Early Fire Detection: A New Indoor Laboratory Dataset and Data Distribution Analysis

Amril Nazir,Hamad Alhebsi,Maen Takruri,Husam Mosleh,Abdul-Halim Jallad

doi:10.3390/fire5010011

Abstract

Fire alarm systems are typically equipped with various sensors such as heat, smoke, and gas detectors. These provide fire alerts and notifications of emergency exits when a fire has been detected. However, such systems do not give early warning in order to allow appropriate action to be taken when an alarm is first triggered, as the fire may have already caused severe damage. This paper analyzes a new dataset gathered from controlled realistic fire experiments conducted in an indoor laboratory environment. The experiments were conducted in a controlled manner by triggering the source of fire using electrical devices and charcoal on paperboard, cardboard or clothing. Important data such as humidity, temperature, MQ139, Total Volatile Organic Compounds (TVOC) and eCO2 were collected using sensor devices. These datasets will be extremely valuable to researchers in the machine learning and data science communities interested in pursuing novel advanced statistical and machine learning techniques and methods for developing early fire detection systems. The analysis of the collected data demonstrates the possibility of using eCO2 and TVOC reading levels for early detection of smoldering fires. The experimental setup was based on Low-Power Wireless Area Networks (LPWAN), which can be used to reliably deliver fire-related data over long ranges without depending on the status of a cellular or WiFi Network.

Highlights

Conventional fire alarms are very effective in detecting flaming fires
While Sensor this is Networks a prototype setup, it has the potentia these reported designs depend on low range ZigBee protocol-based devices that do not develop into an IoT-like system capable of providing an early fire alarm system connec provide the long-range advantages of Low-Power Wireless Area Networks (LPWAN)
We conducted conducted on clothing triggered by electric fire sources, while the remaining four fire exadditional experiments for the electric fire source as the time for the fire alarms to be periments were triggered by charcoal on paperboard cardboard or clothing

Summary

Introduction

Conventional fire alarms are very effective in detecting flaming fires. they cannot detect the presence of smoldering fires, which develop very slowly and do not develop much heat when compared to flaming fires. It is usually initiated by heat sources such as cigarettes, coal or short-circuited wires, which can cause slow combustion of home furniture, linen, cloth or paper-based materials These fires are relatively slow and are not quickly detected by photoelectric smoke detectors and fire alarms except after the fire is fully developed and damage in the property is inevitable [5]. The pattern by the eC correlation analysis conducted on 2the provided here providesshown valuable resultsindicates and observations only the relevant features in early order to build accurate levels that it that can filter be used as amost possible trigger for fire alarms. While Sensor this is Networks a prototype setup, it has the potentia these reported designs depend on low range ZigBee protocol-based devices that do not develop into an IoT-like system capable of providing an early fire alarm system connec provide the long-range advantages of LPWAN.

Experimental Setup

Experiments:

System

Results and Discussion

Data Distribution Analysis

Conclusions