ISSN 2321 3361 © 2020 IJESC

Research Article Volume 10 Issue No.4

Disaster Alert System

Namit Kumar Singh1, Naveen Yadav2
Department of Computer Science and Engineering
IMS Engineering College, Ghaziabad, India

Abstract:
Devastations of natural disasters are the lashes of mother nature that hit us every year with a whip. These are inevitable. There are no
alternative ways to prevent this incident, but we can take proper steps to reduce its damage. Nowadays a great deal of attention has
been given to the potential applications of mobile communication technology. Short Message System (SMS) has a huge impact on the
communication system. This paper proposes an android application which can alert people before a natural disaster such as cyclone
and flood strike, and tell them the optimal route to the nearest shelter via SMS or voice call. In evacuation process, we have used
partition based shortest technique to find nearest shelter place.

Index Terms: Natural Disaster, Notification System, Android, Location Based, Nearest Shelter

I. INTRODUCTION II. RELATED WORK

Natural disasters are the consequences of natural hazards. It does Early disaster warning and evacuation approach are very general
occur a serious breakdown in the sustainability of human beings, disaster management system in disaster-prone areas. Nowadays
animals, and the property. It also occurs economic losses and mobile phones play an essential role in disaster management
disruption of economic and social progress. The overwhelming system in several ways: monitoring, communiate, warning
number of dead or seriously injured and homeless people are dissemination, evacuation and rescue and relief aid. A number of
affected by the occurrence of a natural disaster. A massive notification system has been proposed in our real world. In
amount of money needs to be spent for reconstruction and research paper, Short Message Service (SMS) is sent to all
rehabilitation after the natural disaster. These are nothing but citizens from the server about the awareness of upcoming flood
extreme environmental events that impact human activities. warning. A huge number of SMS transfer from the server cause
Hurricane, earthquakes, tsunamis and volcanic eruptions, as well network congestion and can break the voice communication
as floods, are the most frequent threats. Tornadoes and droughts system in the same network. To avoid this kind of congestion
which are also prevalent. According to the Annual Disaster cell broadcasting service is used to directly send messages to the
Statistical Review 2010 , 330 natural disasters were registered subscribers in a specific area. But this process fails to help in
worldwide in 2003. It was reported that there was a total of evacuation process which provides information about the safe
21,610 people died. And among them, 9,871 people died in region. GSM alarm device is used for evacuation process in
flood, and 8583 people died due to the storm. The estimated which three kinds of warning are sent to the police station or fire
damage was 118.6 billion dollar. Numerous researchers had brigade station Although it can avoid network congestion, the
tried to develop an early disaster warning system for minimizing GSM alarm is not a faster way for evacuation process. Satellite
the potential loss. communication systems will be very fast, reliable and robust. As
They came up with some brilliant ways to do it. In this modern a result, well-developed countries like Australia and South
time, the mobile phone has really changed the way of Korea are planning to use satellite communication for disaster
communications. It is now the most used communication tools. management. This satellite service maintenance is expensive and
Some of the researchers tried to use Short Message System developing countries cannot afford this. Very few researchers
(SMS) as an alerting system for the disaster. Our proposed propose location based services for disaster management on
technique is an android based application that takes the weather mobile phones and evacuation system by visual and audio
updates from websites and calculates the disaster level. warnings useful to both normal and blind people. Natural
Subscribers data can be stored in this applications database. Disasters such as cyclone, storm, earthquake, Tsunami, and
It calculates the optimal route to the shelter from the flood have shown the harmful, damaging mode of nature which
subscriber’s current position and sends voice call/SMS with the has taken millions of lives including people and animals. Above
warning and the shelter-locations to them. This is GSM type all techniques are concern about people warning. Most of them
alerting system, so the subscribers do not need to have an did not discuss evacuation processes by which people get shelter
android device in order to receive the service. It also stores place. Dijkstra’s algorithm based shortest path calculation is
previous weather statistics data in the database. Using these data used to find the nearest shelter place. This process is time-
and real-time data, it can predict the upcoming weather by using consuming because it searches all paths from the source to
machine learning techniques. Here it uses the naive Bayes destination. For this, we propose a location based smart disaster
algorithm to predict the weather by analyzing the previous and management system that can warn all subscribers (people, blind
real-time data. people). In the evacuation process, we used partition based

IJESC, April 2020 25493 http:// ijesc.org/

trajectory to find nearest shelter place. It is very faster than the disaster occur at the different time in the different geographical
previously proposed methods. area. Some recent examples of violent natural disasters are the
2011 Japan earthquake and Tsunami, the 2010 Haiti earthquake,
III. PROPOSED SOLUTION the 2007 cyclone SIDR, the 2004 Indian Ocean Tsunami, the
1991 Bangladesh cyclone. Geo-graphically few South Asian
The main purpose of Smart Disaster Notification System is to countries are situated in between the Himalayas and the ocean,
alert people before the time of disaster and tell them the optimal on the delta of wide rivers, means that the countries are very
route to the nearest shelter. In our system, we divide the whole exposed to flooding. The people live in coastal areas have to
application in some module. First one is database building that face several storms each year and cultivable lands disappear in
consists of subscriber information and locations whose the river due to river erosion. Such countries are most affected
probability of disaster is measured. The second one is based on by the planets climate changes and the number of cyclones.
locations in which our proposed methods take information from Hence, there is also the risk of Tsunami in these countries. Our
the weather website. After that these information will be disaster preparedness system protects the people from upcoming
converted to JSON (JavaScript Object Notation) format. Then, disaster. For this, it uses SMS, voice call or voice alert. Our
according to JSON information, the system will be able to proposed work can be implemented on Android mobile phones.
understand the probability of disaster and then the system will Android is an operating system for mobile devices such as smart
send nearest shelter information to the subscribers. phones and tablet computers developed by Open Handset
Alliance led by Google. As android is more open and
IV. MODULES AND THEIR DESCRIPTION comprehensive than other mobile operating systems, this is the
best selling product worldwide. It also allows the building of
The system comprises of 2 major modules with their sub- new applications at lower cost. Consequently, this is more
modules as follows: interactive for users. Hence, an android mobile platform has
been used in our proposed disaster awareness system.
Admin (PHP)
B. Input into database
Login: Admin can login using credentials.
We will keep the record of subscribers in the database. In there
Manage Safe houses: Admin can add, update, delete and View we will store the subscriber name, location, mobile number.
for safe houses. They can add location of safe house using Based on the subscriber locations we can give them the proper
Google Map warning about the disaster. Smart disaster notification system
will fetch the information from it and sends a notification to the
Add Extreme Weather Updates: Admin can add, update, subscribers. Sending notification will depend on the update from
delete and view extreme weather updates, they can add location the websites.
of using Google Map
C. Update from website
Add Evacuation Updates: Admin can add, update, delete and
View Evacuation updates. Hey can add location of using Google This application will take the update from the website and
Map and add permiter of location that will affect. evaluate the level of disaster. Then it will convert the data into
View Users: Admin can view user. JSON format. The following figure 1 shows the update process.

User (Android) D. Minimum distance calculation

We determine the optimal route to the nearest shelter and show
Register: User can register and obtain credentials. it to the application and in case of the non-android user we just
Login: User can login using credentials. give them the placement of the nearest shelter.

Profile: User can edit their profile.

Change Password: User can change their password in case of
emergency.
Home: Here user can view current weather updates like temp,
wind etc. It will also Show Alerts if Any (Extreme or
Evacuations) according to my location. Nearest Safe houses and
direction in Google Maps. Figure.1.Update from website
Notification: It will also show Alerts Notification The system will send nearest shelter information. Here if we
calculate Euclidean Distance we find nearest shelter point is B.
A. Preliminaries But that existing path is so far than path A. For that reason we
The consequence of natural hazards is called the natural disaster. use trajectory partitioning method to calculate the distance
There are different kinds of a natural disaster such as cyclone, between two places. Trajectory partitioning means path
storm, earthquake, Tsunami, flood etc. Different kinds of natural partitioning, which is very important because proposed
algorithm has used sub-trajectories.

IJESC, April 2020 25494 http:// ijesc.org/

 Figure.2. Calculation of nearest shelter place.

For partitioning, we first took the starting and ending point of whether it would need to partition the trajectory. If the distance
the route. Then from the starting point, we move towards using is greater than the given limit, we would partition the trajectory
the trajectory points. In the first trajectory point, we measured into two parts. Figure 3(a) shows the approximate solution
the perpendicular distance with respect to the line which was structure. The other partition will replace the existing line which
drawn from the starting and ending point of the route. We used was used to measure the perpendicular distance. Then from
equation 1 to measure the perpendicular distance where (m, n) is every point, we follow these steps and determine the trajectory
the coordinate of the trajectory point, d is the perpendicular path and store them in the array and finally shows the result. In
distance, and Ax + By + c = 0 is the equation of the line. Then figure 3(b) P 3 trajectory
we check the perpendicular distance to the given limit to check

Figure. 3. Calculate Minimum distance function for path shelter

IJESC, April 2020 25495 http:// ijesc.org/

 Table.1. Alert Classification F. Sending notification
Disaster SMS Voice Alert
Rainfall Yes Yes After getting updates from websites minimum distance is
calculated and then a notification is sent to the subscriber who
Heavy Rainfall Yes Yes already registered in the database. This notification is both audio
Cyclone Yes Yes and text message because subscriber can be blind.
Wildfire Yes Yes
Flood Yes Yes G. Location tracking of victim

Above Table predicts rain/ flood possibility by analyzing At the time of disaster, subscribers are in the middle of it. Then
weather statistics data. It uses Naive Bayes algorithm to find the for rescue, this application determines the victims location by
pattern from the weather statistics data. Equation 2 is for Naive using GPS for android or triangulate location using the mobile
Bayes classifier. network for non-android phone and send back the data to the
rescue center.

V. IMPLEMENTATION

Where X represents a vector and C represents class. In our implementation, we have used Android based smart
technology. For alerting people, we first insert their data into the
E. Predicting weather applications database. In the database, subscriber locations are
also saved. Then the application communicates with the server.
Smart disaster notification system stores weather data into the Determining the kind of situation server responses with a JSON
database. Then it applies machine learning techniques to find the file containing the weather information. This application reads
pattern from these data. After that, it predicts whether there will the JSON file and converts the data into a application
be any disaster by using real time demonstration.

IJESC, April 2020 25496 http:// ijesc.org/

Message which would be sent to the people whose data are in VI. RESULT
the database. If JSON file gets some disaster like as Cyclone,
Flood, Wildfire it sends SMS or voice call to the subscribers. In our system, we have warned people before disaster via SMS
also told them the optimum route to the shelter. A lot of works
A. Database Design exists with the same types of feature but our contribution is we
have used partition based trajectory distance where others
This is an Android application so we used SQLite database. This system used Dijkstra’s algorithm or Euclidean distance to
database contains three attributes which are subscriber name, measure the optimum path to the shelter position. A comparison
mobile number, region. The attribute mobile number acts as a of our works with other works is shown in table II. Table III
primary key because the mobile number is unique for shows the comparison between Euclidean distance and trajectory
subscribers. In the weather server, we create database tables for distance. We can see that for every test case the Euclidean
storing the weather status and to store the position of the shelter. distance is smaller than the trajectory distance. Because
Euclidean distance only calculates through the direct path using
B. Application Testing coordinate of two places. But there may not be direct paths from
those two places. In this case, trajectory dis-tance is more
Figure shows that statistics of rain on some days. By running accurate than the Euclidean distance.
naive Bayes algorithm it predicts that heavy rainfall may cause
the flood and it can affect some areas. So the JSON file points to Table.II. Comparison with other works
those possible affected areas. If the statistics show no promising
situation of disaster system, it points to no warning in figure.
The system searches if there is anyone exist in affected regions,
it sends them the alert. By this, only the affected people get
alerts, not all the people in the database. Table I shows the
overview of the alert classification. Figure shows inner looking
of the application. For inserting data into the database users need
to press the create account button and give the information.
After completing registration, data of the people are saved in the
database. A user also can edit or delete account information by
choosing Edit Account Menu. After pressing the Edit Account
button there are two options edit account or delete accounts. If
the user wants to edit account, then the user has to provide the
mobile number which is unique (primary key) and then edit the Hence, we can calculate the nearest shelter and recommend the
information. For deleting account user have to again give the subscriber. The approach uses Dijkstra’s algorithm to find the
mobile number. The system will then search the database nearest shelter where as we have used partition based trajectory
corresponding that mobile number and will delete that distance. Also, our application gives weather prediction using
information. When apps requests for weather updates, website real-time and archived weather data.
calculates the situation of the disaster and make JSON report
according to the condition of the weather. Table.III. Trajectory path result

VII. CONCLUSIONS
This was our project of System Design about “Disaster Alert
System” developed in Android as well as web application based
on Java and PHP programming language. The Development of
this system takes a lot of efforts from us. We think this system
gave a lot of satisfaction to all of us. Though every task is never
said to be perfect in this development field even more
improvement may be possible in this application. We learned so
many things and gained a lot of knowledge about development
field. We hope this will prove fruitful to us.

VIII. REFERENCES

[1]. Mohamad Sukeri Bin Khalid and Shazwani Binti Shafiai.

Flood disaster management in malaysia: An evaluation of the

IJESC, April 2020 25497 http:// ijesc.org/

effectiveness flood delivery system. International Journal of
Social Science and Humanity, 5(4):398, 2015.

[2]. Debby Guha-Sapir, Femke Vos, Regina Below, and Sylvain

Ponserre. Annual disaster statistical review 2010. Centre for
Research on the Epidemiology of Disasters, 2011.

[3]. Steven L Zimmers and Daniel W Davis. Alert notification

system, April 21 2015. US Patent 9,015,256.

[4]. Jovilyn Therese B Fajardo and Carlos M Oppus.

Implementation of an android-based disaster management
system. In EHAC’10 Proceedings of the 9th WSEAS
international conference on Electronics, hardware, wireless and
optical communications, pages 126–130, 2010.

[5]. Jung-Hua Lo and Yi-Xiang Tsai. Developing a real-time

emergency re-sponse system using ios as an example. Scientific
Journal of Information Engineering, 5(3), 2015.

[6]. Mark D Laird and Michael Glier. Mobile emergency

notification system, May 22 2007. US Patent 7,221,928.

[7]. Kazi Mujibur Rahman, Tauhidul Alam, and Mashrur

Chowdhury. Lo-cation based early disaster warning and
evacuation system on mobile phones using openstreetmap. In
Open Systems (ICOS), 2012 IEEE Conference on, pages 1–6.
IEEE, 2012.

[8]. IMRAN Mahmud, JHUMANA Akter, and SHAHRIAR

Rawshon. Sms based disaster alert system in developing
countries: A usability analysis. Internation Journal of
Multidisciplinary Management Studies, 2(4), 2012.

[9]. Laurence Beddome Turner. Wireless Telegraphy and

Telephony. Cam-bridge University Press, 2013.

[10]. Nuwan Waidyanatha, Dileeka Dias, and Harsha

Purasinghe. Challenges of optimizing common alerting protocol
for sms based gsm devices in last-mile hazard warnings in sri
lanka. In Wireless World Research Forum Meeting, volume 19,
2007.

[11]. Wikipedia. Naive bayes classifier, 2016. [Online; accessed

24-May-2016].

[12]. Marius Cioca, L-I Cioca, and S-C Buraga. Sms disaster
alert system programming. In 2008 2nd IEEE International
Conference on Digital Ecosystems and Technologies, 2008.

[13]. Duke H Jeong. National disaster warning system in korea,

2009.

[14]. Tobias Scherner and Lothar Frisch. Notifying civilians in

time-disaster warning systems based on a multilaterally secure,
economic, and mobile infastructure. AMCIS 2005 Proceedings,
page 127, 2005.

[15]. Gamini Jayasinghe, Farazy Fahmy, Nuwan Gajaweera, and

Dileeka Dias. A gsm alarm device for disaster early warning. In
Industrial and Information Systems, First International Confer
ence on, pages 383–387. IEEE, 2006.

IJESC, April 2020 25498 http:// ijesc.org/