When ‘push’ does not come to ‘shove’: Revisiting ‘faster is slower’ in collective egress of human crowds
Graphical abstract
Introduction
Prediction and management of pedestrian flows through narrow doorways is an important aspect of crowd control and evacuation management. Narrow doorways are common geometric features of pedestrian and public facilities such as railway station gates, stadium gates, train gates etc (Davidich et al., 2013, Fernández et al., 2015, Seriani and Fernandez, 2015, Seriani and Fujiyama, 2018, Seriani et al., 2017) where crowds may congregate at high densities (Hänseler et al., 2017, Hughes, 2002, Shahhoseini and Sarvi, 2019, Shahhoseini et al., 2017). Therefore, accurate estimation of the egress and ingress flows through doorways and bottlenecks is a crucial aspect of crowd modelling (Boltes et al., 2013, Seriani et al., 2016) along with other modelling dimensions such as walking behaviour (Antonini et al., 2006, Flötteröd and Lämmel, 2015, Robin et al., 2009), exit direction choice (Ehtamo et al., 2010, Haghani and Sarvi, 2017, Haghani and Sarvi, 2018b, Haghani and Sarvi, 2019a, Huang and Guo, 2008, Lo et al., 2006) and pathfinding (Crociani et al., 2016, Daamen et al., 2005, Guo et al., 2012, Haghani and Sarvi, 2016, Hoogendoorn and Bovy, 2004, Kim et al., 2015). Determination of egress throughout rate at bottleneck is also one of the most important elements of evacuation time estimation (Ezaki et al., 2012, Shahhoseini and Sarvi, 2018, Yanagisawa et al., 2009a, Yanagisawa et al., 2009b, Yanagisawa and Nishinari, 2007).
The collective exit of human crowds through a narrow bottleneck is one of the most studied topics in pedestrian crowd dynamics (Daamen and Hoogendoorn, 2012a, Garcimartín et al., 2017, Garcimartín et al., 2014, Guo, 2014, Hoogendoorn and Daamen, 2005b, Nicolas et al., 2017, Oh and Park, 2017, Seyfried et al., 2010, Seyfried et al., 2009, Tanimoto et al., 2010, Tobias et al., 2006, Wang et al., 2015, Zuriguel et al., 2014). Despite the great wealth of research dedicated to this topic, however, a major aspect of pedestrian flow through a bottleneck has remained controversial, in that, there is a mixture of evidence about this question in the literature. The question is whether the efficiency of crowd escape through narrow doorways drop as a result of pedestrians’ rush to exit. This problem has been known as “faster is slower” assumption since the pioneer computational study of Helbing et al. (2000b) suggested based on computer simulations that, contrary to the intuition, when exiting through a narrow doorway, an elevated desire to move faster can cause further delays due to the clogging which reduces the efficiency of the collective discharge. The problem has been studied extensively since using both computational and experimental methods (Gago et al., 2013, Garcimartín et al., 2014, Parisi and Dorso, 2007, Pastor et al., 2015, Soria et al., 2012, Sticco et al., 2017, Suzuno et al., 2013), and has often been cited as a guide to practitioners and managers to discourage people from rushing to exits in cases of emergencies (Fang et al., 2011). It is often treated as a known fact by the media or managing authorities that the crowd will be better off (in minimising their discharge time) if people do not rush to escape in case of an emergency.
This work focuses on the question of ‘competitive egress’ (Hidalgo et al., 2017) through narrow doorways based on an empirical approach. The study is mainly motivated by the mixture of evidence that has emerged in recent years on this question and the fact that the role of ‘exit width’ on this phenomenon has not been systematically explored yet. This topic in general was identified through our recent review of the literature as one of the controversial topics in this field (Haghani and Sarvi, 2018a). The existing mixture of findings will be explained in more details in the following section. We believe that the great practical implications surrounding this topic warrants further examination and testing. The main question at hand is whether higher levels of mere rushing and pushing for exit through a narrow doorway (without any explicitly aggressive or destructive pushing) hinders the process of collective egress by prolonging it. Secondly, we aim to know whether the direction of this effect is dependent on the width of exit. Our hypothesis in designing this research was that the occurrence of the faster is slower may be limited to a certain range of exit width and that for wider exits, the effect might disappear. So, the question was how narrow the exit should be in order to observe the faster-is-slower effect. And our aim was to identify the possible borderline in terms of the exit width at which faster becomes faster as opposed to slower. The answer is crucial in determining whether crowds of evacuees should be deterred to rush at exit points in order to achieve a minimum egress time. Compared to the previous body of work on this topic that has predominantly used numerical techniques or experimental techniques with non-human crowds, we perform empirical laboratory tests with crowds of humans. And compared to the previously existing experiment of this question using human crowds (Garcimartín et al., 2017), for the first time, we systematically investigate the (potential) moderating effect of the exit width by examining varying widths ranging from 60 cm to 120 cm. Empirical tests reported in this work are performed with a crowd size of 114 individuals that is marginally larger than the counterpart comparable experiment that has previously been reported on this topic (Garcimartín et al., 2016, Garcimartín et al., 2017, Garcimartín et al., 2014, Pastor et al., 2015).
Section snippets
Literature review
The original observation of the faster-is-slower phenomenon was through computational analyses of Helbing et al. (2000a) that demonstrated the presence of a minimum in the relation between the desired velocity of simulated agents and the simulated egress time. The abovementioned finding derived from a model that treats the flow of pedestrians as self-moving particles has been re-examined experimentally by a number of earlier studies. Given the difficulties of experimenting this
Methods
The experiments simulated the collective egress of human crowds through narrow doorways and were conducted in 7th March 2017 in the basketball court of the University of Melbourne. A sample of university staff and students were registered through email invitations circulated in various faculties of whom 114 individuals turned up on the day of the experiments. All participants were undergraduate or postgraduate university students. They received monetary reward for their participation. The
Analyses
We observed that, as we instructed, subjects showed completely different patterns of behaviour under the three different behavioural conducted conditions. They moved much faster under conditions C2 and particularly C3 compared to C1. We observed obvious pushing and agitation under C3. Under C2, and particularly C3 larger degrees of crowd jam was created at the exit.
We quantified these differences in the speed and density using the colour-coding of the spatial-temporal averages of these
Conclusions
We empirically tested the assumption that higher degrees of (non-violent) competitiveness in collective egress of human crowds can hinder the process by prolonging it. The practical interpretation of the question that we tested is that, in terms of minimising the egress time, it would determine if the crowd could benefit from an orderly non-competitive conduct compared to a conduct than entails physical contact, greater density and (mild nonexplicit nonviolent) pushing. Also, we examined
Is there a simple answer to the faster-is-slower question?
We discussed in the literature review section of this article the details of the discrepancies that currently exist on the central question of this study: is faster in collective egress of humans slower or faster. We suggest that there may not be a simple yes-or-no answer to this question and there may be moderating factors in play whose roles need to be determined. In this study, we observed that when attempting to move faster (without engaging in explicit shoving), the collective egress was
Acknowledgements
This study was financially supported by Discovery Project research grant DP160103291 awarded by Australian Research Council.
The authors are very thankful to the editor-in-chief and three reviewers of this article for their feedback and insightful remarks.
References (89)
- et al.
Collecting pedestrian trajectories
Neurocomputing
(2013) Phased evacuation: an optimisation model which takes into account the capacity drop phenomenon in pedestrian flows
Fire Safety J.
(2009)- et al.
Level of service of pedestrian facilities: modelling human comfort perception in the evaluation of pedestrian behaviour patterns
Transp. Res. Part F: Traffic Psychol. Behav.
(2018) - et al.
High pressures in room evacuation processes and a first approach to the dynamics around unconscious pedestrians
Phys. A: Stat. Mech. Appl.
(2017) - et al.
Waiting zones for realistic modelling of pedestrian dynamics: a case study using two major German railway stations as examples
Transp. Res. Part C: Emerg. Technol.
(2013) - et al.
Resilience or panic? The public and terrorist attack
Lancet
(2002) - et al.
A proposed pedestrian waiting-time model for improving space–time use efficiency in stadium evacuation scenarios
Build. Environ.
(2011) - et al.
On passenger saturation flow in public transport doors
Transport. Res. Part A: Policy Pract.
(2015) - et al.
Bidirectional pedestrian fundamental diagram
Transport. Res. Part B: Methodol.
(2015) - et al.
Experimental evidence of the “Faster Is Slower” effect
Transport. Res. Proc.
(2014)
Simulation of spatial and temporal separation of pedestrian counter flow through a bottleneck
Phys. A: Stat. Mech. Appl.
Route choice in pedestrian evacuation under conditions of good and zero visibility: experimental and simulation results
Transport. Res. Part B: Methodol.
Questioning the linear relationship between doorway width and achievable flow rate
Fire Safety J.
How perception of peer behaviour influences escape decision making: the role of individual differences
J. Environ. Psychol.
Crowd behaviour and motion: empirical methods
Transport. Res. Part B: Methodol.
Hypothetical bias and decision-rule effect in modelling discrete directional choices
Transport. Res. Part A: Policy Pract.
Imitative (herd) behaviour in direction decision-making hinders efficiency of crowd evacuation processes
Safety Sci.
Simulating pedestrian flow through narrow exits
Phys. Lett. A
A dynamic network loading model for anisotropic and congested pedestrian flows
Transport. Res. Part B: Methodol.
Pedestrian route-choice and activity scheduling theory and models
Transp. Res. Part B: Methodol.
A continuum theory for the flow of pedestrians
Transport. Res. Part B: Methodol.
Microscopic events under high-density condition in uni-directional pedestrian flow experiment
Phys. A: Stat. Mech. Appl.
Does crowding affect the path choice of metro passengers?
Transport. Res. Part A: Policy Pract.
An experimental study of the “faster-is-slower” effect using mice under panic
Phys. A: Stat. Mech. Appl.
A game theory based exit selection model for evacuation
Fire Safety J.
Estimation of crowd density applying wavelet transform and machine learning
Phys. A: Stat. Mech. Appl.
Pedestrian flows through a narrow doorway: effect of individual behaviours on the global flow and microscopic dynamics
Transport. Res. Part B: Methodol.
Faster-is-slower effect in escaping ants revisited: ants do not behave like humans
Safety Sci.
Specification, estimation and validation of a pedestrian walking behavior model
Transport. Res. Part B: Methodol.
Pedestrian traffic management of boarding and alighting in metro stations
Transport. Res. Part C: Emerg. Technol.
Pedestrian crowd flows in shared spaces: investigating the impact of geometry based on micro and macro scale measures
Transport. Res. Part B: Methodol.
Pedestrian crowd dynamics in merging sections: revisiting the “faster-is-slower” phenomenon
Phys. A: Stat. Mech. Appl.
Experimental evidence of the “Faster is Slower” effect in the evacuation of ants
Safety Sci.
Clogging transition of pedestrian flow in T-shaped channel
Phys. A: Stat. Mech. Appl.
Study of bottleneck effect at an emergency evacuation exit using cellular automata model, mean field approximation analysis, and game theory
Phys. A: Stat. Mech. Appl.
Predicting collective behaviour at the Hajj: place, space and the process of cooperation
Phil. Trans. R. Soc. B
Discrete choice models of pedestrian walking behavior
Transp. Res., Part B, Methodol.
Role of driving force on the clogging of inert particles in a bottleneck
Phys. Rev. E
Efficient egress of escaping ants stressed with temperature
PLoS ONE
Analysis of crowd dynamics with laboratory experiments
Social identity and intergroup relationships in the management of crowds during mass emergencies and disasters: recommendations for emergency planners and responders
Polic.: J. Policy Pract.
A pedestrian movement model that takes into account the capacity drop phenomenon in the motion of crowd
Revisit the faster-is-slower effect for an exit at a corner
J. Stat. Mech.: Theory Exp.
Route choice in pedestrian simulation: design and evaluation of a model based on empirical observations
Intell. Artif.
Cited by (64)
A study on the arch mechanism of pedestrian evacuation and congestion alleviation strategies at building exits
2024, Journal of Building EngineeringHuman behaviour in fire: Knowledge foundation and temporal evolution
2024, Fire Safety JournalEffects of step time and neighbourhood rules on pedestrian evacuation using an extended cellular automata model considering aggressiveness
2024, Physica A: Statistical Mechanics and its ApplicationsImproved social force model considering the influence of COVID-19 pandemic: Pedestrian evacuation under regulation
2023, Applied Mathematical ModellingFaster and safer evacuations induced by closed vestibules
2023, Simulation Modelling Practice and Theory