The groundbreaking device that we have designed is the Tree of Life. True to its name, our team’s device exemplifies the idea that all life on Earth is connected and is our contribution toward having a positive impact on life on Earth. Our novel IoT device will not only monitor carbon emissions in corporate and community environments but will also identify the best way to sequester carbon based on the areas where the device is placed.
Our device will be a "tree" that sticks half into the ground and half above ground. The in-ground portion will house a 7-in-1 soil sensor that measures moisture level, electrical conductivity, temperature, nitrogen, phosphorus, potassium, and pH. The above-ground portion will house a Raspberry Pi 4B single-board computer as the main processor of the system. The Raspberry Pi was chosen given its capabilities as a single-board computer given its 1.5Ghz 64-bit quad-core ARM CortextA72 processor. The SCD30 NDIR CO2 Temperature and Humidity sensor is utilized to measure ambient conditions. This sensor measures the CO2 composition of the ambient air in PPM (parts-per-million). It also provides temperature and humidity calculations as it contains a built-in SHT31 temperature and humidity sensor. It is interfaced with the Raspberry Pi via I2C connection. To add GPS capabilities to the Pi, it is equipped with an Adafruit Ultimate GPS breakout board. The device is also equipped with four cameras to capture images to conduct object detection. Overall, the prototype of the device is estimated to cost $372. However, the cost per unit can decrease significantly when utilizing wholesale parts upon mass production.
By utilizing the data from the device an algorithm will determine the best tree or plant to set in the given area with the ultimate goal of biosequestration. This term refers to the net removal of CO2 from the atmosphere by plants and microorganisms, as well as the storage of CO2 in vegetative biomass and soils. Through our team’s research, we have concluded that our device has the potential to be an extremely valuable tool in promoting soil health which in turn can lead to gigatons of carbon capture if used to scale. The factors which will be taken into account to determine this are: pH Value, soil temperature, moisture, nitrogen, phosphorus, potassium, electrical conductivity, hardiness zone, light exposure, soil preference, tolerances that may be presented, type of tree/plant, and size of the tree/plant. Tree and plant diversification will also be considered which is why the device includes four cameras and a GPS module. The cameras will use Google Cloud vision to identify trees/plants in the area when an Internet connection is available. It is very easy for a non-expert to deploy this device in a secure fashion as no knowledge apart from turning the device on is required.
To utilize artificial intelligence, the device uses a selection algorithm powered by the “Plant and Protect” database compiled by the Morton Arboretum. The Morton Arboretum is a local arboretum located in the Chicagoland area which has an initiative entitled Plant and Protect. The architecture of the selection algorithm entails using Python libraries such as Beautiful Soup, pandas, and Selenium. Using these libraries the algorithm loops through and scrapes plant data from the Morton Arboretum website and creates a CSV file. Pandas then obtains the data from the created lists and creates a dataframe database. The database has information such as name, location, soil preference, size range, light exposure, and tentative NPK test results. This database includes 639 entries containing both native and nonnative trees and plants. One might think that nonnative trees and plants would be detrimental to an ecosystem, but the word “nonnative” is often confused with the word “invasive”. Rather, nonnative trees and plants can be quite beneficial to an ecosystem. The following conclusion was reached by the Journal of Pollination Ecology in 2019: "We propose that using native and non-native plants improves habitat gardening by increasing opportunities for attracting a richer diversity of bee species and for longer periods." Next, the data is pulled to create a geolocation heatmap. The heatmap's intent is to show where the optimal locations are for the trees or plants.
When the device is not connected to the internet, it is able to suggest a suitable tree or plant to the user based on the local database, which is currently a total of 639 entries. However, this database can be expanded upon easily in the future. Since Google Lens relies on an internet connection, diversification will not be able to be considered immediately if an internet connection is not available. However, once a connection has been established, the results can be further refined using our more in-depth selection algorithm. Additionally, once a wifi connection is established, the device will send the more detailed findings to their master database, housed in Firebase and accessed off of our website. Firebase allows an encrypted data transmission along with overall security of findings per user or group. With proper authentication, the users are able to view their readings, live results, calculated heat maps, and much more.