0% found this document useful (0 votes)
68 views3 pages

AI Powered Disaster Management System

Disasters or hazards have the potential to cause catastrophic damage and significant socioeconomic loss.
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
68 views3 pages

AI Powered Disaster Management System

Disasters or hazards have the potential to cause catastrophic damage and significant socioeconomic loss.
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 3

Volume 8, Issue 4, April – 2023 International Journal of Innovative Science and Research Technology

ISSN No:-2456-2165

AI Powered Disaster Management System


Ashna Lakshmanan Sangeerth S Nambiar
TLY19CS0016 TLY19CS048
Department of Computer Science Department of Computer Science
College of Engineering Thalassery College of Engineering Thalassery

Nasariya Parvin M Sanya


TLY19CS037 TLY19CS049
Department of Computer Science Department of Computer Science
College of Engineering Thalassery College of Engineering Thalassery

Nikhil Dharman M K
Prof., Department of Computer Science
College of Engineering Thalassery

Abstract:- Disasters or hazards have the potential to damage to life and property. Floods are an annual event in
cause catastrophic damage and significant socioeconomic India and disaster management has changed little from year
loss. The year 2022 has been recorded as the eighth to year. Nearly 15% of India is at risk of flooding. The
consecutive year with 10 or more billion-dollar weather Times of India reports that 2,000 lives are lost each year,
or climate related disasters. As a result, an AI powered 80,000 hectares of farmland are damaged, and costs around
Disaster Management system is developed. It aims to Rs 180 crore.
strengthen disaster mitigation strategies using AI
technology. It helps in detecting and preparing for the The system is actually an AI powered disaster
extreme weather and other hazards, and also to management service, that is through the internet. This is a
communicate to people and communities effectively method of providing services to affected victims in the
about the necessary response. AI helps response teams to situation of a disaster. It is the procedure to predict any
understand the hazards or accidents, monitor events in possible disaster and provide quick responses to the public.
real time and anticipate specific pitfalls in the face of It aims to strengthen disaster mitigation strategies using AI
impending or on-going disasters. The disasters can either technology. It helps in detecting and preparing for the
be predicted with the help of AI technology by training extreme weather and other hazards, and also to
machine learning models or be detected from live news communicate to people and communities effectively about
feeds. AI is used during different phases of disaster the necessary response. AI helps response teams to
operation first, vaticination and protuberance; also, to understand the hazards or accidents, monitor events in real
help communicate what has passed; and in the time and anticipate specific risks in the face of impending or
monitoring and early discovery of implicit new pitfalls. on-going disasters.

I. INTRODUCTION II. LITERATURE SURVEY

Natural disasters are situations that no one can control. The paper [1] proposes an IoT-based flood monitoring
Natural disasters kill thousands of people every year and and artificial neural network(ANN) based flood forecasting
cause billions of dollars in economic losses. No country or designed to increase scalability and reliability of flood
community is immune. In recent years, the number of control systems. The main objective of this system is to
natural disasters has increased, and along with that, the monitor humidity, temperature, pressure, precipitation, river
damage caused by urbanization and population growth has water level and find their temporal correlation information
increased, and the impact of natural disasters has come to be for floods predictive analysis. The IoT approach is deployed
felt more strongly. Of all the continents, Asia is considered to collect data from sensors and Wi-Fi communication and
the most disaster-prone. Between 1991 and 2000, Asia ANN approach is used for data analysis in flood prediction.
accounted for 83% of the world's population affected by
disasters. India is very prone to natural disasters and the The paper [2] is focused on the development of an
country regularly experiences very severe natural disasters. effective flood prediction system using machine learning
In the case of India, among the different types of natural (ML) algorithms that can help prevent loss of human life
disasters that affect different parts of the country, floods, and property. The paper use k-nearest neighbors (KNN),
hurricanes, earthquakes and droughts cause the greatest support vector machines (SVM), random forests (RF) and
damage to life and property. Heat waves, cold snaps, decision trees (DT) to build ML models. And to solve the
avalanches, landslides and fires regularly cause great problem of oversampling and low accuracy, a stacking

IJISRT23APR2182 www.ijisrt.com 1954


Volume 8, Issue 4, April – 2023 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
classifier will be used. The paper uses precision, f1-score, IV. PROPOSED SYSTEM
recall and precision to compare between these models. The
results suggest that layered models are best for forecasting In the proposed system, we introduce an efficient
floods due to real-time rainfall in a given area. system to manage disasters with minimum manual work. In
the proposed system, it uses the technology of AI to enhance
Paper [3] proposes a fully functional and efficient the working of the existing processes of disaster
earthquake detector cum forecaster based on deep neural management. Phases like: mitigation, preparedness,
networks of long-short-term memory (LSTM) units. The response and recovery are processes that are improved by
model captures inherent temporal characteristics of the AI technology. Our proposed system helps enhance
earthquake data.the proposed LSTM model shows disaster mitigation strategies using AI technology. Helps
satisfactory performance for small to medium-sized you recognize and prepare for extreme weather and other
earthquakes and also implements a baseline artificial neural hazards. Send notifications of nearby disasters, effectively
network (ANN) model to perform a suitable comparison. informing people and communities about required actions.
The AI techniques used to support disaster management in
The paper[4] proposed six different machine learning all phases consist of several ML and DL methods. The main
classifiers namely Artificial Neural Network, Random Tree, disasters to be considered for this project are earthquakes
CHAID, Discriminant, XGBoost Tree, and Tree-AS on six and floods. With the help of an artificial neural network, the
datasets of different regions of India. All the algorithms datasets are trained and modelled for it to predict the
have been applied to each dataset. The objective of this happening of the disaster. Flood prediction uses K-Means
paper is to predict the value of magnitude for the future and earthquake prediction uses LSTM. LSTM stands for
earthquake in India and nearby regions from the historical “Long Short-Term Memory”
data on earthquakes. From the result, It has been observed
that for Andaman & Nikobar dataset XGBoost Tree The proposed system also uses live news feeds and
achieved the highest accuracy with 99.10%, for the Gujarat social media analytics to send disaster notifications. AI can
dataset Tree-AS achieved the highest accuracy with 91.67%, evaluate and authenticate photos and comments on Twitter,
for the North India dataset Artificial Neural Network Instagram, and YouTube in real time to distinguish between
achieved the highest accuracy with 99.13%, for North East true and false information. These key statistics help field
India dataset XGBoost Tree achieved the highest accuracy responders get to problem locations faster and focus their
with 99.04%, for Nepal-UP-Bihar dataset XGBoost Tree efforts on the most vulnerable. This 4 information also helps
achieved the highest accuracy with 99.01%, for Nearby reduce the time it takes rescue teams to find victims.
India’s Country dataset XGBoost Tree achieved the highest Additionally, artificial intelligence (AI) and predictive
accuracy with 92.12%. From all the results, it has been analytics tools can scan digital information from Twitter,
noted that the XGBoost tree classifier performed well in Facebook, and YouTube to provide early warning, ground-
most datasets., the Curve has been made between magnitude level location, and real time incident verification .
& gap, magnitude & magnitude error, and magnitude and
depth error for finding the mathematical relation between V. CONCLUSION
them.
The Project: An AI powered disaster management
III. EXISTING SYSTEMS system aims towards providing a system which can provide
responses quickly. Disaster management manages disasters
In the existing system, in India, the main goal of over time. The four phases of disaster management:
disaster management is to mitigate or avoid damages that mitigation, preparedness, response and recovery are widely
may be caused by hazards and disasters. It also ensures that accepted. The AI techniques used to support disaster
prompt, appropriate and prompt action can be taken for management in all phases consist of several ML and DL
effective recovery. The National Disaster Management methods. ML methods like K-Means clustering algorithms
Agency (NDMA) is currently the supreme agency and DL methods like Long Short-Term Memory (LSTM).
responsible for disaster management in India. The NDMA is ML and DL have enabled the development of systems
chaired by the Prime Minister of India. The Disaster capable of predicting, responding to, and recovering from
Management Act 2005 defines disaster management as the disasters using large and complex data sets, resulting in
integrated process of planning, organizing, coordinating and practical decision support tools. These ways work the
implementing actions to effectively manage disasters. In capability to manipulate different types of data from
India, government agencies lack proper disaster multiple sources and determine patterns that can give
management training and are ill-equipped to deal with information that can not else be bared.
natural disasters through effective mitigation and
preparedness measures. Disasters and their management are
commonly discussed in terms of their consequences, but in
practice this should lead to the planning and preparation of
strategies to deal with and mitigate disasters in a responsible
and effective manner.

IJISRT23APR2182 www.ijisrt.com 1955


Volume 8, Issue 4, April – 2023 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
REFERENCES

[1]. S. Bande and V. V. Shete, "Smart flood disaster


prediction system using IoT & neural networks," 2017
International Conference On Smart Technologies For
Smart Nation (SmartTechCon), 2017, pp. 189-194, doi:
10.1109/SmartTechCon.2017.8358367.
[2]. R. Balamurugan, K. Choudhary and S. P. Raja,
"Prediction of Flooding Due to Heavy Rainfall in India
Using Machine Learning Algorithms: Providing
Advanced Warning," in IEEE Systems, Man, and
Cybernetics Magazine, vol. 8, no. 4, pp. 26-33, Oct.
2022, doi: 10.1109/MSMC.2022.3183806.
[3]. B. Bhargava and S. Pasari, "Earthquake Prediction
Using Deep Neural Networks," 2022 8th International
Conference on Advanced Computing and
Communication Systems (ICACCS), Coimbatore,
India, 2022, pp. 476-479, doi:
10.1109/ICACCS54159.2022.9785011.
[4]. Chittora, Pankaj & Chakrabarti, Tulika & Debnath,
Papiya & Gupta, Amit & Chakrabarti, Prasun &
Praveen, s & Margala, Martin & Elngar, Ahmed.
(2022). Experimental analysis of earthquake prediction
using machine learning classifiers, curve fitting, and
neural modeling. 10.21203/rs.3.rs-1896823/v1.

IJISRT23APR2182 www.ijisrt.com 1956

You might also like