J 2025

Accelerating Emergency Response in Airport Environments: An Experimental Study on Intelligent Sound Detection Systems

SMAŽINKA, Dalibor, Radomír ŠČUREK and Martin HRINKO

Basic information

Original name

Accelerating Emergency Response in Airport Environments: An Experimental Study on Intelligent Sound Detection Systems

Authors

SMAŽINKA, Dalibor (203 Czech Republic, guarantor), Radomír ŠČUREK (203 Czech Republic) and Martin HRINKO (203 Czech Republic, belonging to the institution)

Edition

International Journal of Safety and Security Engineering, 2025, 1258-5769

Other information

Language

English

Type of outcome

Article in a journal

Field of Study

20104 Transport engineering

Country of publisher

Canada

Confidentiality degree

is not subject to a state or trade secret

Organization unit

CEVRO University

Keywords in English

AI-driven security systems_airport security_crisis management-first responders reaction_multimodal detection_public safety and security_sound event detection systems
Changed: 19/6/2025 10:48, doc. Ing. Martin Hrinko, Ph.D., MBA, LL.M.

Abstract

V originále

Reducing emergency response times is critical to enhancing the efficiency of integrated rescue systems (IRS) and mitigating the impact of crisis events. This study investigates the deployment of intelligent sound event detection (SED) systems capable of recognizing specific sounds, such as gunshots and shouting, within public and commercial spaces. Through controlled simulations in an airport administrative building, the research demonstrates that SED systems significantly outperform traditional notification methods, reducing average response times by over 97%—from 175 seconds to just 5 seconds. These findings highlight the potential of SED systems to revolutionize emergency response strategies. The study introduces a novel approach by integrating sound detection with video surveillance into multimodal systems. This combination enhances situational awareness and allows for more precise responses to emergencies, addressing limitations of standalone detection systems. However, the study acknowledges key limitations—primarily that SED systems are less effective in silent incidents. The results emphasize the scalability of SED systems for diverse real-world applications in critical locations such as public institutions, shopping centers, and transportation hubs, where rapid decision-making is essential. Future research should explore optimizing these systems for noisy and unpredictable environments and advancing machine learning algorithms to improve reliability, adaptability, and detection accuracy, ensuring robust crisis management in varied scenarios.