Crowd Evacuation in Stadiums Using Fire Alarm Prediction DOI Creative Commons

Afnan Alazbah,

Osama Rabie, Abdullah Al-Barakati

и другие.

Sensors, Год журнала: 2025, Номер 25(9), С. 2810 - 2810

Опубликована: Апрель 29, 2025

Ensuring rapid and efficient evacuation in high-density environments, such as stadiums, is critical for public safety during fire emergencies. Traditional alarm systems rely on reactive detection mechanisms, often resulting delayed response times, increased panic, overcrowding. This study introduces an AI-driven predictive model that leverages machine learning algorithms real-time environmental sensor data to anticipate hazards before ignition, improving emergency efficiency. To detect early risk indicators, the system processes from 62,630 measurements across 15 ecological parameters, including temperature, humidity, total volatile organic compounds (TVOC), CO2 levels, particulate matter. A comparative analysis of six models—Logistic Regression, Support Vector Machines (SVM), Random Forest, proposed EvacuNet—demonstrates EvacuNet outperforms all other models, achieving accuracy 99.99%, precision 1.00, recall AUC-ROC score close 1.00. The significantly reduces false rates enhances speed, allowing responders take preemptive action. Moreover, integrating optimization minimizes bottlenecks congestion, improves structured crowd movement. These findings underscore necessity intelligent high-occupancy venues, demonstrating AI-based modeling can drastically improve Future research should focus IoT-enabled navigation, reinforcement algorithms, management further enhance minimize casualties. By adopting advanced technologies, large-scale venues preparedness, reduce delays, safety.

Язык: Английский

Crowd Evacuation in Stadiums Using Fire Alarm Prediction DOI Creative Commons

Afnan Alazbah,

Osama Rabie, Abdullah Al-Barakati

и другие.

Sensors, Год журнала: 2025, Номер 25(9), С. 2810 - 2810

Опубликована: Апрель 29, 2025

Ensuring rapid and efficient evacuation in high-density environments, such as stadiums, is critical for public safety during fire emergencies. Traditional alarm systems rely on reactive detection mechanisms, often resulting delayed response times, increased panic, overcrowding. This study introduces an AI-driven predictive model that leverages machine learning algorithms real-time environmental sensor data to anticipate hazards before ignition, improving emergency efficiency. To detect early risk indicators, the system processes from 62,630 measurements across 15 ecological parameters, including temperature, humidity, total volatile organic compounds (TVOC), CO2 levels, particulate matter. A comparative analysis of six models—Logistic Regression, Support Vector Machines (SVM), Random Forest, proposed EvacuNet—demonstrates EvacuNet outperforms all other models, achieving accuracy 99.99%, precision 1.00, recall AUC-ROC score close 1.00. The significantly reduces false rates enhances speed, allowing responders take preemptive action. Moreover, integrating optimization minimizes bottlenecks congestion, improves structured crowd movement. These findings underscore necessity intelligent high-occupancy venues, demonstrating AI-based modeling can drastically improve Future research should focus IoT-enabled navigation, reinforcement algorithms, management further enhance minimize casualties. By adopting advanced technologies, large-scale venues preparedness, reduce delays, safety.

Язык: Английский

Процитировано

0