Inspiration
The inspiration for this project comes from EOG's Prompt to optimize and analyze Methane gas emissions. We chose this project due to our passion for environmental safety, which aligns with EOG's Environment-friendly mission.
What it does
The project takes in sensor readings and weather parameters to alert when potential gas leaks are possible while predicting the potential problem areas for gas leaks.
How we built it
We were able to identify and map various severities for various gas concentration thresholds to identify potential gas leak moments. We then performed an XGBoost model to predict possible gas leak locations to locate and ensure safety when a leak occurs quickly.
Challenges we ran into
Pre-processing data sets in a meaningful way to perform machine learning. We initially struggled with data exploration but found it to be one of the most crucial steps in identifying out features.
Accomplishments that we're proud of
We are incredibly proud of being able to complete an ML project. Learning about different models and the pros and cons of their usage in this challenge was one of our most significant accomplishments.
What we learned
We learned many nuances of successfully predicting future incidents using machine learning as a team where everyone is new to ML. We are proud of how much we have learned and the engaging nature of the challenge. We learned how to efficiently use the sci-kit-learn library along with many other data exploration concepts for providing meaningful insights about leak events.
Log in or sign up for Devpost to join the conversation.