Crowd Dynamics Research Articles

Effect of age and stereopsis on a multiple-object tracking task.

3D vision develops during childhood and tends to diminish after 65 years of age. It is still relatively unknown how stereopsis is used in more complex/ecological contexts such as when walking about in crowds where objects are in motion and occlusions occur. One task that shares characteristics with the requirements for processing dynamic crowds is the multiple object-tracking task (MOT). In the present study we evaluated the impact of stereopsis on a MOT task as a function of age. A total of 60 observers consisting of three groups of 20 subjects (7–12 years old, 18–40 years old and 65 years and older) completed the task in both conditions (with and without stereoscopic effects). The adult group obtained the better scores, followed by the children and the older adult group. The performance difference between the stereoscopic and non-stereoscopic conditions was significant and similar for the adults and the children but was non significant for the older observers. These results show that stereopsis helps children and adults accomplish a MOT task, but has no impact on older adults’ performances. The present results have implications as to how populations differ in their efficiency of using stereoscopic cues for disambiguating complex dynamic scenes.

Modified two-layer social force model for emergency earthquake evacuation

Studies of crowd behavior with related research on computer simulation provide an effective basis for architectural design and effective crowd management. Based on low-density group organization patterns, a modified two-layer social force model is proposed in this paper to simulate and reproduce a group gathering process. First, this paper studies evacuation videos from the Luan’xian earthquake in 2012, and extends the study of group organization patterns to a higher density. Furthermore, taking full advantage of the strength in crowd gathering simulations, a new method on grouping and guidance is proposed while using crowd dynamics. Second, a real-life grouping situation in earthquake evacuation is simulated and reproduced. Comparing with the fundamental social force model and existing guided crowd model, the modified model reduces congestion time and truly reflects group behaviors. Furthermore, the experiment result also shows that a stable group pattern and a suitable leader could decrease collision and allow a safer evacuation process.

Modeling detour behavior of pedestrian dynamics under different conditions

Pedestrian simulation approach has been widely used to reveal the human behavior and evaluate the performance of crowd evacuation. In the existing pedestrian simulation models, the social force model is capable of predicting many collective phenomena. Detour behavior occurs in many cases, and the important behavior is a dominate factor of the crowd evacuation efficiency. However, limited attention has been attracted for analyzing and modeling the characteristics of detour behavior. In this paper, a modified social force model integrated by Voronoi diagram is proposed to calculate the detour direction and preferred velocity. Besides, with the consideration of locations and velocities of neighbor pedestrians, a Logit-based choice model is built to describe the detour direction choice. The proposed model is applied to analyze pedestrian dynamics in a corridor scenario with either unidirectional or bidirectional flow, and a building scenario in real-world. Simulation results show that the modified social force model including detour behavior could reduce the frequency of collision and deadlock, increase the average speed of the crowd, and predict more practical crowd dynamics with detour behavior. This model can also be potentially applied to understand the pedestrian dynamics and design emergent management strategies for crowd evacuations.

Modeling and simulation of group formation and crowd dynamics with smoke effect

In order to depict crowd dynamics in the case of smoke effect, a novel evacuation model is presented based on the extended cellular automaton. In the model, the moving probability of the target position in the next time step is given by combining distance gain and smoke gain at first. Through considering individual kinship and group effect, group disbanding and regrouping rules are defined, and the optimal evacuation route planning is determined. At last, the simulation was conducted to study the key factors influencing evacuation efficiency. The results showed that when the group size was small and the smoke spreading velocity was slow, priority selection of the shortest route strategy was good for reducing evacuation time, and large-scale group should be avoided as much as possible in evacuation process.

Dynamic crowd evacuation approach for the emergency route planning problem: Application to case studies

Disastrous situations, either natural or man-made, have claimed countless lives. In a planning context, this triggers the need for an evacuation situation that involves the emergency route planning (ERP) problem. ERP emphasizes the effectiveness of the evacuation plan to mitigate risk and ensure crowd safety. Reviews on previous ERP works have elicited a plethora of approaches, such as decision support, exact, heuristic, and meta-heuristic approaches. However, previous works on ERP focused on either crowd modeling techniques or routing approaches, but not both. This elicits the need to bridge modeling and routing approaches by means of a dynamic concept. To capture this dynamism concept, group cohesiveness and flexible routing approaches have been proposed, namely as an integrated evacuation planning with dynamism (iEvaP+). Group cohesiveness is utilized by considering group structure and intra-group compliance. Meanwhile, an artificial immune system (AIS) algorithm is adopted as the route optimizer. The iEvaP+ approach is applied to three case studies in order to identify the impact of crowd dynamism in the ERP problem. The result indicates that dynamism has direct influence on the evacuation plan performance. Besides the group cohesiveness, the properties of the evacuation network (such as exit) are also important in mitigating risks and crowd survivability.

Foot force models of crowd dynamics on a wobbly bridge.

Modern pedestrian and suspension bridges are designed using industry standard packages, yet disastrous resonant vibrations are observed, necessitating multimillion dollar repairs. Recent examples include pedestrian-induced vibrations during the opening of the Solférino Bridge in Paris in 1999 and the increased bouncing of the Squibb Park Bridge in Brooklyn in 2014. The most prominent example of an unstable lively bridge is the London Millennium Bridge, which started wobbling as a result of pedestrian-bridge interactions. Pedestrian phase locking due to footstep phase adjustment is suspected to be the main cause of its large lateral vibrations; however, its role in the initiation of wobbling was debated. We develop foot force models of pedestrians' response to bridge motion and detailed, yet analytically tractable, models of crowd phase locking. We use biomechanically inspired models of crowd lateral movement to investigate to what degree pedestrian synchrony must be present for a bridge to wobble significantly and what is a critical crowd size. Our results can be used as a safety guideline for designing pedestrian bridges or limiting the maximum occupancy of an existing bridge. The pedestrian models can be used as "crash test dummies" when numerically probing a specific bridge design. This is particularly important because the U.S. code for designing pedestrian bridges does not contain explicit guidelines that account for the collective pedestrian behavior.

Incorporating Intelligence into Exit Choice Model for Typical Evacuation

Integrating an exit choice model into a microscopic crowd dynamics model is an essential approach for obtaining more efficient evacuation model. We describe various aspects of decision-making capability of an existing rule-based exit choice model for evacuation processes. In simulations, however, the simulated evacuees clogging at exits have behaved non-intelligently, namely they do not give up their exits for better ones for safer egress. We refine the model to endow the individuals with the ability to leave their exits due to dynamic changes by modifying the model of their excitement resulted from the source of panic. This facilitates the approximately equal crowd size at exits for being until the end of the evacuation process, and thereby the model accomplishes more optimal evacuation. For further intelligence, we introduce the prediction factor that enables higher probability of equally distributing evacuees at exits. A simulation to validate the contribution is performed, and the results are analyzed and compared with the original model.

Patch-based topic model for group detection

Pedestrians in crowd scenes tend to connect with each other and form coherent groups. In order to investigate the collective behaviors in crowds, plenty of studies have been conducted on group detection. However, most of the existing methods are limited to discover the underlying semantic priors of individuals. By segmenting the crowd image into patches, this paper proposes the Patch-based Topic Model (PTM) for group detection. The main contributions of this study are threefold: (1) the crowd dynamics are represented by patch-level descriptor, which provides a macroscopic-level representation; (2) the semantic topic label of each patch are inferred by integrating the Latent Dirichlet Allocation (LDA) model and the Markov Random Fields (MRF); (3) the optimal group number is determined automatically with an intro-class distance evaluation criterion. Experimental results on real-world crowd videos demonstrate the superior performance of the proposed method over the state-of-the-arts.

A Feedback Control-Based Crowd Dynamics Management in IoT System

The development of technologies related to the Internet of Things (IoT) provides a new perspective on applications pertaining to smart cities. Smart city applications focus on resolving issues facing people in everyday life, and have attracted a considerable amount of research interest. The typical issue encountered in such places of daily use, such as stations, shopping malls, and stadiums is crowd dynamics management. Therefore, we focus on crowd dynamics management to resolve the problem of congestion using IoT technologies. Real-time crowd dynamics management can be achieved by gathering information relating to congestion and propose less crowded places. Although many crowd dynamics management applications have been proposed in various scenarios and many models have been devised to this end, a general model for evaluating the control effectiveness of crowd dynamics management has not yet been developed in IoT research. Therefore, in this paper, we propose a model to evaluate the performance of crowd dynamics management applications. In other words, the objective of this paper is to present the proof-of-concept of control effectiveness of crowd dynamics management. Our model uses feedback control theory, and enables an integrated evaluation of the control effectiveness of crowd dynamics management methods under various scenarios. We also provide extensive numerical results to verify the effectiveness of the model.

Pedestrian crowd dynamics in merging sections: Revisiting the “faster-is-slower” phenomenon

The study of the discharge of active or self-driven matter in narrow passages has become of the growing interest in a variety of fields. The question has particularly important practical applications for the safety of pedestrian human flows notably in emergency scenarios. It has been suggested predominantly through simulation in some theoretical studies as well as through few experimentations that under certain circumstances, an elevated vigour to escape may exacerbate the outflow and cause further delay although the experimental evidence is rather mixed. The dimensions of this complex phenomenon known as the “faster-is slower” effect are of crucial importance to be understood owing to its potential practical implications for the emergency management. The contextual requirements of observing this phenomenon are yet to be identified. It is not clear whether a “do not speed up” policy is universally beneficial and advisable in an evacuation scenario. Here for the first time we experimentally examine this phenomenon in relation to the pedestrian flows at merging sections as a common geometric feature of crowd egress. Various merging angles and three different speed regimes were examined in high-density laboratory experiments. The measurements of flow interruptions and egress efficiency all indicated that the pedestrians were discharged faster when moving at elevated speed levels. We also observed clear dependencies between the discharge rate and the physical layout of the merging with certain designs clearly outperforming others. But regardless of the design, we observed faster throughput and greater avalanche sizes when we instructed pedestrians to run. Our results give the suggestion that observation of the faster-is-slower effect may necessitate certain critical conditions including passages being overly narrow relative to the size of participles (pedestrians) to create long-lasting blockages. The faster-is-slower assumption may not be universal and there may be circumstances where faster is, in fact, faster for evacuees. In the light of these findings, we suggest that it is important to identify and formulate those conditions so they can be disentangled from one another in the models. Misguided overgeneralisations may have unintended adverse ramifications for the safe evacuation management, and this highlights the need for further exploration of this phenomenon.

ProactiveCrowd: Modelling Proactive Steering Behaviours for Agent‐Based Crowd Simulation

AbstractHow to realistically model an agent's steering behaviour is a critical issue in agent‐based crowd simulation. In this work, we investigate some proactive steering strategies for agents to minimize potential collisions. To this end, a behaviour‐based modelling framework is first introduced to model the process of how humans select and execute a proactive steering strategy in crowded situations and execute the corresponding behaviour accordingly. We then propose behaviour models for two inter‐related proactive steering behaviours, namely gap seeking and following. These behaviours can be frequently observed in real‐life scenarios, and they can easily affect overall crowd dynamics. We validate our work by evaluating the simulation results of our model with the real‐world data and comparing the performance of our model with that of two state‐of‐the‐art crowd models. The results show that the performance of our model is better or at least comparable to the compared models in terms of the realism at both individual and crowd levels.

Finite volume approximations of the Euler system with variable congestion

We are interested in the numerical simulations of the Euler system with variable congestion encoded by a singular pressure (Degond et al., 2016). This model describes for instance the macroscopic motion of a crowd with individual congestion preferences. We propose an asymptotic preserving (AP) scheme based on a conservative formulation of the system in terms of density, momentum and density fraction. A second order accuracy version of the scheme is also presented. We validate the scheme on one-dimensionnal test-cases and compare it with a scheme previously proposed in Degond et al. (2016) and extended here to higher order accuracy. We finally carry out two dimensional numerical simulations and show that the model exhibit typical crowd dynamics.

Mean-Field Pontryagin Maximum Principle

We derive a maximum principle for optimal control problems with constraints given by the coupling of a system of ordinary differential equations and a partial differential equation of Vlasov type with smooth interaction kernel. Such problems arise naturally as Gamma-limits of optimal control problems constrained by ordinary differential equations, modeling, for instance, external interventions on crowd dynamics by means of leaders. We obtain these first-order optimality conditions in the form of Hamiltonian flows in the Wasserstein space of probability measures with forward–backward boundary conditions with respect to the first and second marginals, respectively. In particular, we recover the equations and their solutions by means of a constructive procedure, which can be seen as the mean-field limit of the Pontryagin Maximum Principle applied to the optimal control problem for the discretized density, under a suitable scaling of the adjoint variables.

Almost “Nothing New Under the Sun”: American politics and the election of Donald Trump

Donald Trump’s surprising 2016 election as President of the United States was unusual both in the set of states he won and in clearly winning the electoral vote while decisively losing the popular vote. His victory is somewhat less surprising given recent Republican domination of American politics, a context which provides Trump both leadership opportunities and constraints. A large factor in Trump’s rise is the leader–follower dynamics of crowds, seen throughout time, which enabled him to win an uncritical and devoted following. An important part of that dynamic was Trump’s validation of the social identity of the white working class in the United States, especially in comparison to Hillary Clinton’s both implicit and explicit denigration of that base of Trump support. Trump’s identity story for his base is unusually exclusive, highlighted by ingroup vs. outgroup hostility. His appeal is compared to inclusive identity stories successfully related by other US presidents, which suggest how future leaders might effectively touch “the better angels of our nature.”

Overlap-avoiding Tickmodel: an Agent- and GIS-Based Method for Evacuation Simulations

Evacuation simulation is a method to determine evacuation times for areas, buildings, or vessels. It is based on the simulation of crowd dynamics and pedestrian motion; in this paper, we investigated the evacuation characteristics with a new motion model. The motion model and modeling space were implemented in an agent-based environment. The model is simple and generally applicable, it navigates the agent towards the destinations (safe zones) in a mixed macro-micro approach. The simulations were tested in a geospatially modeled lecture hall of the Budapest University of Technology and Economics (BME). The evacuation times and the panic rate were both estimated; a new way of measurement was applied for panic. Finally, conclusions were made on the person count ~ evacuation time and person count ~ panic rate relations. The paper introduces the key factors of this complex modeling phenomenon and demonstrates how to set up an agent-based evacuation model. The results can simulate the real phenomenon and constitute valuable assets for decision-making in public safety issues (architectural design, evacuation protocol, regulations of space).

A revolutionary crowd model: Implemented to contrast oscillating to consistent media influence on crowd behavior

Since Le Bon introduced his profound theory about the crowd in 1895, the phenomenon has been investigated across a range of scientific disciplines. Nevertheless, the mystery of the popular mind seems so far unrevealed, especially with the emergence of unfamiliar crowd movements, such as Twitter revolutions, which are triggered by novel types of media and interpersonal communication. Such original collective behavior, along with the current turbulent sociopolitical global environment, has necessitated the development of explanatory contemporary models. This research introduces a new revolutionary crowd model with a unique set of internal and external factors that can fit the modern uprisings in order to enable the understanding of the conditions that could lead to or prohibit the formation of revolutionary crowds. The model is implemented to examine the effectiveness of oscillating intensified mass media on the crowd pattern and dynamics. Among several emergent behaviors, the model shows an insignificant impact of disrupted intensified media on the crowd, in contrast with consistent low-intensity media, the failure of contagion theory in sustaining a revolution without a persistent stimulus, and the refutation of current claims of the insignificancy of leaders’ roles in igniting and maintaining modern crowds.

An experimental study of the impact of an obstacle on the escape efficiency by using mice under high competition

Crowd dynamics is an area of interest for scientists as a series of accidents caused by crowd within the last decades. Interestingly, some previous experiments and simulations have suggested that the presence of an obstacle in front of an outlet may improve the flow rate. A number of experiments were conducted to drive the mice to pass through an exit, with or without an obstacle before it. The impact of the obstacle on the efficiency of evacuation was studied under two typical geometry conditions, i.e., The setting I and the setting II. The setting I is an area of 1.5 m wide and 80 cm long and the setting II is a laterally confined space compared with Setting I. For evacuation at Setting I, a benchmarking study without the obstacle was conducted and the average evacuation time per mouse is 4.7s±0.44s. Further study was conducted by placing a 3.2 cm-obstacle at varying distances of 3 cm–6 cm and the quickest evacuation time per mouse is reduced to 3s±0.09s, around reduce by 36% compared with no obstacle condition. The presence of an obstacle before an exit improves the flow rate, which is consistent with the previous finding in silo flow and crowd flow. For evacuation at Setting II, similarly, a benchmarking study without obstacle was conducted and the average evacuation time per mouse is 3.9s±0.36s. When the 3.2 cm-obstacle was placed at varying distances from 4 cm to 10 cm, the average evacuation time per mouse increases to 4.9s±0.28s, around 26% higher than no obstacle condition. Further experiments were conducted at Setting II by using a 2 cm-diameter at varying distances from 4 cm to 8 cm and the presence of 2-cm obstacle still increases the evacuation time. The study reveals that the presence of an obstacle in front of an exit can improve or deteriorate the evacuation efficiency depending on the surrounded geometry. The time intervals of two consecutive mice and the burst sizes are also studied.

Automatic model construction for the behavior of human crowds

Designing suitable behavioral rules of agents so as to generate realistic behaviors is a fundamental and challenging task in many forms of computational modeling. This paper proposes a novel methodology to automatically generate a descriptive model, in the form of behavioral rules, from video data of human crowds. In the proposed methodology, the problem of modeling crowd behaviors is formulated as a symbolic regression problem and the self-learning gene expression programming is utilized to solve the problem and automatically obtain behavioral rules that match data. To evaluate its effectiveness, we apply the proposed method to generate a model from a video dataset in Switzerland and then test the generality of the model by validating against video data from the United States. The results demonstrate that, based on the observed movement of people in one scenario, the proposed methodology can automatically construct a general model capable of describing the crowd dynamics of another scenario in a different context (e.g., Switzerland vs. U.S.) as long as that the crowd behavior patterns are similar.

Enabling large-scale transit microsimulation for disruption response support using the Nexus platform

In many large cities, public transit systems have been carrying an ever-increasing burden of commuters. In such systems, service disruptions can negatively impact system performance and transit users well after they are resolved. Currently, transit agencies handle these disruption episodes in an ad-hoc fashion, largely due to the lack of adequate analytical tools to aid in analyzing and selecting appropriate response strategies. This paper presents a proof-of-concept case study of the Greater Toronto transit network using Nexus, a new crowd dynamics and transit network simulation platform. Nexus enables detailed simulation of all transit system actors using a novel method of linking together established simulators of surface transit, fully separated rail transit, and stations. Transit users, as agents in the model, move between the different simulators and have their routes determined by an external dynamic routing module. The case study focuses on interfacing Nexus with a commercial pedestrian simulator, MassMotion, to allow for detailed crowd simulation at key stations, and illustrating how the platform could be used for disruption management. To this end, the impact of disruptions of various lengths was analyzed, and a simple response strategy was implemented to provide an example of how the system could be used to test mitigating strategies.

Fluctuations around mean walking behaviors in diluted pedestrian flows.

Understanding and modeling the dynamics of pedestrian crowds can help with designing and increasing the safety of civil facilities. A key feature of a crowd is its intrinsic stochasticity, appearing even under very diluted conditions, due to the variability in individual behaviors. Individual stochasticity becomes even more important under densely crowded conditions, since it can be nonlinearly magnified and may lead to potentially dangerous collective behaviors. To understand quantitatively crowd stochasticity, we study the real-life dynamics of a large ensemble of pedestrians walking undisturbed, and we perform a statistical analysis of the fully resolved pedestrian trajectories obtained by a yearlong high-resolution measurement campaign. Our measurements have been carried out in a corridor of the Eindhoven University of Technology via a combination of Microsoft Kinect 3D range sensor and automatic head-tracking algorithms. The temporal homogeneity of our large database of trajectories allows us to robustly define and separate average walking behaviors from fluctuations parallel and orthogonal with respect to the average walking path. Fluctuations include rare events when individuals suddenly change their minds and invert their walking directions. Such tendency to invert direction has been poorly studied so far, even if it may have important implications on the functioning and safety of facilities. We propose a model for the dynamics of undisturbed pedestrians, based on stochastic differential equations, that provides a good agreement with our field observations, including the occurrence of rare events.