Open WarpWing opened 1 year ago
Apologies for the inconvenience, due to some unforeseen technical difficulties we had to move our Streamlit frontend to https://blazewatch.deltaprojects.dev/ as the old frontend seems to be running a beta version of our front end from a while back.
Thank you for understanding.
:+1: for this.
BlazeWatch: A Blazingly Fast Wildfire Predictive System
Special thanks to @officialdelta (Alan Alwakeel) for the early warning system and formatting of the READMEs)
Project Type
Project Description
π₯ Project Name: BlazeWatch
π Description: BlazeWatch is a three-pronged wildfire predictive system designed to detect potential wildfires in Zakar Island before they intensify. By amalgamating data from temperature sensors, social media messages, and our state-of-the-art ML model, we aim to protect the residents, their property, and the environment.
Components: π Carbon: The frontend is built with Streamlit, offering an intuitive interface for users and officials to visualize and interact with wildfire data. π₯οΈ Oxygen: The robust backend system that efficiently processes data, coordinates between the frontend and the ML model, and ensures real-time performance. π§ Blaze: Our specialized machine learning model trained on data sets provided, leveraging both temperature readings and social media cues to predict potential wildfire outbreaks.
Features: Real-Time Predictions: By harnessing the power of real-time temperature readings and timely analysis of micro-blog posts, Blazewatch can detect potential wildfire hotspots. Interactive Dashboard: Through Carbon, our Streamlit interface, users can view temperature distributions, tweets, and predicted wildfire alerts across Zakar. Early Warning System: Upon detection of potential wildfire risks, notifications are generated and forwarded to the designated zakar-fire-predictions station, ensuring timely interventions.
How it Works: Data Collection: Temperature readings from zakar-temperature-readings, micro-blog posts from zakar-tweets, and past wildfire notifications from zakar-fire-alerts are ingested into our system. Processing & Analysis: Oxygen processes this data, identifying potential anomalies and patterns. Key textual clues from micro-blog posts hinting at wildfires are detected. Prediction with Blaze: Our ML model evaluates the processed data, gauging the likelihood of a wildfire breakout, ensuring high accuracy. Alert Generation: In case of a probable wildfire, an alert is generated with details like event_day, geospatial_x, and geospatial_y.
Streamlit link
https://blazewatch.warpwing.cloud/
Source code link
https://github.com/BlazeWatch/carbon
Instructions for the reviewers
Introducing BlazeWatch: A Revolutionary Fire Detection System π₯
At BlazeWatch, we're igniting the future of fire detection with a comprehensive and innovative solution that intertwines the power of machine learning, robust backend management, and a sleek frontend interface.
Components
BlazeWatch consists of three interconnected components, each playing a vital role in the system:
Blaze - Machine Learning Model: Our state-of-the-art ML model that predicts and analyzes fire events with unparalleled accuracy. Learn more and get started with Blaze
Oxygen - Backend Component: The lifeline of BlazeWatch, managing both incoming messages and broadcasting to social media platforms, and orchestrating data with Postgres. Explore Oxygen and breathe life into your backend
Carbon - Frontend Component: The user-facing interface, designed for smooth interaction and rich visualization. Engage with Carbon, the combustion catalyst of BlazeWatch
Each component comes with an extensive README.md, providing detailed instructions for setup and use. Following them to the letter ensures a seamless integration of the entire BlazeWatch system.
Requirements
BlazeWatch is more than just technology; it's a movement to redefine fire safety and awareness. With precision and elegance, we're shaping the future, one flame at a time. Join us in this exciting endeavor and make the world a safer place! π₯
What made you decide to build this particular app?
The decision to build BlazeWatch stemmed from a profound realization of the critical gap in real-time fire detection and response systems. While technologies for fire prevention and mitigation existed, we identified a need for a solution that could fuse the power of machine learning with modern data processing to create a more intelligent, responsive, and holistic system.
Our team was motivated by the potential to save lives and preserve nature by detecting fires at their inception. By leveraging the state-of-the-art ML model Blaze for predicting and analyzing fire events, along with Oxygen and Carbon to manage and present data seamlessly, we believed we could create something that went beyond traditional fire detection.
BlazeWatch isn't just a product; it's our contribution to a safer, more resilient world where technology plays a pivotal role in safeguarding communities from the devastation of fire. The thrill of innovation, combined with the responsibility to make a tangible difference, fueled our passion to bring BlazeWatch to life. π₯
Other information
To delve into the capabilities of BlazeWatch and explore the proficiency of our model, we invite you to take a closer look:
Blaze Repo: Within the Blaze repository, the predict.ipynb file offers usage examples that work seamlessly, giving you an instant understanding of the model's functionality. You can find the file here.
Visualization & Demo: While Streamlit links primarily serve for data visualization projects only, we've gone a step further to set ourselves apart from the competition. A live demo of our project is thoughtfully included in our Carbon frontend, giving you a hands-on experience of BlazeWatch in action.
Our commitment to excellence shines through these features, providing you with tangible insights into what makes BlazeWatch a revolutionary fire detection system. Explore, engage, and feel the heat of innovation with BlazeWatch! π₯
Confirm submission of private form