**Name of Project: HYGROPRO DEFARM DATA CRUNCH
**Proposal in one sentence: Work with John to analyze Nanue’s Farm controlled environment data and find ways to improve farm efficiency.
Description of the project and what problem is it solving:
The HYGROPRO DEFARM is a mission to get all controlled environment farms to share their data on the ocean marketplace. John has been working with Trevor at Nanue’s Farm who has agreed to share his farm data for analysis. The farm is a shipping container farm designed and distributed by Freight Farms out of Boston, MA. The goal of the data crunch is to experiment with different methods of analysis that can be useful to freight farms so that they are convinced there is a benefit to publishing their data on the marketplace.
The Freight Farm controlled environment data is collected on an agrotek controller and stored on Freight Farm’s servers, but can be downloaded by the Farmer. The data contains timeseries sensor and actuator information as well as power consumption and simple crop information depending on how meticulous of a record keeper the farmer is. My proposed starting point to look at is the derivative of the electrical conductivity in the water as conditions change.
Trevor is also allowing John to test his crop at NCSU for chemical composition. This will provide a baseline test to compare data for. The chemical composition test will provide mass content analysis of the crop compared to the nutrient regiment it received in the farm. Another angle for the data analysis is to try and maximize the nutrient absorption of the crops by leveraging pH absorption windows. Once a hypothesis is discovered, Trevor will try to mimic the conditions required and John will collect another sample for testing to confirm or deny the hypothesis.
Grant Deliverables:
- Visualization of Freight Farm Data
- Optimization Proposal for Freight Farmer
- Hypothesis to test with plant chemical composition information from NCSU
Spread the Love
John needs at least 1 other person with data analysis experience to tackle this challenge. The awarded funds for this proposal will go towards compensating the team that helps take the first steps into understanding the pain points of the Freight Farm. Anything learned from this experience will help John convince Freight Farm that publishing data to the Ocean Marketplace has a big upside for FF. Finally, the principles explored in this initial inquiry will serve as the foundation for future ML techniques.