[Column] Disaster management by AIoT
Kim Yong-suk
The author is a professor of electronic engineering at SungKyunKwan University and vice president of the Institute of Semiconductor Engineers.
It’s been two months since the Itaewon disaster. It was a truly large-scale social disaster following the collapse of Seongsu Bridge (1994), the collapse of Sampoong Department Store (1995), the Daegu subway fire (2003) and the Sewol Ferry disaster (2014). Clarifying the cause and investigating the legal responsibility of those involved is important, but the focus should be put on the prevention of similar accidents. It is wiser to eliminate factors leading to accidents in advance rather than resolving them after a problem arises.
Regarding industrial accidents, there is a rule called the Heinrich Law of the 1:29:300 ratio. If a major industrial disaster occurs, there were 29 minor accidents and 300 potential incidents. It means that before a major disaster, there are always minor signs related to it. When these signs are spotted and corrected, disasters can be prevented. Can we use advanced science and technology to tell the possibility of a major disaster in advance?
An accident does not suddenly happen at an unforeseen moment, because it is preceded by multiple warning signs. By analyzing a series of smaller incidents that are precursors to a major disaster, a serious incident can be avoided. That can be achieved by using existing technologies.
For example, we can use the following method for a gathering of a large crowd. The location-based signal data from the three mobile carriers can measure the floating population, and closed-circuit television can be used to identify crowd density in a specific area. CCTV installed in the area can count how many people are gathering, and once a certain density is exceeded, a warning can be broadcast automatically.
After the nearby subway stations receive an alert, trains can pass without stopping. Text messages can be sent to personal cell phones. That can prevent more people from moving to the area. At the same time, police and firefighters can be deployed promptly to save more lives.
AIoT technology combines artificial intelligence and the Internet of Things, and can be used for disaster prevention. A series of smaller incidents that are precursors to a major disaster can be identified in advance and prevent a major disaster. Big data obtained through censors can be analyzed and understood with AI. AIoT technology can be utilized when a disaster such as a fire, flood, explosion, gas leak, environmental pollution or landslide occur.
Let’s look at the case of fire prevention. The Seoul Metropolitan Fire and Disaster Headquarters introduced IoT to firefighting facilities in 2017 and carried out a pilot program of checking the maintenance and management of the facilities at all times without on-site inspections by firefighting officials. I have experience of reviewing the technical side of the system with other experts at the time. Now it is in the stage of actual application.
The Act on the Installation and Management of Firefighting Facilities, which includes IoT fire prevention systems, is about to be implemented. The firefighting management system standard is being developed and different manufacturers will produce interchangeable facilities to ease operation and management. It is ready for commercialization.
Sinkholes can also be prevented by analyzing CCTV images to automatically detect sewage pipe defects. Accidents in smaller construction sites can be prevented too. CCTV footage from construction sites can be analyzed by AI in real time, detecting signs of danger and immediately notifying safety managers and workers of the danger.
In the wake of the Itaewon tragedy, it is necessary to establish a joint government-civilian organization such as a national disaster and accident prevention committee. Such entities should be formed to comprehensively promote policymaking, technology development, standardization, legislation and commercialization of products. Korea is an advanced ICT country. The strength in advanced science technology such as AI and IoT should be utilized for disaster prevention and response. By minimizing human intervention and maximizing science and technology, I hope this could serve as an opportunity to build an effective system to prevent disasters at personal and national levels.
Translation by the Korea JoongAng daily staff.