Digital tools for canopy management to increase competitiveness

Summary

Objective

To develop and implement new features for the VitiCanopy smartphone application to allow higher processing capabilities, increase flexibility and generate data for different grapevine densities and training systems.

Background

A previous Wine Australia project (UA 1207) made significant progress towards providing tools for vineyard management:

• Development of a free app called VitiCanopy, which characterises canopy architecture with geotagged data across a vineyard.
• Re-evaluation of vine balance using VitiCanopy. New indices such as Leaf area index/Yield and Canopy Porosity/Yield were correlated with fruit quality – through chemical composition and fruit grade at harvest.
• Accurate preliminary models to assess pruning weight at the plant level using parameters obtained from VitiCanopy as inputs and real pruning weights per plant as targets, through Artificial Neural Network modelling techniques.

Research approach

The VitiCanopy App will be further developed with new modules added to increase flexibility, generate accurate data for different training systems and vine densities and assess pruning weights. An important aim is to provide information about the links between canopy architecture, vine balance and berry/wine non-volatile and volatile chemical composition, sensory attributes and quality for different varieties, training systems and wine regions using machine learning methods.

A web-based software tool (VitiWeb) will be developed to allow batch analysis of multiple images taken in the field and the option of Geographic Information Systems (GIS) generated maps of vigour and other architectural parameters, allowing users to assess spatial differences in canopies across a vineyard. VitiWeb will also be able to analyse canopy architecture using remote sensing imagery from users, based either on Unmanned Aerial Systems platforms or satellite information.

The link between bud fruitfulness and canopy management will also be investigated to achieve more accurate early yield predictions, through bud dissection, image analysis and machine learning.

The new tools will be validated and delivered to key stakeholders via a benchmarking study across multiple wine regions and industry partners, using vineyards that produce grapes of different grades/quality. The study will provide objective region and variety-specific information on the best canopy structure to achieve the desired fruit and wine quality.

Sector benefits

This project will deliver refined vineyard management practices and tools that will assist the sector to achieve consistent production of high quality grapes and wine.