MUD — Moisture Under Detection

Co-authors: Peter Phillman Pedersen, Céline Portenier

As part of the HackEO Zurich 2024 "Pixels for the planet" we (Peter, Celine, and I) have build a mock-up of a product and designed a technology that explains the need for more accurate moisture detection in the soil. We believe that by constructing a synthetic dataset of the soil moisture we can train a machine learning model to predict the moisture content changes of the soil with accuracy better than 0.05 \(\frac{\text{m}^3}{\text{m}^3}\) using publically available satellite data. This would allow to see the rainfall, and derive yield index for backcasting the yield of the crops.

External resource: https://itsmud.com