On October 29, at around 10.15 pm, a pre-Halloween crowd surged through a popular narrow street in Seoul, South Korea. The street, lined with bars and restaurants, sees a large footfall on most days but the crowding witnessed on that was unprecedented. The resulting crowd crush claimed over 150 lives and left many more injured. Close on the heels of the Seoul tragedy, a pedestrian bridge in Gujarat’s Morbi collapsed, killing 135 people.
In the past, crowd-related disasters have occurred at concerts, pubs and stadiums around the world. In 2017, a packed foot over-bridge at Mumbai’s Prabhadevi railway station (formerly known as Elphinstone) witnessed a stampede that killed 22 people
Crowd densities of under four people per square metre is usually considered a safe limit. As it approaches six people per square metre, the risks increase significantly. It is common to observe densities that breach this threshold in South Asian cities – especially in areas that see high footfall such as popular markets, public places and transit stations.
World Resources Institute, India, for which we work, is testing an algorithm that captures crowding densities by counting the number of people using images from video feeds. This algorithm was recently applied to a video of a crowded suburban Mumbai station where the density was close to 10 people per square metre. The same algorithm showed that the crowd in Seoul reached densities of up to 11 people per square metre.
Apart from crowd density, peripheral factors also play a role in making a tight situation perilous. In Seoul, crowd waves (a movement in overcrowded situations where people are lifted by a human wave), crowd collapse (people falling because of lack of balance due to overcrowding) and the lack of adequate crowd management led to the incident. In the case of Mumbai, misinformation and confusion sowed panic among commuters.
Cities can look at preventing such disasters by adopting a few critical measures:
An understanding of metrics such as densities, temporal and spatial patterns of crowds can help frame more robust mitigation strategies. A database of benchmarks on average crowding levels, safe thresholds, scenario building, and predicting reaction to trigger events can prevent catastrophic disasters.
CCTV footage can be used to address crowding challenges. Using the algorithm on the footage will lead to a better understanding of how crowd density changes through the day and when it is likely to breach safe levels. It also helps to identify vulnerabilities such as bottlenecks, design flaws and the effect of external conditions like the weather or delays in transit services that can potentially lead to crowding.
Mitigation strategies for crowd management need to be developed for city authorities, the community and individuals. Density as an indicator authorities a vocabulary and informs better crowd management and disaster preparedness. These indicators can be used to develop early warning systems and institute preemptive crowd control protocols in the event of a possible emergency.
While people do tend to adapt to crowds – like natural lanes forming in a crowded corridor – citizens must be trained to act appropriately in such situations. These include measures such as identifying the closest exit points, staying away from walls, bracing oneself to prevent injuries to the chest and moving diagonally between pockets of people. Information about these mechanisms must be effectively communicated to the citizens along with ensuring crowd regulation.
Mitigation efforts also need to be complemented by including infrastructure interventions. In a city like Mumbai, where over 50% of the people walk, pedestrian infrastructure is markedly inadequate. The images below show how the lack of safe pedestrian infrastructure compels people to walk on the roads, bringing them into direct conflict with vehicles.
Pedestrian counts can help design infrastructure to accommodate different volumes of crowds. It can also identify conflict points where pedestrians are at risk. At transit stations, a deeper understanding of pressure points, bottlenecks and clearing times can support quick interventions like better signages and the installation of mirrors at strategic points, to increase the sense of space for seamless movement of commuter traffic.
Crowd planning, monitoring and control should be prioritised to ensure public safety in cities. Quantifying crowding, using indicators like densities, is the first step for effective crowd management. Indian cities are among the top surveilled cities in the world. The infrastructure already in place can be expanded to predict, control and mitigate such disasters and ensure the safety of citizens.
Madhav Pai is Executive Director, WRI India Ross Centre and Vishal Ramprasad is Senior Program Manager, Sustainable Cities and Transport at the World Resources Institute India (WRI India). Views expressed by the authors are personal.