Expansion of Generative AI–Powered CCTV; LaonPeople’s AI Technology Gains Recognition

As Korea continues to face large-scale fire incidents—from aging commercial buildings in Euljiro and underground parking garage fires in Cheongna, to factory fires and massive wildfires burning millions of hectares—public attention is increasingly shifting toward domestic companies capable of detecting smoke and fire in the earliest stages using generative AI.

AI solutions company LaonPeople has launched OdinAi, a generative AI–powered system designed to detect fires and precursor signals at an early stage. Representing an evolution beyond traditional intelligent CCTV, OdinAi identifies early signs of smoke or fire even in high-risk environments, understands contextual relationships within video feeds, and provides rapid, accurate assessments along with recommended response actions.

Conventional intelligent CCTV systems—based mainly on deep-learning object detection—have struggled with irregular fire indicators such as smoke and flames. These systems have frequently misclassified reflections, lighting artifacts, or dust as fire, leading to false alarms, while occasionally missing real fire signals, resulting in false negatives.
In many documented cases, repeated false alarms caused system fatigue among operators, prompting them to disable alarms or detection features altogether—leading to catastrophic failures in actual fire emergencies when early suppression systems did not activate.

OdinAi Emerges as a Solution for High-Risk, High-Impact Environments

In environments where rapid detection and immediate response are critical—such as large apartment complexes, shopping malls, or wildfire-prone regions—OdinAi is being evaluated as a breakthrough solution.

Powered by LLM (Large Language Models) and VLM (Vision-Language Models), OdinAi fundamentally improves accuracy by analyzing not just the presence of flames but the context, cause, and relationships between surrounding objects.
For example, OdinAi can distinguish whether smoke near a ventilation duct is an actual fire precursor or simply exhaust, understanding environmental context to make autonomous judgments.

Early Fire Detection and Disaster Prevention: A New Standard

With its ability to perform rapid and precise analysis, OdinAi detects flames and smoke early to prevent escalation, setting a new standard for intelligent fire surveillance across both public and private sectors.

OdinAi Expands Across Public, Private, and Global Sectors

LaonPeople recently completed field deployments after extensive technical testing. At the Yongsan District Office parking facility, OdinAi is now in use for fire and smoke monitoring as well as continuous assessment of accidents and potential crimes. It also became the first AI system linked to CCTV in public EV-charging stations in Korea, detecting fires early and enabling real-time emergency response.

OdinAi is also operational at Incheon International Airport, where it analyzes relationships and intent between objects in video streams, detects undefined events, and autonomously interprets situations. It supports safety management by identifying intrusions into restricted areas, emergencies, or fire risks in shuttle-train zones—enhancing passenger safety and operational efficiency.

International deployment is accelerating as well.
Thailand—where typhoons and floods last year displaced 700,000 people and inflicted KRW 10 trillion in damage—has begun adopting OdinAi for multi-modal disaster detection, traffic monitoring, and crime prevention. The government has designated Bueng Kan Province as a pilot region for flood and crime detection, with plans to expand OdinAi across Thailand and to neighboring countries such as Laos and Vietnam.

Beyond Alarms: AI-Generated Reports and Automated Response Guidance

OdinAi goes far beyond simple alarm systems. Using its AI reporting engine, the platform:

• Generates natural-language descriptions of risk events
• Issues automated alerts
• Produces AI-generated incident reports
• Generates recommended response actions
• Supports real-time situational monitoring and post-incident analysis

With its natural-language interface, users can perform tasks such as CCTV control, alarm review, and Q&A via simple chat or voice prompts—allowing small facilities or unmanned/overnight monitoring centers to operate without specialized staff.

“Technology Creates Safety” — A New Standard for Industrial Safety

OdinAi is not just an alerting system but a next-generation video-analytics platform capable of genuine situational understanding and autonomous response.
LaonPeople has already proven the system’s capabilities across multiple industrial environments worldwide, demonstrating that generative AI can evolve beyond passive monitoring into active, context-aware decision-making.

According to LaonPeople, demand for generative AI–powered CCTV monitoring is rapidly growing among industries where fire poses serious risk—logistics warehouses, semiconductor fabs, and battery manufacturing lines.
The company added that OdinAi drew strong interest from attendees at Smart Factory & Automation World (SFAW 2025) and AI EXPO KOREA, reflecting growing market momentum.