Inspiration
California is one of the most wildfire-prone landscapes in the world. With climate change, southern California is seeing warmer temperatures, reduced snow pack, drought, and increased Santa Ana winds, all of which increase wildfires. San Diego has had three historical "100 year" wildfires in the last 20 years, creating billions of dollars of damage. Most wildfires are not preventable, so we need to improve our response system to prevent a fire getting out of control. San Diego's Climate Action Plan addresses climate resiliency and prompts us to ask, what can we do to mitigate the negative consequences of wildfires?
What it does
Our solution is to create a large network of sensors in San Diego forests to report key data for detecting fires, hence a smart forest. This uses the internet of things (aka IoT) - the idea that if we have enough distributed sensors, we can collect information about an environment and automate software to have the environment draw it's own conclusions (such as wildfire detection) and alert emergency response. Once the sensors and software are installed, the process is fully automated - no humans required to oversee the data.
This solution is twofold: For an immediate effect, we can detect a wildfire early and notify response services of it's location, growth, and movement. This helps control the size of the fire and improve resource allocation during high risk times. For a long term effect, we'll be collecting more detailed data for improved predictive capability. For example, we could learn about locations where fires started but burned out.
We can now implement a system like this because sensors and wireless networks have become very cheap in the last few years. Before this system may have been too cost prohibitive to consider.
How I built it
An infrared sensor connects to a secure wireless network. The sensor collects temperature readings and transmits over the network to a server using RESTful APIs. The server is written in node js. It collects and augments the data (with location, timestamp, etc), stores it, analyzes it, generates events on it and stores the events for later analysis.
Challenges I ran into
The biggest challenge was building a scalable solution that can work with a large number of devices and respond in real time to events generated by the sensors. Also selecting the correct technologies to enable this system is important.
Accomplishments that I'm proud of
We completed an end-to-end solution, not just one part, for this problem. We integrated with another project during the hackathon that alerts emergency response teams with the location and issue on a visual display (such as iPad).
What I learned
We learned about an issue with great importance to the future of San Diego. We learned about existing solutions and their shortcomings. We came up with an innovative idea to address a gap in those solutions to be a part of a complete fire detection picture.
What's next for Smart Tree Network
This project has growth potential also for an urban tree network. One challenge related to the Climate Action Plan is how to determine and quantify the influence of urban tree stormwater capture. The same platform can be extended for this problem. For example, use a humidity sensor in the soil/roots of urban trees to track water absorption over time. From the data, we could create a model determining the effect of urban trees on flooding and water capture. The implementation would be similar to the wildfire project - install sensors and collect data across the city for analysis. Sensors could immediately alert authorities to flooding, or long term find trends in water flow.
Log in or sign up for Devpost to join the conversation.