Let's stop the fire misalarm with the AI sensor...Development of Intelligent Fire Detection Technology

[Anchor]
Due to frequent malfunctions, there are cases where the fire detector is turned off and leads to a large fire.

Korean researchers have developed an intelligent fire detection technology that can reduce false alarms that occur when there is no fire.

Reporter Lim Null-sol reports.

[Reporter]
A hotel in Bucheon last August

A red flame breaks through the window.

The alarm went off immediately after the fire, but the hotel staff decided it was a malfunction and turned off the fire alarm, increasing the damage.

Fire alarms often malfunction due to aging detection facilities, environmental factors, and sensor errors.

In fact, 96.6% of the 250,000 dispatches caused by fire alarms were caused by malfunctions between 2021 and July 2022.

Since nine out of ten malfunctions, fire alarms such as shepherd boys are often ignored, which can lead to major accidents.

Researchers in Korea have developed an "artificial intelligence sensor for preventing false alarm" that accurately distinguishes between when a fire breaks out and when a fire does not occur by measuring particles that vary depending on the wavelength of light.

This technology uses sensors to measure the characteristics of small particles in the air and determine whether there is a fire.

I'll check it myself.

When spraying in the air, the particles were measured in the detector using artificial intelligence learning data and recognized that there was no fire.

[Yang Hoe-sung / ETRI Lead researcher at the Defense Security Intelligence Laboratory: By projecting various wavelengths of light onto each particle, the technology can collect different scattering data for each particle and combine them with artificial intelligence, so it can distinguish whether it is a particle caused by fire or non-fire.]

In addition, it can detect the characteristics of smoke particles, so it can also determine whether it is a fire caused by wood.

Firefighters' on-site response is expected to be faster as they can predict not only the fire but also the initial fire factors.

[Lee Kang-bok / ETRI Director of Defense Safety Intelligence Research: I think it will be the first time to apply ICT technology to the fire safety field, and it will be applicable to facilities that deal with expensive logistics such as logistics facility warehouses]

The research team explained that the developed artificial intelligence sensor will be applied first to air intake sensors, which can accelerate localization away from import dependence.

In addition, it was expected to reduce the cost of firefighting and waste of firefighting power due to malfunctions.

I'm YTN Science Lim Neul-sol.


Video recording: Hwang Yu-min

Design: Eun