Feasibility study for Application of AI in Protected Cropping

A series of hypotheses for applying advanced Artificial Intelligence (AI) and other technologies to protected cropping were evaluated for technical and commercial feasibility. It was concluded that the most appropriate course of action was to use state of the art technologies to develop a common strategy and concerted framework for the collection of high quality data to enable future efforts to develop AI solutions.

Summary points:

1. We recommend building a consortium to collect high-quality data that can enable advanced analysis for the purposes of crop protection, yield prediction, improved harvesting and SmartHort.
2. Building a value-chain wide consortium would enable an optimised solution that reduces waste and costs, benefiting growers.
3. Improved scouting with automated systems already in development is possible, and will support the project suggested in point 1.
4. Improved crop-walking solutions would bring short-term efficiency gains and provide better data for future analysis, and will support the consortium suggested in point 1.
5. Powdery mildew prediction was one of the goals set for this project. Despite our extensive efforts, it was not possible to get access to a suitable data set from growers to enable this analysis. Accurate prediction would be one of the outcomes of the project in point 1.
6. Pest prediction is similar to powdery mildew prediction, but given the development of connected electronic insect traps, it may be possible to address this more quickly. Accurate prediction would be one of the outcomes of the project in point 1.
7. Automated sensor drift detection was initially identified as a potential direction for the project. However, we believe more promising directions were identified.
8. Data governance principles must be established to protect growers' interests while enabling the collection and use of data suitable to develop AI applications beneficial for all.

The purpose of this study was to identify areas where AI could be deployed or improved to accelerate the performance and growth of the protected cropping industry. A thorough assessment as conducted by consulting experts in the field and conducting a literature review in order to identify how AI could be developed to positively impact the industry have been formalised and assessed for technical and commercial feasibility.