Samsung Fire & Marine has developed a system that uses artificial intelligence (AI) to analyze employee evacuation behavior and quantify training effectiveness.
The company announced on September 21 that it has created an 'AI-based Evacuation Training Quantitative Evaluation System' that analyzes evacuation training processes using CCTV footage and building spatial data.
Unlike traditional evacuation drills that focus on whether the training was conducted and site inspections, this new system is characterized by its ability to analyze actual behaviors that occur during the evacuation process. It can identify the time taken to start evacuating after a fire alarm, the evacuation route, and bottlenecks in specific areas. The evaluation focuses on five key areas: evacuation initiation, movement routes, bottlenecks, emergency exit usage, and response spread.
In July, Samsung Fire & Marine filed a patent for the related technology, 'Evacuation Evaluation Method and Device.' This technology quantitatively analyzes evacuation behavior by combining CCTV footage with spatial data.
In October, the company plans to conduct a proof of concept (PoC) for evacuation training at actual worksites. After confirming the system's analytical accuracy and applicability in the field, it aims to expand its usage.
A Samsung Fire & Marine representative stated, "We expect to identify specific vulnerabilities at each workplace and will work to enhance field applicability to promote a data-driven evacuation training system."
* This article has been translated by AI.
Copyright ⓒ Aju Press All rights reserved.