AI models predict optimal harvest dates for fruit growers, aiding planning and productivity

AI for Agricultural Planning

Okanagan Specialty Fruits, an apple grower with over 1,250 acres in Washington State, is implementing artificial intelligence (AI) models to forecast ideal harvest dates. The company aims to maximize productivity and manage challenges such as adverse weather conditions during harvest periods.

Joel Carter from Okanagan Specialty Fruits noted that traditional harvest planning can be disrupted by factors like extreme heat, which can halt operations. AI models that integrate weather data and predict ripeness are seen as beneficial for determining not just when fruit is ripe, but also the available picking window.

Technology in Use

The company is experimenting with camera systems from Canadian firm Vivid Machines. These cameras are mounted on tractors and capture imagery of apple trees as they pass through orchards. AI then analyzes this footage to identify buds, flowers, and fruit.

Carter stated that the Vivid Machines system provides crop estimates and harvest dates. It is particularly effective at detecting small flower buds that are difficult to see with the naked eye, offering early insights into potential yields.

Data Dependency and Accuracy

The accuracy of these AI-driven forecasts relies heavily on the quality of historical data fed into the system. Carter emphasized that the AI models require farm-specific information rather than general internet data, making the solution bespoke to each farm's conditions.

While apples, such as Granny Smiths, have a harvest window of about three weeks, other fruits like berries have much shorter windows, sometimes only a few days. Precise timing is critical for these crops to avoid spoilage and maximize profit.