IOP Conference Series: Materials Science and Engineering
PAPER • OPEN ACCESS
IoT based Forest Fire Detection System in Cloud Paradigm
To cite this article: H Singh et al 2021 IOP Conf. Ser.: Mater. Sci. Eng. 1022 012068
View the article online for updates and enhancements.
This content was downloaded from IP address 131.181.108.90 on 14/04/2021 at 11:27
ICCRDA 2020 IOP Publishing
IOP Conf. Series: Materials Science and Engineering 1022 (2021) 012068 doi:10.1088/1757-899X/1022/1/012068
IoT based Forest Fire Detection System in Cloud Paradigm
H Singh1, A Shukla2, S Kumar3
1
ABES Engineering College, Ghaziabad, UP, India
2,3
Jaypee University of Engineering and Technology, Guna, MP, India
Email: harikeshsingh@yahoo.co.in
Abstract. The most common hazard in forest fires are accident as the forest themselves
destroys the forests and can be great threat to wild life and peoples. The Internet of Things
(IoT) is the physical device which is used to connect, store and enable the objects to collect
information for exchanging the data through the internet based system. The main objective of
paper is to predict the developing trend of the fire by monitoring temperature, humidity, etc.
There are several types of sensors which can be used for the collections of temperature and
humidity data as an input for IoT based system which increases the efficiency of process of fire
detection. The proposed mechanism can be used a threshold value of each sensor to validate
the data and take initiative to avoid the prerequisite of achieving high reliability prevention
using this mechanism.
1. Introduction
The forests are major a part of environment on the world which also balancing the ecological system.
The fire is typically observed after covering the outsized area and then its controlling is almost
impossible sometimes. As a result other consequences of forest fires are long-term disastrous effects
which impacts directly on local weather patterns, heating and extinction of rare species of the flora
and fauna. The complete forests are filled with highly combustible things like dry leaves and branches
of the several trees which are act as a fuel source for fire ignition and later fire stages.
As a part of the event of the fire detection system, it is mandatory to analyse the characteristics of
the environment under normal and fire conditions. It is also required to develop a strong proposal,
information on temperature and ratio under normal conditions was collected during the summer in
Colombia and following observations are taken as [1][2]:
The temperature and humidity values are reflecting the cyclical behavior at different stages of
the day under normal conditions.
It also maintains an inverse relationship between their magnitudes with temperature and
humidity values under normal conditions. So, as the temperature increases, it decreases the
humidity.
The speed of variation is higher and as sensor node is activated under the sun-light, the rate of
change in temperature observed at a fire place.
Content from this work may be used under the terms of the Creative Commons Attribution 3.0 licence. Any further distribution
of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.
Published under licence by IOP Publishing Ltd 1
ICCRDA 2020 IOP Publishing
IOP Conf. Series: Materials Science and Engineering 1022 (2021) 012068 doi:10.1088/1757-899X/1022/1/012068
Most forest fires are occurred due to atmospheric temperature and occasional humidity which
provide auspicious environments for fire to begin. These forests and the bio-diversity of India are
increasing a greater chance of fire underneath tremendous pressure. It is significant to detect
woodland fire by using early caution schemes. In India, there are normally conventional behaviors to
reduce the impact of wooded area fire [3].
There are several methods to detect the forest fires including the watch towers monitoring using
satellite images but it is still not sufficient because of complex forest infrastructure which requires lot
of trained persons to operate. In order to protect these huge stretches of land and biodiversity, there is
an important want to have a tremendous surveillance and early caution structures for the forest fire
whilst in its establishing diploma of spreading shall be managed which will save you its unfold.
Usually so that you can prevent fire in a forest vicinity there may additionally be a massive
dependency on man power [4].
An efficient mechanism to avert wooded area fire may also be laid down in a greater way for it
should be beneficial to officials and woodland involved authority. Through this approach, we create a
machine if you desire to assist the woodland fire protection people to manipulate undesirable results
on forest ecosystems and the natural world. For this purpose, sensing surroundings can be deployed
with a large vast range of wireless sensor nodes [5].
A wooded fire area is a natural disaster consisting of a fire which destroys a forested location, and
can be a splendid risk to those who live in forests as well as the natural world. Forest fires are usually
started out by using lightning, however additionally by way of human negligence or arson, and might
burn hundreds of square kilometers [6].
2. Related Work
Various researchers presented their solutions to deal with the problem. Some major contributions
have been presented as follows:
x Fire forecasting is one of the critical research challenges of the Wireless Sensor Networks (WSN)
network to save much of our environment. In the WSN fire-detection system, sensors nodes are
installed inside a remote forest to transmit audible information to the station below, which is
located by the forest department. Although the sensors nodes within the forest are made with GPS
connectivity, the high cost of its supply motivates the authors of this paper to create a unique style
of localization using Support Vector Machine. This information transfer system extends the life of
the network and reduces congestion during data transfer from the forest area to the sub-station
[7][8].
x The fire alarms and alarm systems have been introduced that support a combination of smoke
sensor and alarm device to create a life safety system style. However, such fire alarms are
sometimes flawed and may respond to unrealistic fire signals that are classified as false alarms.
there is a need for high-quality and intelligent fire systems that use multiple sensor values to detect
real fire incidents. The model has sensors for collecting critical data from sensor nodes where
symbolic logic converts data during linguistic diversity trained at ANFIS to attract potential fire
hazards. The proposed view also creates alerts with a message sent to the users smartphone. The
system uses a smaller size, less expensive sensors and ensures that this solution regenerates.
MATLAB-based simulation was used for the tests and therefore the results show a satisfactory
result [9][10].
2
ICCRDA 2020 IOP Publishing
IOP Conf. Series: Materials Science and Engineering 1022 (2021) 012068 doi:10.1088/1757-899X/1022/1/012068
x The Internet of Things is about connecting things to other people through the internet, forcing
itself because it is a new business approach in many fields. to create a quick and effective response
in real time, IoT improves the process and provides emergency management with the necessary
information and communication to use these assets. with this paper it is proposed that a rapid
response to fire hazards be tested and evaluated using an IoT-based model. Fire is one of the major
causes of accidental death in the world. To make the proposed program use an expensive Wi-Fi
module, gas detection sensor, Flame sensor, awareness buzzer and temperature sensors. The
sensors detect and alert the local emergency with information collected by the system, and alert
organizations such as fire departments, police stations and hospitals by sending accurate location
to both users and users using the module all connected. Therefore, a smart integrated system is
designed for IoT to manage such risks where innocent people and property live [11][12].
x Significant losses and high threats to the environment are common consequences of forest fires.
This work describes a fire control that supports symbolic minds and decision-making processes
that aim to improve fire prevention, detection and combat strategies. In this proposal, the natural
monitoring of a number of hazardous substances is carried out by wireless sensor networks and
analyzed with the proposed wireless controller. In this regard, climate variables, greenhouse gases
and therefore oxygen levels are measured in real time to measure the presence of short-term fire
hazards and to detect the recent occurrence of fires in various forest areas. In addition, the
Analytic Hierarchy Process method has been used to detect fireside spreads, and, where necessary,
environmental alerts are sent via the Web service and received via a mobile system [13][14].
x The proposed method [15] help to detect fires and prevent the spread of fire, with an early warning
system that warns local fire fighters. It also records a surveillance video in front of the center,
helping to identify the causes of the fire, such as heat. Using 3600 rotating cameras and
radiometry thermal sensors, the system detects abnormal temperature in a fireplace and informs
the center station of location details. It also operates a water cannon system in addition to
notifying local pump stations to assist fire fighters. Greece can be especially dangerous in the
Mediterranean world when it comes to wildfires. The danger has been mounting under the
influence of global climate change, and in the summer of 2007 about 200,000 hectares of
cultivated land were burned.
x The proposed system [16] which makes use of temperature and smoke sensors to detect fire
emergencies and sends a wireless response to a server computer and server mobile which will
instantly inform up to 5 remote authorized users by an SMS or a call. The authorized user has
complete access to the system where the user takes appropriate actions suchas turn on or off alert
notification modify the threshold and reset the system. The system lacks direct automatic
communication with the civil defense station as they have no specific system to receive and
interpret the signal. The system does not communicate with the nearby buildings. The computer
server used in the system could be it-self affected by the fire, on the other hand it is not efficient to
have a separate server for each installed system.
x In this system [17], the model is used to detect the fire, directly notify the fire department about
the fire using an IoT server. The system uses different types of sensors to detect fire and alert the
fire department over IoT by activating the fire buzzer there. It sends information about the incident
by using Wireless Fidelity (Wi-Fi) connection to access the IoT server and transmit data over the
3
ICCRDA 2020 IOP Publishing
IOP Conf. Series: Materials Science and Engineering 1022 (2021) 012068 doi:10.1088/1757-899X/1022/1/012068
Internet so that the fire department team can use this information to take the suitable action. This
system has less benefit in the absence of Internet service or slow Internet connection.
x The proposed system [18] used to early detect fire accidents, communicate with respective
authorities and store the data that is related to the fire accident. The system uses smoke and Down
Hill Control (DHC) sensors to detect fire. Whenever the temperature exceeds the limit, the system
communicates with respective authorities. Then it activates fire sprinklers and keep them in
position till temperature is reduced below a specific value. Fire accident related data is stored in a
remote server for future use. In this system, heat sensor damage due to the fire may cause the
sprinkler never turns-off.
3. Implementation results
A fire outbreak can be a major concern in homes, offices, industries etc. It is dangerous and requires
high safety and control to avoid the destruction of life and property. One of the ways to prevent the
risk is to install an automatic alarm detector in the danger areas, so the availability of Arduino based
fire alarm and system was proposed. it is able to automatically detect heat in a given area, sounds an
alarm, cuts to building pipes and sprays water to measure the intensity of a fire. The system uses DHT
11 sensor, buzzer, 5v DC (Direct Current motor), GSM (Global System for Mobile) Module sim800l
to send SMS (Short Message Service) with LCD screen 16X2 and Atmeg328p Microcontroller. At the
top, the goals of this program have been achieved which is why the program has worked so well
[19][20].
The fire detectors include air and soil temperature moisture sensors. Both prices are worth it has
high alarm levels. Both prices must be high in order avoid false alarms caused natural features that can
be increased only one flexibility. For example, because of open air activity, effects such as fog, fog or
rain is often seen as rising within the amount of smoke I felt therefore continuous daylight can do it
rising temperatures. These events should be reported without alarm alarm.
A temperature sensors (LM35) are common devices integrated into a silicon substrate. The result of
temperature sensor are used for alarming the detector to check the actual conditions of environment
and other sensor data and also set as simple comparison for operation of the thermostat or as a heating
event interrupt. Then further, the fire detectot enable the ultrasonic and humidity sensors for
monitoring the actual conditions of forest.
An ultrasonic sensor is used for measuring the distance of forest trees based on ultrasonic waves
and it is faster than audible sounds. An accelerometer sensor (ADXL335) is also used for calculating
the acceleration of wind in the forest to sense the actual speed of the air.
The Humidity Sensors (DHTII) is designed specifically for high volume users. It measures the
humidity of current forest environment and generates an alert if humidity increases which may causes
the fire to the forest. There is a soil moisture sensor also used for testing the soil level of forest ground
as it found more dry level which enhances the occurrences of fire in the forests.
The Light Dependent Resistant (LDR) is used to complete the circuitry of the connected
components and it is light sensitive so as the more light observed its resistance becomes higher and
more voltages are recorded which also alter the fire detectors.
Fire safety can be an important part of the lifestyle. Security measures are being taken against
organizations, businesses and relatives. But forest fires go unnoticed. Forests are essential for human
4
ICCRDA 2020 IOP Publishing
IOP Conf. Series: Materials Science and Engineering 1022 (2021) 012068 doi:10.1088/1757-899X/1022/1/012068
survival and social development. Forests protect the balance of the earth and the environment. This
paper presents an inexpensive automatic fire protection system for the purpose of the forest using the
GSM fire detection system
The outcome of proposed mechanism is a successful running system which gives a solution to our
problem statement and gives justice to our problem objective. A system has been developed that helps
us in the early detection and prediction of forest fires. The results have been summarized here:
All hardware components connected in our work are giving correct values and there is no
defect in them.
All software components are working correctly there by showing no error as such.
The sensors are detecting the environment for the conditions and sending the environmental
conditions values to our cloud Thingspeak with the help of NodeMCU .
The cloud API is continuously displaying the current values in the form of a graph which
helps us in continuous monitoring of the environment.
Different graphs for different sensors used in our paper of early detection and prediction of
fires (Figure 1-4).
The threshold values of all the sensors used have been preset in cloud API.
As soon as the present values exceed the preset values, an email is triggered to the admin
which gets information about the environmental conditions not being in the ideal state.
As long as present values are less than preset threshold values, there shall be display of all
values with respect to time in the form of graphs.
An easy and continuous monitoring of the environmental conditions of forests can be
achieved.
Thus, presented mechanism helps us to achieve the objective of making this paper which helps in
prediction and early detection of forest fires.
Figure 1. Analysis of result of accelerometer sensor
5
ICCRDA 2020 IOP Publishing
IOP Conf. Series: Materials Science and Engineering 1022 (2021) 012068 doi:10.1088/1757-899X/1022/1/012068
Figure 2. Analysis of result of Soil Moisture Sensor
Figure 3. Analysis of result of LDR Sensor
Figure 4. Analysis of result of UltraSonic Sensor
6
ICCRDA 2020 IOP Publishing
IOP Conf. Series: Materials Science and Engineering 1022 (2021) 012068 doi:10.1088/1757-899X/1022/1/012068
It has observed that all the sensors are tested and performed efficiently over the
ThingSpeak as shown in figure 1-4. Its results are recorded and analyzed which indicates the
fire detection system with IoT can perform and can be simulated.
4. Conclusion and Future Scope
In this paper, an IoT Infrastructure for forest fire detection system has been designed to detect fire at
earlier stage. The most important objective in fire surveillance is early and reliable detection and
localization of the fire. The data collected through different sensors located at different locations and
information will be sent to Arduino Uno placed in various places. As a result the proposed model is
tested with ThingSpeak web interface and sensor data is recorded for testing of fire detection.
The system will process the information and send mail using GSM module to the nearby fire
service station with the location of the fire using GPS module. Several types of sensors used in the
system are DHTII: Humidity sensor, soil moisture sensor and LM35: Temperature sensor, LDR
sensor, ADXL335: Accelerometer sensor, Ultrasonic Range sensor. Early detection and prediction
will lower the count of forest fires in the entire world and save our planet Earth.
In future, we can develop this model to minimize the energy consumption of all sensors and
complete networks considering the node distribution among clusters using distributed sensing.
References
[1] Varela N, Martinez D, Jorge L, Ospino A and Zelaya N A L 2020 Procedia Comp. Sc. 175 pp
435-440
[2] Vikram R, Sinha D, De D and Das A K 2020 Wireless Networks 26 pp 5177-5205
[3] Dauda M S and Toro U S 2020 Int. J. Engg. Applied Sc. and Technol. 4 11 pp 447-453
[4] Shukla A, Kumar S and Singh H 2020 Int. Arab J. Inf. Technol. 17 2 pp 225-232
[5] Saranya N, Sahana S, Suganthi B, 2020 Int. J. Innov. Technol. and Expl. Engg. 9 6 pp 1835-1838
[6] Sarwar B, Bajwa I S, Jamil N, Ramzan S and Sarwar N 2019 Sensors 19 p 3150
[7] Shukla A, Kumar S and Singh H 2019 J. Chin. Inst. of Engg. 42 7 pp 583-592
[8] Ghosrani S, Jadhav A, Lekhak K and Chheda D 2019 Int. J. Research in Engg. Sc. and Mgmt. 2 4
pp 442-445
[9] Mahajan R, Yadav A, Baghel D P, Chauhan N, Sharma K and Sharma A 2019 Int. Research J.
Engg. And Technol. 6 8 pp 4623-4625
[10] Saikumar T and Sriramaya P 2019 Int. J. Recent Technol. and Engg. 7 6S4 pp 429-433
[11] Sathya D 2019 Int. J. Engg. and Technol. 8 6S3 pp 1138-1142
[12] Castro J T, Gil P C, Perez N R, Gonzalez I S, Goya C H and Colomo R A 2018 Complexity 2018
pp 1-17
[13] AlKhatib A A A 2017 Forest Fire Monitoring Intech Open Chapter 3 pp 54-72
[14] Singh D, Sharma N, Gupta M and Sharma S 2017 Int. J. Engg. Sc. and Comp. 7 5 pp 10857-
10860
[15] Pico A M, Frau D C, Araujo A, Alejandre J and Rozas A 2016 J. Sensors 2016 pp 1-8
[16] Dener M, Ozkok Y and Bostancioglu 2015 Procedia Soc. and Behav. Sc. 195 pp 1846-1850
7
ICCRDA 2020 IOP Publishing
IOP Conf. Series: Materials Science and Engineering 1022 (2021) 012068 doi:10.1088/1757-899X/1022/1/012068
[17] AlKhatib A A A 2014 Int. J. Distr. Sensor Networks 10 3 pp 1-12
[18] Sifakis N I, Iossifidis C, Kontoes C and Keramitsoglou 2011 Remote Sens. 3 pp 524-538
[19] Sarkar S M et al 2010 Int. Conf. Comp. Infor. Technol. pp 193-197
[20] Sahin Y G and Ince T 2009 Sensors 9 3 pp 1485-1498
