Towards Fairness-aware Crowd Management System and Surge Prevention in Smart Cities

TL;DR

The paper tackles safety gaps in crowd management for smart cities by proposing fairness-aware evacuation and surge-prevention strategies. It introduces the Normalized Evacuation Time Disparity (NETD) metric and evaluates three gate-assignment strategies (RGA, VEGA, CGA) using high-fidelity simulations (Vadere) and a NetLogo-based prevention model. Findings show that VEGA improves fairness significantly in scenarios where vulnerable individuals cluster near exits, while CGA can outperform others in center or evenly dispersed layouts; a single strategy cannot fit all situations, motivating adaptive decision-making. Additionally, a preventive approach that stagesswitches performances to reduce crowding lowers both surge risk and the need for reactive evacuations, with Map C reducing stage-switching frequency by ~26% and APS by ~34%. The work demonstrates practical implications for fair, efficient, and proactive crowd management in CPS-enabled smart cities and highlights directions for real-world deployment, sensing integration, and privacy-aware design.

Abstract

Instances of casualties resulting from large crowds persist, highlighting the existing limitations of current crowd management practices in Smart Cities. One notable drawback is the insufficient provision for disadvantaged individuals who may require additional time to evacuate due to their slower running speed. Moreover, the existing escape strategies may fall short of ensuring the safety of all individuals during a crowd surge. To address these pressing concerns, this paper proposes two crowd management methodologies. Firstly, we advocate for implementing a fair evacuation strategy following a surge event, which considers the diverse needs of all individuals, ensuring inclusivity and mitigating potential risks. Secondly, we propose a preventative approach involving the adjustment of attraction locations and switching between stage performances in large-crowded events to minimize the occurrence of surges and enhance crowd dispersion. We used high-fidelity crowd management simulators to assess the effectiveness of our proposals. Our findings demonstrate the positive impact of the fair evacuation strategy on safety measures and inclusivity, which increases fairness by 41.8% on average. Furthermore, adjusting attraction locations and stage performances has shown a significant reduction in surges by 34% on average, enhancing overall crowd safety.

Figures (5)

• Figure 1: The four scenarios of crowd distribution in Vadere, where blue dots represent individuals, and orange blocks represent exit locations. (a): Center crowd gathering; (b): Non-center crowd gathering; (c): Evenly crowd dispersing; (d): Unevenly crowd dispersing
• Figure 2: Frequency of switching performance to another stage (F) and Average Panic/Surge value with four different Switch Index (SI) (10, 20, 30, 40) using different parameters. (a): PN500, BRF50, PT10, ST30; (b): PN750, BRF50, PT10, ST30; (c): PN500, BRF30, PT10, ST30; (d): PN500, BRF50, PT10, ST40; (e): PN500, BRF50, PT20, ST30.
• Figure 3: Our proposed simulator tool interface utilizing NetLogo. The top-left and bottom-right red rectangles on the simulation map denote two stages. Additionally, the upper yellow dot and lower blue dot symbolize the restroom and bar, respectively. The blue area indicates patches proximal to the left stage, while the green area depicts patches adjacent to the right stage.
• Figure 4: The illustration of the 2021 Astroworld Festival accident.
• Figure 5: Three different stage setups.