Controlling the crowds
Commuters travelling by city rail and metro systems often face the challenge of packed platforms, long queues and significant congestion during peak times. As well as causing frustrating delays, this can also pose risks to safety through overcrowding and passengers pushing and rushing to doors.
A new study, presented at Ergonomics & Human Factors 2026, has looked at how dealing with these situations could be improved and whether AI can help crowd management.
Researchers from Newcastle University collected in-depth insights from stakeholders involved in urban transport systems. They used work domain analysis (WDA) to map the metro station environment and control task analysis (ConTA) to examine the decisions made when congestion was as its peak. Through this, they developed a systems-based approach to dealing with congestion and also created a roadmap for adopting AI tools to help predict and manage crowds.
The study said: “Passenger congestion in metro systems remains a persistent operational and safety challenge shaped by the interaction of infrastructure design, operational practices, passenger behaviour and organisational capacity. This study demonstrates that congestion is not solely a physical or operational issue, but a systemic challenge embedded within socio-technical structures.”
