Pre-emptive satellite prediction of food waste
Use satellite imagery to predict, at the farm or local geography level (e.g. <20mile radius level), harvest time and yield of produce. Use this information to predict when and where too much food is going to be produced and likely result in food waste, and set up operations to pick up and utilise this food. Unknowns: - Where exactly food waste happens close to the farm level. Food waste closer to the farm is more prevalent in the global south, in these geographies what do farmers currently do with excess produce? Are there already local solutions that solve this problem e.g. do farmers text friends to sell of their goods cheap, do they successfully sell them at market just for lower cost, or do they fail to sell them at market? Only in this last scenario does this project, in its current wording, add value. Known issues: - Getting traction with tech use for traditional (or older) farmers - Setting up infrastructure on the ground. How do you actually pick up the excess produce? Transport it? Maybe a too good to go style app that alerts entrepreneurial local businesses to buy and sell cheap produce could work around this problem and mean no physical assets are required. Existing companies: - Apollo Agriculture, don't do quite the same thing but tangentially related. Use satellite and other data to predict which crops would grow best on different pieces of land (thanks AG for the info on them!).