Artificial intelligence (AI) tools are emerging to assist farmers in predicting optimal harvest dates for fruit. These models aim to consider various factors, including weather conditions, which can impact the safety of workers and the timing of ripeness.
Okanagan Specialty Fruits, an apple orchard company in Washington State, is experimenting with cameras mounted on tractors. These cameras, from Canadian firm Vivid Machines, use AI to identify buds, flowers, or fruit and provide crop estimates and harvest dates. Joel Carter from Okanagan Specialty Fruits noted that the system is effective at detecting tiny flower buds.
In the UK, FruitCast offers harvest forecasts for growers of strawberries, raspberries, blackberries, blueberries, and tomatoes. Raymond Martin, co-founder of FruitCast, stated that their system can perform ripeness analysis on a scale that individual farmers might not achieve across large or varied growing areas. The company claims its forecasts are within 10% of actual picked volume one week out and within 17% three weeks out.
Researchers are also exploring advanced methods for fruit analysis, such as using millimetre waves to penetrate deeper into fruit to measure humidity, water, and sugar content. However, the adoption of new technology can be complex, with growers needing confidence in its effectiveness, according to Jing Zhang at North Carolina State University.