AI Case Study
SunSelect enables greenhouse farmers to more accurately forecast their future harvests with machine learning
SunSelect leverages machine learning technology from Motorleaf to decrease harvest prediction errors. The software collects data from indoor growing conditions and identifies patterns in order to predict harvests size. SunSelect has managed to reduce its weekly prediction error by 50%.
"The yield prediction system is built upon crop data—such as variety of tomato, density, etc.—and environmental data—such as light levels, temperature, and C02. This is combined with visual data, captured by several cameras deployed in the greenhouse. Finally, Motorleaf's hardware adds additional data not normally collected by greenhouses, such as RGB light spectrum. With this information, Motorleaf created a unique machine learning algorithm for SunSelect. Based on an ever-increasing amount of data captured week-by-week, the prediction generated by the system improved each week.
Initial trials of the technology since October 2017 in a 70-acre California greenhouse cultivating tomatoes demonstrated its value to farming."
"SunSelect reduced its error in predicting weekly tomato yield by half, resulting in significant cost savings for the grower.
As a result of the improved predictability using Motorleaf’s technology, SunSelect has since abandoned manual yield predictions in favour of Motorleaf’s algorithms."
Motorleaf created a unique machine learning algorithm for SunSelect
Budgeting And Forecasting
"Predicting the amount of vegetables from a harvest is currently a time-consuming process. Agronomists count samples of vegetables, leaves and flowers in a small area and that sample then serves to estimate the expected yield of the entire grow operation. Often imprecise, farmers are unsure if they will produce enough vegetables to meet contract obligations or know how much labour they will need to package their produce.
If they produce too much, farmers try to sell their perishable goods quickly at rock-bottom prices. Like most large greenhouse vegetable growers, Victor Krahn and his team struggled to match their supply and demand dynamics to produce consistent, predictable crops for retailers."
"crop data—such as variety of tomato, density, etc.—and environmental data—such as light levels, temperature, and C02. This is combined with visual data, captured by several cameras deployed in the greenhouse. "