Original Article
Abstract
References
Information
Purpose: This study is to analyze the current status and effects of artificial intelligence (AI) applications in the 119 emergency call-taking system and to empirically examine how an AI-based call-taking support system influences the efficiency of disaster response in real operational environments. To this end, this study reviews recent trends in key AI technologies—including speech-to-text (STT), natural language processing (NLP), machine learning, and large language models (LLMs)—and analyzes a practical application case of an AI-based 119 emergency call-taking system implemented at the Seoul Emergency Operations Center. Method: Consists of both quantitative and qualitative analyses. Quantitatively, operational data from an AI callbot pilot were analyzed to assess changes in call waiting time, the accuracy of emergency priority classification, and the effectiveness of non-emergency call triage. Qualitatively, organizational and operational changes in the emergency operations center following the introduction of AI were examined. Results: Indicate that the introduction of the AI-based system reduced the average call waiting time by approximately 7.33%. In addition, the automated pre-screening of non-emergency calls through the AI callbot significantly alleviated call-takers’ workloads. Even under conditions of heavy call surges, the AI-supported initial response improved the concentration of resources on high-priority emergency calls, while the emergency classification accuracy achieved a level exceeding 70% in the initial model implementation. Conclusion: This study demonstrates that AI in the 119 emergency call-taking system can function not merely as an automation tool but as a collaborative decision-support technology that assists human call-takers. Furthermore, the findings suggest that future advancements in AI-based emergency call-taking systems should focus on enhancing training data quality, adopting explainable AI, and establishing institutional and governance frameworks to ensure sustainable and reliable deployment in public safety services.
연구목적: 119 신고접수체계에 적용된 인공지능(AI) 기술의 현황과 효과를 분석하고, 실제 운영 환경에서 AI 기반 신고접수 지원 시스템이 재난 대응 효율성에 미치는 영향을 실증적으로 검토하는 데 있다. 이를 위해 음성인식(STT), 자연어처리(NLP), 기계학습, 대규모 언어모델(LLM) 등 주요 AI 기술의 적용 동향을 고찰하고, 서울종합방재센터를 중심으로 AI 기반 119 신고접수 시스템의 도입 사례를 분석하였다. 연구방법: AI 콜봇 실증 운영 데이터를 활용하여 신고접수 대기시간 변화, 긴급도 판단 정확도, 비긴급 신고 선별 효과를 정량적으로 분석하였으며, 동시에 AI 도입 이후 종합상황실의 업무 구조와 운영 방식 변화를 정성적으로 검토하였다. 분석결과: AI 시스템 도입 이후 평균 신고접수 대기시간이 약 7.33% 단축되었으며, AI 콜봇을 활용한 비긴급 신고 자동 선별을 통해 상담요원의 업무 부담이 완화되는 효과가 확인되었다. 특히 신고 폭주 상황에서도 AI 기반 초기 대응을 통해 긴급 신고 처리 집중도가 향상되었으며, 긴급도 판단 정확도 또한 초기 모델 기준 약 70% 이상의 수준을 보였다. 결론: 본 연구는 119 신고접수체계에서 AI가 단순한 자동화 수단을 넘어 접수요원의 판단을 보조하는 협업형 의사결정 지원 기술로 기능할 수 있음을 실증적으로 제시한다. 아울러 향후 AI 기반 신고접수체계의 안정적 확산을 위해 학습데이터 고도화, 설명가능한 AI 도입, 제도적·거버넌스 기반 마련이 필요함을 정책적·기술적 과제로 제안한다.
- Choi, W.-S. (2020). “A study on the intelligent disaster management system based on artificial intelligence.” Korean Society of Hazard Mitigation, Vol. 20, No. 1, pp. 127-140. 10.9798/KOSHAM.2020.20.1.127
- Choi, Y.-U., Lee, J.-W., Baek, J.-H., Lee, Y.-G., Park, J.-H., Kwon, Y.-W. (2023). “Implementation of an intelligent 119 callback system utilizing urgency keyword-based speech AI intent analysis.” Proceedings of KIIT Conference, Jeju, pp. 271-273.
- Electronics and Telecommunications Research Institute (2023). Trends in Development of Intelligent Response Technology for 112 and 119 Emergency Calls. Electronics and Telecommunications Trends, Daejeon. 10.22648/ETRI.2023.J.380307
- Hwang, J.-Y. (2024). “Application of AI in the 112 emergency operations center and expected benefits.” Korean Journal of Police Administration, Vol. 21, No. 2, pp. 179-198.
- Jeong, J.-H., Park, H.-H. (2024). “Development of a false alarm classification and prediction model using transformer-based large language model.” The Journal of Korean Institute of Communications and Information Sciences, Vol. 49, No. 4, pp. 489-502 10.7840/kics.2024.49.4.489
- Kim, B.C. (2021). Efficient Operation Strategies for 119 Emergency Operation Centers based on the Characteristics of Disaster Response. Master Thesis, Woosuk University.
- Kim, T.-H., Kim, J.-G., Woo, N.-K., Kim, J.-H. (2017). “A study on the improvement of work and training to make better Korean 119 dispatcher’s response ability.” Journal of the Society of Disaster Information, Vol. 13, No. 4, pp. 550-561.
- Kim, Y.-C., Han, Y.-T. (2024). “A comparison of training ad dispatch criteria for security personnel in 119 domestic and overseas general situation rooms.” Fire Science and Engineering, Vol. 38, No. 2, pp. 70-79. 10.7731/KIFSE.9384786e
- Kwon, S.-J., Kang, Y.-H., Lee, Y.-H., Lee, M.-H., Park, S.-H., Kang, M.-J. (2020). “Analysis of disaster safety situation classification algorithms based on natural language processing using 119 calls data.” Korea Information Processing Society, Vol. 9, No. 10, pp. 317-322.
- National Fire Agency (2025). National Fire Agency Statistical YearBook. Sejong.
- Seoul Fire & Disaster Headquarters (2025). Seoul Fire Administration White Paper. Seoul.
- Publisher :The Korean Society of Disaster Information
- Publisher(Ko) :한국재난정보학회
- Journal Title :Journal of the Society of Disaster Information
- Journal Title(Ko) :한국재난정보학회논문집
- Volume : 22
- No :1
- Pages :389-398
- DOI :https://doi.org/10.15683/kosdi.2026.3.31.389
Journal of the Society of Disaster Information
