Earticle

On Aug. 6, 2019, a large-scale fire broke out at a plant in Anseong, killing one firefighter, injuring 10 people and burning a warehouse. NISI reported by debris analysis that the AIBN(Azobisisobutyronitrile) was stored in the estimated ignition location. The AIBN is a “self-reactive substance”, which is sensitive to shock or friction and has a high risk of explosion, due to adverse reactions at air temperatures above 40℃. The risk of explosion with the reference material, BPO(Benzoyl peroxide) is compared, by measuring the explosive pressure, velocity and onset temperature according to the capacity of the AIBN, and proposed a policy for preventing fire by deriving the appraisal technique through qualitative analysis of GC/MS.

For the fire detection, the principles and functions of smart new detectors are studied. New methods that have not been done before are reviewed. The remote photographing system of mobile phone is applied, which collected related data to find out how to apply this function in the single alarm type detector. By using the single alarm type detector, the feasibility of introducing the imaging system is verified through basic experiments and re-experiments. After discussing with the relevant companies, the company decided to develop a new single alarm type detector with the imaging system attached to the company and release it on the market.

Increased use of hazardous materials increases overall number of safety accidents such as leaks, fires, and explosions. Also, it causes environmental damage such as water pollution and air pollution, exacerbating extensive and diverse damage. In this study, a wide range of surveys were conducted on the entire groups related to safety management of hazardous materials, including safety managers and carriers responsible for safety management at hazardous materials industrial sites, and safety management agents and tank performance test workers who are specialized of hazardous materials, and civil service personnel and inspectors at fire stations. Because the sample size was large for each group, a multi-faceted survey was analyzed using data analytics such as correlation analysis and regression. The results show the findings in the aspect of safety capabilities and institutional environment to be improved for the enhanced safety management of hazardous materials.

The technologies underlying fire and smoke detection systems play crucial roles in ensuring that these systems deliver optimal performance in modern. In fact, fire can cause significant damage to lives and properties. In majority of cities, camera-monitoring systems have been already installed, to take advantage of availability of these kinds of systems encourage us to develop a cost-effective vision detection methods. However, this is a complex vision task by the reason of perspective deformations, unusual camera angles and viewpoints, and seasonal changes. To overcome these limitations, we propose a method-based on deep learning that uses a convolutional neural network, which employs dilated convolutions. We evaluated our method, by training and testing it on our custom-built dataset. Consisting of a collection of fire and smoke images that we collected and labeled manually. The performances of methods proposed in previous studies were compared with those of well-known state-of-the-art architectures; our experimental results indicate that the classification performance and complexity of our method was superior to those of previous methods. In addition, our method is designed to be well generalized for unseen data, that it offers effective generalization and also reduces the number of false alarms.

In the ship fire, the initial fire extinguishing is very important because evacuation space is limited. Large ships are equipped with various extinguishing devices such as portable and fixed extinguishing equipment. The CO2 extinguishing equipment which lowers the oxygen density is one of the devices used in the ship fires. In actually, when a lot of fires occur in the engine room of a ship, a CO2 extinguishing system is often used to extinguish the fire. However, there is a space in the ship that can not be sealed in a large space such as a cargo storage compartment. The fire often occur in cargo compartments. In general, it is understood that CO2 extinguishing must be used in sealed spaces to be effective. Almost studies on the effectiveness of CO2 extinguishing systems for ship fires have been conducted only in a completely closed system such as an engine room. At present, studies on the effectiveness of CO2 extinguishing when there is no closed system has been insufficient. If the initial suppression of the fire is not successful, the combustion will increase, making it difficult to identify the exact cause of the fire. It is important to identify the cause of all fires, but especially for ship fires, it is very important to extinguish the fire early in order to identify the cause of fire. Therefore, in this paper, we analyzed the factors affecting CO2 extinguishing in an unsealed space based on the case of ship fire and investigated the effectiveness of CO2 extinguishing by approximating actual fire cases.

The purpose of fire investigation is to find the evidence related to the fire and analyze scientifically to find the cause of the fire. There is no systematic excavation methodology of fire investigation and the excavation technique are developed through the experimental study of actual fire cases and prepared to enhance the reliability and public confidence of the fire investigation. For the systematic survey, three step analytic methods are applied, which makes it possible to collect useful data and keep debris in the safe conditions for the future survey. Objective proofs are acquired by using the drawing, which can be compared with the real pictures. Also, the specific numbers are assigned to each proofs.

Statistics of missing persons information system by the National Police Agency indicated a total of 42,390 missing reports of children, intellectually disabled and dementia patients in 2019. Also, according to the date statistics by the National Fire Agency, 40,000 to 50,000 fires broke out annually in Korea, resulting in massive loss of lives and property. For this reason, research on various algorithms on disaster sites and searching methods for missing persons continues, especially drone technology with automatic search and analysis functions are being developed by using artificial intelligence(AI). This study aims at utilizing a drone which is one of many unmanned equipment to collect the most accurate and the largest data in time by minimizing the site damage. To collect the data for searching missing persons and identifying disaster sites, aviation video and photography, 3D mapping, and special equipment can be mounted on drones to take 360-degree panoramic photographs and images. For data acquisition, drone control patterns were studied for searching for missing persons and identifying disaster sites by applying data elements extracted from the researcher's preceding research, so that Deep Learning pattern recognition algorithm can be applied to the latest AI technology. This pattern is the most critical element for unmanned mobile device control technology including AI drones for missing person research and disaster sites. The images for 2D and 3D modeling of videos ad photographs, which were taken by drones by applying Deep Learning elements to drone search patterns, were analyzed based on the production of modeling results and data with PIX4D software. In many cases, the cause of the serious loss of life at the disaster and missing search site is the golden time delay in the search and the Human Error during the inspection/diagnosis of the site. Therefore, there is a need of pattern development of the unmanned drones for the reconnaissance patrol by using Deep Learning elements, in order to compensate for the lack of professional manpower for diagnosis/inspection as well as search equipment. For this reason, we suggest the joint search patterns of aviation drones and floating/underwater drones for the follow-up studies.