Saturday, December 2, 2023

Tech aims to boost grape yield forecasting

Avatar photo
Robotics comes to the aid of what can be a costly problem for viticulture.
University of Canterbury Computer Science and Software Engineering academic Professor Richard Green, research engineer Dr Josh McCulloch and research assistant Matt Lansley are ‘predicting the unseen’ with a new method for accurate yield estimation in viticulture/horticulture.
Reading Time: 2 minutes

Blending viticulture with robotics, 3D imaging, and data analytics could assist with the potentially costly exercise of forecasting grape yield.

University of Canterbury researchers have been awarded  $6.1 million over five years by the Ministry of Business, Innovation and Employment’s Endeavour Fund to investigate a new approach to yield prediction.

The wine industry is one of New Zealand’s most important and valuable horticulture sectors, adding up to $2.4 billion in exports this year. 

However, forecasting grape yield is a struggle and inaccurate techniques can be costly, affecting profits.

“We can’t automate what we can’t see in agriculture because every plant is different, but this research could change that,” University of Canterbury Computer Science and Software Engineering academic Professor Richard Green said.

For his research project, Predicting the Unseen: A new method for accurate yield estimation in viticulture/horticulture, a team will develop a new approach by blending a 3D-imaging-based detection system with a physiological growth prediction model.

Green said it is a complex, interlinked, and challenging measurement and data problem – and this is the first time it has been approached this way. 

“We’re encouraged by the results of our preliminary research. It’s already clear that there is huge potential with this new approach, and I believe this funding will help us go further than people may think,” he said. 

“Until now we’ve only been able to scan 3D images of vines that have no leaves, so we know where to prune. Now, with this funding, we can scan through the year, which means you can perfectly align it to see how much it’s grown and changed. 

“This will help us forecast yield, and we’ll gain access to data that will help us understand the crop on a whole new level.” 

Working with the right people is important, Green said. The research team includes cross-disciplinary, multi-institutional members, including mātauranga Māori, robotics, data analytics and viticulture experts covering all technical aspects of the research to ensure the technology accurately serves the needs of the industry. 

Green said the project may go further than forecasting grape yield, and New Zealand has the capability to become a world leader in using these systems for multiple benefits. 

“Not only will we be able to predict it, we’ll increase average yield, and improve operational and financial planning for wineries and vineyards. We’ll also accelerate vineyard automation to help to mitigate labour shortages and costs, and we will better prepare our vineyards for climate change.” 

If all goes well solving this expensive problem for the viticulture sector, the flow-on effect may create new revenue and export opportunities.

Next, the research team will aim to take on other horticulture challenges. 

“I see true potential that as we solve these issues, this project will serve as a template for future research and become a cornerstone for more extensively automating our future agriculture high-tech sector,”  Green said.

People are also reading