As California residents, our interest in developing The Hermes Network spurred from the prominent 2018 California WildFire season that ended with the deaths of over 85 individuals and the destruction of 8.6 million acres of land. The limited amount of fire department resources and the sheer scale of wildfires prompted the development of our long-range, low-powered IoT mesh network that aims to help fire departments improve their response times and predict the likelihood of wildfires in certain regions given sensor data aggregated through various low cost, low powered IoT nodes.
What it does
We have built an extensible IoT mesh network that aggregates sensor data and detects wildfires and predicts the likelihood of wildfires given various environmental parameters. The data collected by the sensors are stored in MongoDB, and the data is used to update the front end Flutter UI.
How we built it
Our project is built upon three primary components: our hardware product with sensors, our prediction algorithm, and our UI integration with MongoDB. Our topology (attached below) explains our general approach.
Challenges we ran into
Our team is relatively new to hardware, and this is our first hardware hack. Furthermore, we had difficulty integrating all the hardware and software subsystems, particularly data processing and UI output.
Accomplishments that we're proud of
A functioning extensible mesh network. Integration with front-end Flutter UI Functioning static prediction algorithm given historical data. Good cross-validation Implementation of MBSE per Northrop Grumman specifications.
What we learned
This hackathon was quite a fruitful learning experience as we familiarized ourselves with working with hardware sensors and mesh network technology. Furthermore, we learned to integrate hardware data with a database, algorithm, and UI.
What's next for The Hermes Network
We plan to scale up the application to include better quality sensors and improve the detection algorithm through a DNN regression algorithm. Furthermore, we plan to improve UI components and improve data integration. Overall, we plan to make this platform as accessible as possible with cross-platform support.