AI for Climate Change and Disaster Prediction

Inspiration

The increasing frequency of natural disasters due to climate change inspired us to develop a solution that utilizes AI to predict and mitigate their impact. Witnessing communities struggle with disasters motivated us to leverage technology to create a more resilient future.

What it does

Our project is an AI-powered disaster prediction system that analyzes satellite imagery, weather data, and environmental factors to provide early warnings for natural disasters. By delivering accurate predictions, we aim to help communities prepare and respond effectively, reducing loss of life and property.

How we built it

We gathered historical and real-time data from sources like NOAA, Google Earth Engine, and OpenWeatherMap. Utilizing machine learning techniques such as CNNs for image analysis and LSTM for time-series forecasting, we trained our models on this data. A user-friendly dashboard was created to display predictions and alerts in real-time.

Challenges we ran into

We faced challenges in data preprocessing, especially ensuring the quality and accuracy of satellite images. Balancing model complexity with real-time processing requirements was another hurdle. Additionally, integrating various data sources into a cohesive platform required significant effort.

Accomplishments that we're proud of

We successfully built a prototype that accurately predicts specific types of disasters, demonstrating a significant improvement in prediction accuracy compared to existing models. The user interface allows for seamless interaction and real-time updates, which is a significant milestone for our project.

What we learned

This project taught us the importance of interdisciplinary collaboration between AI and environmental science. We gained valuable experience in handling large datasets and learned to adapt our models based on real-world conditions. We also discovered the necessity of user-centric design in developing impactful solutions.

What's next for AI for Climate Change and Disaster Prediction

We plan to refine our model for broader disaster types and integrate additional data sources. Future developments will include partnerships with local governments and NGOs for real-world testing and deployment, ultimately aiming to expand our platform's reach globally.

Built With

Submitted to

Google's Women Techmakers presents She Builds AI

Created by

As a key member of the team developing the AI-powered disaster prediction system, I conducted extensive research on natural disasters and identified key data sources, including satellite imagery and weather data, to inform our model’s development. I played a significant role in selecting appropriate machine learning techniques, particularly Convolutional Neural Networks (CNNs) for image analysis and LSTM models for time-series forecasting. I took charge of data preprocessing, ensuring the datasets were clean and well-structured, and I implemented feature engineering to enhance model performance. Additionally, I collaborated with the team to design a user-friendly dashboard, focusing on intuitive access to critical information for vulnerable communities. My contributions also included developing and executing testing protocols to validate the accuracy and functionality of the AI model and user interface. Throughout the project, I fostered collaboration and actively engaged with my team members, facilitating discussions and promoting a positive team dynamic. Overall, my efforts spanned research, model development, user interface design, and teamwork, which were essential in bringing our project to fruition.

Neha Mohapatra

Updates

Neha Mohapatra started this project — Oct 11, 2024 02:25 AM EDT

Leave feedback in the comments!

Log in or sign up for Devpost to join the conversation.