Forest Fire Monitoring System using WSNs Technology

Evizal Abdul Kadir

, Sri Listia Rosa

and Mahmod Othman

Department of Informatics Engineering, Faculty of Engineering, Universitas Islam Riau, Pekanbaru, Indonesia

Department of Fundamental and Applied Science Universiti Teknologi PETRONAS Seri Iskandar, Perak, 32610, Malaysia

Keywords:

WSNs, Forest Fire, Sensors, Detection.

Abstract:

Forest ﬁres contribute to air pollution, which is one of the disasters, and adversely affects the environment

because foggy particles along with carbon particles in a ﬁre. Forest ﬁres in the dry season occur in most of

Indonesia’s forestry areas. Riau Province is located on the island of Sumatra, Indonesia, in an area with a

high likelihood of forest ﬁres due to typical peatlands. The purpose of this research is to design and contribute

to new technologies for ﬁre detection using Wireless Sensor Networks (WSNs) Technology and intelligent

software for accurate ﬁre detection. This study proposes WSNs for the detection of forest ﬁres in peat areas

using sensor nodes with several embedded sensors for accurate ﬁre detection. The sensor node prototype was

designed and tested in a laboratory to check results and calibrate it to the real environment. Four sensors

are embedded with temperature and humidity sensors, ﬁre and smoke detection sensors and particle sensors.

It analyses with intelligent software to get accurate information and data from the ﬁre, including location,

detection of values from all sensors. The results show that WSNs sensor nodes can detect ﬁres and send

information about all parameters that indicate forest ﬁres. The design and development of WSN sensor nodes

is to assist local governments or institutions to overcome existing problems, particularly in Riau Province and

Indonesia, due to forest ﬁres.

1 INTRODUCTION

In Indonesia forest ﬁre is a disaster that incident

most of every year occur, especially in dry season.

According to the data, the total loss because of forest

ﬁre in year 1997 is USD2.45 billion (Yulianti et al.,

2012), but this loss of data still less than compare

to previous year in 1995, the total loss is USD19.1

billion. Riau Province in Sumatera is one of the areas

with the greatest risk of suffering from this disaster

due to peat and types of ﬂammable soil. According

to government agencies, the total loss in economic in

year 2016 for Riau province was due to forest ﬁres

of up to US $ 1,650 million. Apart from economic

losses, most activities stopped due to bad environment

(fog) and the closure of all schools, and there were no

activities in government ofﬁces and other institutions.

The forest ﬁre impact applies is not only to Indonesia

or the Riau province, but also to other countries, such

as Singapore and Malaysia, because Riau directly

limits these countries. The satellite uses current

procedures to obtain data on forest ﬁres to identify

critical points, then the information collected is sent

to the authorities and the team goes to a place to

take the steps needed to stop the ﬁre; Because peat

swamps can have their own ﬁres in the area, they must

socialize and campaign.

In this research focuses on development of

intelligent on the surface and level monitoring

systems for detection forest ﬁres, WSN smart

sensor nodes with new designs and smart systems

to collect accurate ﬁre data. The integration of

WSN sensor nodes and information exchange will

beneﬁt local communities and to local authorities

to access the information through sophisticated

real-time databases. He hopes this will be a fast

and cheap solution then obtaining ordinary satellite

data, and this will certainly beneﬁt to community and

economic enhancement. Furthermore, development

of a real-time monitoring system will involve the

backing of the government as the person responsible

for policy formation to apprehend how does the

system run and at the same time understand the

behavior of the results to take appropriate steps.

Kadir, E., Rosa, S. and Othman, M.

Forest Fire Monitoring System using WSNs Technology.

DOI: 10.5220/0009145201350139

In Proceedings of the Second International Conference on Science, Engineering and Technology (ICoSET 2019), pages 135-139

ISBN: 978-989-758-463-3

 2020 by SCITEPRESS – Science and Technology Publications, Lda. All r ights reserved

135

2 RELATED WORKS

WSNs applied for many uses, for example

applications in remote environments, automatic

industrial control, remote sensing and targets.

Applications that are similar to environmental

monitoring systems for forest ﬁre detection are

capable of real time monitoring and detection. In

most scenarios, WSNs consists of several small

number of nodes where the nodes are placed in far

location and unreachable hostile locations or in large

geographical areas. A number of WSNs nodes to

detect the changes in the environment and provide

information to the master cluster node or sensor base

station, then through the gate and for data transfer

to the server, which should be easily maintained and

scaled (Kadir et al., 2019; Kadir et al., 2018a).

A new method for action in the forest ﬁre

monitoring and detection as elaborate in (Liu et al.,

2018) is using data aggregation in WSNs. The

proposed method can be providing a faster and

more effective reaction to forest ﬁres by consuming

validated WSNs energy that is conﬁrmed in large

number of experiments in simulation. WSNs can

deliver better solutions for managing disaster and

operations rescue, such as alarm systems, ﬂood

detection, earthquake detection, forest ﬁre detection,

and landslide detection, water level sensors used to

measure various parameters. and discussed in (Pant

et al., 2017; Aranzazu-Suescun and Cardei, 2017).

Several research on WSNs as discuss in (Kadir

et al., 2018b), The WSN simulation addresses key

design issue, such as the monitored area related to

the sensor’s initial position, the number of sensor

required for a particular application and changes in

coverage over time. WSN uses an algorithm to

identify the injection of malicious data and provide

measures that are unaffected to various sensor and

even when they are hide in attack. The methodology

for applying this algorithm in this different contexts

and also evaluation of results in three different data

sets from different WSN distributions. (Illiano and

Lupu, 2015; Kadir et al., 2016).

Another research that already did in this

application of WSN in prediction of natural tragedies

such as hail, rainfall, ﬁre etc. WSN is rare and also

stochastic (Kansal et al., 2015). WSN implementation

in energy savings reduces delays in data transfer

and extends network life. The routing agent chain

(CCMAR) is used for the adaptive hierarchy of

energy saving clusters (LEACH) and energy saving

collections in sensor information systems (PEGASIS)

(Sasirekha and Swamynathan, 2017).

3 WSN IN FOREST FIRE

DETECTION ANALYSIS

Some of the ﬁctitious satellite forest ﬁres observed in

Riau province extend to most areas, especially in the

south. Figure 1 shows the number of critical points in

accordance with the distribution plan distributed in all

regions of Riau province.

Figure 1: Number of ﬁre hotspots in Riau Province based

on satellite image.

The access point coverage estimate that a series of

WSNs sensors are installed in a environmental area in

Riau province to monitoring this area. The function

of coverage is P given as:

P = f (x, y, t) = {(x

, y

), ...(x

, y

)},

, y

) = f (t), k = 1, 2, 3, ..., n

(1)

(x, y) is the sensor coordinates in the area of

monitored and t is the time. This model uses 2D

spatial projection from the ﬁre control area, 3D

sphere. In this issue, the networks do not move except

the WSN cellular sensor, but the position of the sensor

depends on time, because the sensor node must stop

working from time to time. There may be different

reasons for completing this process: hardware failure,

accident, battery consumption and accidental sensor

removal, etc.

Assume that you specify the scope of the IP matrix

as a value of scalar that represents of percentage in

coverage area observed in a certain time:

IP =

area covered with sensors

the total area of the surveillance region

100%

(2)

The basic component of model can be write in

WSNs as sensor node for deﬁned a vector:

S = (d, E(t)) (3)

the area covered can write as d by radio signals

that exchange data with neighboring nodes when the

ICoSET 2019 - The Second International Conference on Science, Engineering and Technology

136

sensor is in the transmission range or transmission

range. E (t) is the available energy to power the

sensor. Assume that there is a homogeneous of sensor

network to n integrated type sensors concentrators to

communicate with distribution nodes (Kadir et al.,

2019).

The parameters of network can be described in the

vector as:

M = (n, f

, ∆E) (4)

n can be deﬁning as the number of sensor; for

normal transmission frequencies and the consumption

of energy per transmission and transmission. Assume

that the sensor node period sends to the collected of

data to adjacent of nodes. Consumption of energy is

∆E include the spent energy in data collection and

process. In each node has 2 parts:

(a) feel the transmission and environmental data.

(b) receives data from the neighboring and

forwarding nodes.

The function of WSNs sensor center nodes is to

collects data from each of sensor nodes then send

it to the data coordination center or base station.

Data packets received and sent by the coordinator

node, which contains the measurement values and

address (humidity, temperature, and CO

) of the

original sensor node. WSN central nodes have

uninterruptible power supplies and communication

channels between the central node and the unlimited

coordination center. Therefore, the simulation regards

the sensor center as ”always available”. Main purpose

of this simulation and measurement is for optimizing

the networks path to send data from the sensors node

to the hub (Aksamovic et al., 2017).

4 DEVELOPMENT WSN NODE

FOR FOREST FIRE

DETECTION

Forest ﬁre is a natural or man-made events in several

cases throughout the global. Fire areas are found

major in climate, then the rainfall is high to provide

a important amount of vegetation, but in summer very

hot and in dry environments can create hazardous

fuel loads. Global of warming will assist to growth

the number of importance of this phenomena. Every

dry season a forest ﬁre is destroyed not only by

thousands of hectares of forest land, but also by public

assets, goods, resources and facilities. In addition,

ﬁreﬁghters and civilians face the risk of facing horriﬁc

victims every year. Figure 2 shows a diagram of a

series of WSNs sensor used in the forest area for the

detection purposes of ﬁre.

Figure 2: The sample of topology in the WSNs sensor nodes

deploy in a forest for disaster detection.

Forest ﬁres are a common and active phenomenon

that can change their nature and behavior from one

place to another and over time. The truth is that in

some places there is limited fuel for forests, so ﬁres

that continue to burn must spread to the nearest fuel.

The achieved by spreading to the complex heating to

neighboring in housing and community obtained from

the complex behavior of the ﬁre. Another case to

approach is based on the WSNs paradigm designed

and developed in a research project involving all key

players in the forest and ﬁreﬁghters for operations.

Figure 3: A WSNs sensor nodes propose use ZigBee

standard.

Another scenario in Figure 3 illustrates the

proposed schematic structure for multi sensors

node, controllers, routers, cluster heads, and remote

servers for the application WSNs based systems

Forest Fire Monitoring System using WSNs Technology

137

for protection management and forest ﬁre detection

and. Decision making This tree topology network

cluster structure proposes a project to reduce energy

loss and data packets during transmission. The

standard of ZigBee technique is a widely standard

based on IEEE 802.15.4, applicable to low-level PAN

(Wireless). ZigBee is one of the wireless network

standards for low-power sensors that is applied at

868/915 MHz and multi-frequency 2.4 GHz. The

technical advantage recommended by ZigBee is that

ZigBee offers a battery system that is durable, small,

and low battery. Cost, automatic or semi-automatic

installation, and high reliability. Therefore, in the

development and of WSNs node design is used by

multisensory systems to get the most appropriate

choice for the detection and monitoring of forest ﬁres

(Kadir, 2017).

The hardware used to detect and monitor ﬁres

at WSN nodes is available in many kinds on the

market. Where humidity, smoke, temperature, and

carbon sensors are positioned at the node to detect

all parameters that are strongly associated with forest

ﬁres. Figure 4 shows the actual formation of the

sensor in the environmental parameter calibration

test, before the sensor node is positioned in the

ﬁeld, the sensor node must be conﬁgured according

to design and requirements. All nodes send data

or messages to the coordinator WSNs, which has

the function of receiving all information from the

scattered nodes.

Figure 4: A Prototype of WSNs sensor nodes with multiple

sensors use Arduino processor.

5 CONCLUSIONS

It has been proposed to develop a WSN node to detect

ﬁres and monitor from various sensors for correct

detection. Projects that include mathematical analysis

and regional approaches must cover the entire Riau

province. Sensors of humidity, temperature, smoke,

and carbon are the focus of attention in this issue

of these parameters are the main parameter for

ﬁre conditions both on land and in the forest.

Recommended sensor nodes using the low-power

ZigBee model, sensor nodes can be used as long

battery-powered nodes. At least in each region,

a network coordinator must be formed to cover

the entire Riau province and the gateway must be

available to hospital the server (cloud database) and

monitor the computer. The highly applicable WSN

concept proposed to detect forest of ﬁres in province

of Riau is very useful for preparing presentations.

ACKNOWLEDGEMENTS

Authors would like to say thank you very much

to KEMENRISTEKDIKTI Indonesia and Universiti

Teknologi Petronas, Malaysia for funding this

research as well as Universitas Islam Riau to support

the facilities.

REFERENCES

Aksamovic, A., Hebibovic, M., and Boskovic, D. (2017).

Forest ﬁre early detection system design utilising

the wsn simulator. In 2017 XXVI International

Conference on Information, Communication and

Automation Technologies (ICAT), pages 1–5. IEEE.

Aranzazu-Suescun, C. and Cardei, M. (2017). Distributed

algorithms for event reporting in mobile-sink wsns for

internet of things. Tsinghua Science and Technology,

22(4):413–426.

Illiano, V. P. and Lupu, E. C. (2015). Detecting malicious

data injections in event detection wireless sensor

networks. IEEE Transactions on Network and service

management, 12(3):496–510.

Kadir, E. A. (2017). A reconﬁgurable mimo antenna

system for wireless communications. In 2017 4th

International Conference on Electrical Engineering,

Computer Science and Informatics (EECSI), pages

1–4. IEEE.

Kadir, E. A., Irie, H., Rahim, S. K. A., Arta, Y.,

and Rosa, S. L. (2018a). Reconﬁgurable mimo

antenna for wireless communication based on arduino

microcontroller. In 2018 IEEE International RF and

Microwave Conference (RFM), pages 119–122. IEEE.

ICoSET 2019 - The Second International Conference on Science, Engineering and Technology

138

Kadir, E. A., Irie, H., and Rosa, S. L. (2019). Modeling

of wireless sensor networks for detection land and

forest ﬁre hotspot. In 2019 International Conference

on Electronics, Information, and Communication

(ICEIC), pages 1–5. IEEE.

Kadir, E. A., Rosa, S. L., and Gunawan, H. (2016).

Application of rﬁd technology and e-seal in container

terminal process. In 2016 4th International

Conference on Information and Communication

Technology (ICoICT), pages 1–6. IEEE.

Kadir, E. A., Rosa, S. L., and Yulianti, A. (2018b).

Application of wsns for detection land and forest ﬁre

in riau province indonesia. In 2018 International

Conference on Electrical Engineering and Computer

Science (ICECOS), pages 25–28. IEEE.

Kansal, A., Singh, Y., Kumar, N., and Mohindru, V. (2015).

Detection of forest ﬁres using machine learning

technique: A perspective. In 2015 Third International

Conference on Image Information Processing (ICIIP),

pages 241–245. IEEE.

Liu, Y., Liu, Y., Xu, H., and Teo, K. L. (2018). Forest ﬁre

monitoring, detection and decision making systems

by wireless sensor network. In 2018 Chinese Control

And Decision Conference (CCDC), pages 5482–5486.

IEEE.

Pant, D., Verma, S., and Dhuliya, P. (2017). A study

on disaster detection and management using wsn

in himalayan region of uttarakhand. In 2017 3rd

International conference on advances in computing,

communication & automation (ICACCA)(Fall), pages

1–6. IEEE.

Sasirekha, S. and Swamynathan, S. (2017). Cluster-chain

mobile agent routing algorithm for efﬁcient data

aggregation in wireless sensor network. Journal of

Communications and Networks, 19(4):392–401.

Yulianti, N., Hayasaka, H., and Usup, A. (2012). Recent

forest and peat ﬁre trends in indonesia the latest

decade by modis hotspot data. Global environmental

research, 16(1):105–116.

Forest Fire Monitoring System using WSNs Technology

139