and modeling of paired effect on evacuation from a three-dimensional space

Abstract A novel three-dimensional cellular automata evacuation model was proposed based on stairs factor for paired effect and variety velocities in pedestrian evacuation. In the model pedestrians' moving probability of target position at the next moment was defined based on distance profit and repulsive force profit, and evacuation strategy was elaborated in detail through analyzing variety velocities and repulsive phenomenon in moving process. At last, experiments with the simulation platform were conducted to study the relationships of evacuation time, average velocity and pedestrian velocity. The results showed that when the ratio of single pedestrian was higher in the system, the shortest route strategy was good for improving evacuation efficiency; in turn, if ratio of paired pedestrians was higher, it is good for improving evacuation efficiency to adopt strategy that avoided conflicts, and priority should be given to scattered evacuation.

Left-turn gap acceptance models considering pedestrian movement characteristics

Pedestrian–vehicle conflicts are considered as a common safety problem at signalized intersections. The threat to pedestrian safety is mainly related to the interaction with turning vehicles, especially left-turners (left-hand traffic system). This study aims to analyze the lag/gap acceptance behavior of left-turners considering pedestrian movement at signalized crosswalks. Furthermore, the severity of pedestrian–vehicle conflicts is addressed by analyzing vehicle speeds at the conflict points. User behavior at several signalized intersections in Japan is observed by using video cameras. It is assumed that pedestrian movements have their origin at either the near-side (the side of the exiting vehicular traffic) or far-side of the crosswalk. Accepted/rejected lags and gaps are extracted, classified depending on the direction of pedestrian movement, and modeled by using Cumulative Weibull distribution function. The results show that drivers tend to accept shorter lags/gaps between near-side pedestrians compared to far side pedestrians. Furthermore, drivers tend to accept short lags while being conservative about short gaps. Simultaneously vehicles clear the conflict area with significantly higher speeds when accepting lags with single pedestrians. This indicates that the conflicts that occur at low pedestrian demand levels are more severe compared to those at high demand levels.

Enhancement factors for the vertical response of footbridges subjected to stochastic crowd loading

The vertical acceleration response of a hypothetical footbridge is predicted for a sample of single pedestrians and a crowd of pedestrians using a probabilistic approach. This approach uses statistical distributions to account for the fact that pedestrian parameters are not identical for all pedestrians. Enhancement factors are proposed for predicting the response due to a crowd based on the predicted accelerations of a single pedestrian. The significant contribution of this work is the generation of response curves identifying enhancement factors for a range of crowd densities and synchronization levels.

Experimental study on an ultra high-rise building evacuation in China

With the development of modern cities, more and more ultra high-rise buildings have been built. Emergency evacuation of such buildings becomes one of the major concerns for building designers, building occupants and governments, especially after the disaster of 9/11. In this study we designed three experiments to investigate the process of ultra high-rise building evacuation. The experiments were performed in Shanghai World Financial Center, which is about 470m tall. In the first experiment, the evacuation of a single pedestrian from the top floor to the first floor was performed. Movement characteristics such as mean speed and the time needed for evacuation were analyzed. In the second experiment, the mass evacuation process was captured by video cameras and the data were extracted out manually. The evacuees were distributed initially on floors 12–17, and were asked to evacuate through a staircase onto the refuge floor. On the refuge floor, i.e. the 6th floor, the evacuees were required to transit into another staircase to keep moving down to the ground floor. The characteristic space–time curves for each evacuee were extracted and analyzed. Parameters such as the mean speed and the evacuation time, as well as the characteristic of the transit process from one staircase to another were investigated. We at last mimicked and discussed the process of lift evacuation for an ultra high-rise building in the 3rd experiment. Evacuees located on the 41st and 65th floor were asked to move downward by stairs until they came to the refuge floors, where they would evacuate to the ground floor by lifts. The time characteristic of the mixed evacuation strategy was discussed. The basic data obtained from the experiments are useful for building designers and can be used to validate and refine ultra high-rise building evacuation models.

Predicting Pedestrian Counts in Crowded Scenes With Rich and High-Dimensional Features

Estimating the number of pedestrians in surveillance images and videos has important applications in intelligent transportation systems. This problem is particularly challenging when the scenes are densely crowded, in which the techniques of tracking a single pedestrian has limited effectiveness. Alternative approaches employ statistical learning algorithms to infer pedestrian counts directly from visual features computed on images or scenes. In this paper, we describe a system for predicting pedestrian counts that significantly extends the utility of those ideas. Our approach incorporates a richer set of features for statistical modeling. While these features give rise to regression problems in a high-dimensional space, we leverage learning techniques to reduce dimensionality while still attaining high accuracy for predicting the number of pedestrians. Empirical results have validated our strategy. Specifically, our system outperforms state-of-the-art methods on standard benchmark tasks by a large margin.

Crossing Behaviors of Pedestrians at Signalized Intersections: Observational Study and Survey in China

In China pedestrian crashes make up a large proportion of road casualties and are more likely to occur at signalized intersections. This study examines the behavioral characteristics of pedestrians at such sites and the factors affecting pedestrian behavior. Data on pedestrian behavior were obtained from video images of pedestrian movements recorded at 26 signalized intersections in three Chinese cities, and 598 pedestrians who were using crosswalks were surveyed for their perceptions about the crossing environment. The behavior data indicated that most pedestrians walked normally when they crossed roads; the rate of compliance with traffic rules for all pedestrians was 62.8%. The differences in crossing behavior by gender, age, and the type of pedestrian group were then identified. The results showed that women and middle-aged individuals were more likely to violate traffic rules. In comparison with a single pedestrian and pedestrians in a pair, pedestrians in a group tended not to look at traffic signals. Analysis of the rates of compliance with traffic rules at crosswalks also indicated that crossing distance, signal timing, the presence of traffic police, and pedestrian volume had different effects. Moreover, the major reasons for crossing on red and opinions on waiting time were obtained by analysis of the survey data. For example, the largest proportion (30.25%) of surveyed individuals indicated that they violated traffic rules to save time and for convenience. Findings from this study can help researchers and practitioners understand pedestrian behavior at crosswalks at signalized intersections and thus create a better street-crossing environment for all pedestrians.

Experimental identification of pedestrian-induced lateral forces on footbridges

This paper presents a comprehensive experimental analysis of lateral forces generated by single pedestrians during continuous walking on a treadmill. Two different conditions are investigated; initially the treadmill is fixed and then it is laterally driven in a sinusoidal motion at varying combinations of frequencies (0.33–1.07 Hz) and amplitudes (4.5–48 mm). The experimental campaign involved 71 male and female human adults and covered approximately 55 km of walking distributed between 4954 individual tests. When walking on a laterally moving surface, motion-induced forces develop at the frequency of the movement and are herewith quantified through equivalent velocity and acceleration proportional coefficients. Their dependency on the vibration frequency and amplitude is presented, both in terms of mean values and probabilistically to illustrate the randomness associated with intra- and inter-subject variability. It is shown that the motion-induced portion of the pedestrian load (on average) inputs energy into the structure in the frequency range (normalised by the mean walking frequency) between approximately 0.6 and 1.2. Furthermore, it is shown that the load component in phase with the acceleration of the treadmill depends on the frequency of the movement, such that pedestrians (on average) subtract from the overall modal mass for low frequency motion and add to the overall modal mass at higher frequencies.