Digital twins for mitigation of orchard spray drift

Abstract

The annual sale of agricultural pesticides in Australia doubled from 2010 to 11 to 2021-22. While an analysis found that 73% of Australian crop production in 2015-16 relied on herbicides, fungicides and insecticides, a large proportion of pesticide applications miss the right targets and become “spray drift”. For horticulture, according to previous studies, about 50 to 70% of the applied chemicals can either land directly on the ground or drift into the air. The off-target spray brings no gain to growers but instead is unnecessary waste, unnecessary pollution, and unnecessary risks in potentially damaging neighbouring farms. Many growers hope to maximise the use of their existing sprayers with minimal cost before considering retrofitting or replacing them. To facilitate this objective, a digital-twin-based solution has been developed to simulate orchard characteristics and spray processes that closely mirror reality. It enables high simulation fidelity and aims to serve as a decision-support tool for practical operations. This approach is powered by high performance computing to ensure both accuracy and complexity, without compromising one for the other. It had been field-tested in a tropical research orchard with a strong correlation to real-world measurement. A subsequent digital-twin-based investigation has demonstrated that the spray drift in the same orchard environment can be reduced by 36% through simple modifications of sprayer settings. Such outputs cannot only be used to optimise the adjustment of traditional equipment, but can also be loaded to robotic/automatic sprayers to make them more precise and responsive, yet simpler and cheaper. In addition to reviewing the pain points in reduction of spray drift, introducing the digital twin solution and reporting insights that have been gained from it, the research investigates potential development pathways for similar digital twins in a broader range of orchard and vineyard systems.