South Korea's firefighting technology and the capabilities of its firefighters are among the best in the world. Advanced fire trucks swiftly navigate to reported fire locations, confirming real-time information, while veteran firefighters engage in a coordinated battle, plunging into the flames. Paradoxically, however, large-scale fire disasters repeat annually, threatening our safety.

Why is this? The reason is clear. No matter how excellent the firefighting capabilities and equipment, South Korea's current fire system is trapped in an inherent and structural cycle of post-incident response, essentially "dispatching only after a fire has started."

It is time to boldly dismantle the anachronistic firefighting paradigm and enact a legal and institutional transformation towards "proactive and preventive disaster management" that detects and intercepts accidents and disasters as they invisibly unfold. The biggest contradiction in the fire incidents we face today stems from the "distortion of the golden time." Currently, the fire department's golden time is "dispatching within 5 minutes of receiving a fire report." The golden time for future firefighting must evolve to "taking action within 5 minutes of detecting abnormal signs."

Wired Fire Alarms Have Become the 'Boy Who Cried Wolf'

Recent large-scale fires are reported to fire stations only after smoke fills buildings and flames are visible to the naked eye. Even if fire trucks arrive at the scene swiftly, navigating roads and alleys, it's often the case that the fire has reached its "peak" – long after ignition, becoming intense and widespread.

In essence, due to the current fire system's structure, firefighters are always forced to start a disadvantageous battle and sometimes suffer casualties while fighting uncontrollable fires. Allowing this structure to persist is itself a major disaster.

High-rise buildings and apartments, areas densely packed with traditional markets and old buildings, and multi-use facilities and accommodations with high foot traffic all inevitably collapse in the face of these structural contradictions when a fire breaks out.

It is absolutely critical to secure the initial golden time in areas with a high risk of fire, which directly leads to major casualties and property damage when fires occur. However, in reality, we are relying on life-saving fire detection systems that are mechanical, reactive, and constrained by standard construction costs.

Wired detectors installed in many current buildings are prone to malfunction due to aging and have an excessively high rate of "non-fire alarms." This leads to firefighters being dispatched on false alarms, increasing their fatigue. Managers, tired of the incessant false alarms, often turn off the receivers entirely. This ultimately results in the critical consequence of completely losing the opportunity for initial suppression when a real fire occurs.

Even if an alarm sounds promptly, if the physical wired lines are severed by the rising heat and toxic gases during a fire, the signal between the receiver and the detectors is lost at the most crucial moment, rendering the entire system useless. This is precisely why the reactive suppression method, relying solely on the heroic dedication of firefighters and the speed of fire trucks, has reached its limit.

The key to breaking this structural contradiction is the full adoption and mandatory implementation of an "fire prevention system" that integrates AI (Artificial Intelligence) and wireless IIoT (Industrial Internet of Things) technology. We must transition to an AI detection system that analyzes real-time data on flame and smoke patterns, thermal imaging, and various environmental factors.

We need to establish a scientific safety net that captures early signs of fire, such as subtle abnormal electrical currents, changes in gas concentration, and localized temperature increases, before smoke or heat is visible to the naked eye. This AI system should warn of danger and provide early alerts before ignition.

Now, institutional innovation is needed to integrate this advanced technology into reality. Instead of waiting for revised bills that reflect the concept of disaster and fire prediction in the government's "AI Basic Act," "Disaster and Safety Act," and "Fire Facility Act," local governments must take the lead by revising "Fire Prevention Ordinances" to secure strong executive power. We propose three specific directions for ordinance revisions and alternatives that can be implemented at the local government level.

First, the responsibilities of local government heads must include the establishment of an AI-based fire prevention system.

Ordinances should be revised to stipulate the introduction of advanced disaster prevention technology as an institutional responsibility for safely protecting citizens' lives and property. Furthermore, the establishment of a fire prevention basic plan must clearly include "matters concerning the introduction and operation of AI-based fire detection and proactive fire prediction systems."

Second, the scope of mandatory installation for AI detectors and fire prediction systems must be defined, and a bold legal basis for financial support must be established accordingly.

AI detectors and fire prediction systems should be preferentially mandated and recommended for installation in public buildings commissioned by local governments, municipal and provincial multi-use facilities, fire prevention reinforcement zones within their jurisdiction, and areas with a high concentration of electric vehicle charging facilities, which have recently emerged as a new risk factor.

In parallel, ordinances should include practical mutual welfare measures to provide full or partial support from local government budgets for the installation of AI detectors and fire prediction systems in small traditional markets, densely populated old wooden buildings, small lodging facilities, and residential facilities for vulnerable groups such as single elderly individuals and people with disabilities.

Third, the establishment of a real-time data linkage integrated infrastructure must be mandated.

To maximize on-site response capabilities, the establishment of a real-time data linkage integrated infrastructure between the AI fire prediction system and the disaster control centers (119 Situation Rooms) operated by regional fire stations and provincial offices must be mandated. Only when the AI prediction algorithm detects signs of fire and this real-time risk information and abnormal sign alarm are immediately shared with the control system can the initial dispatch and suppression speed of the fire department be elevated to a legendary level.

Revising the fire prevention ordinances of all 226 basic local governments nationwide (si, gun, gu) and 17 metropolitan governments is the best solution to fundamentally prevent large-scale fires, minimize loss of life and property, and drastically reduce chronic false alarms, thus preventing wasted dispatches of firefighting resources and administrative inefficiencies. Furthermore, by leading proactive and advanced firefighting administration based on artificial intelligence, we can achieve not only a safe South Korea but also the export of cutting-edge firefighting equipment. The government itself declared a transformation into an AI-based industry a long time ago.

Disasters are cruel to the unprepared, but they lose their malevolence when faced by those who predict them through technology and defend against them through systems. Changing South Korea's firefighting system, which remains in a reactive state, to a proactive and preventive one is no longer a matter of choice. It is an urgent task of the times, involving the right to life of the people.

The technology is already at our fingertips. It is time for local governments and assemblies to respond with ordinance revisions, lead the era of scientific firefighting, and for the government to respond with legislative revisions.

◆ Youngjin Cho

CEO, RoseAI