This project is supported by funding from the Australian Government Department of Agriculture and Water Resources as part of its Rural R&D for Profit program .
Modern sensors backed by ever-increasing computing power have great potential in the vineyard – so much potential, in fact, that it’s almost hard to know where to start.
Two complementary projects under way at the CSIRO – supported with funding from the Australian Government Department of Agriculture and Water Resources as part of its Rural R&D for Profit program and Wine Australia – are designed to get a better handle on what’s possible, what’s feasible in terms of time, cost and effort, and how we can best match the technical side with the reality of what grapegrowers actually want.
The first, run by Dr Everard Edwards, is focusing on vine canopy, including size, composition and management. The second, run by Dr Mark Thomas, is interested in issues around the fruit, including composition, disease and yield estimation.
The common theme is access to expertise from different areas of CSIRO – primarily Data 61, Australia’s largest data innovation group, which is supporting the Australian Government's Cyber Security Strategy, but with input also from the Manufacturing and Energy business units.
‘We are very much looking at technologies in which the CSIRO already has expertise and how it can be adapted rather than trying to build a capability base from nothing or emulating work others have been doing’, Dr Edwards said.
There will be external collaborations, however, notably with Germany’s Fraunhofer Institute, which is providing access to its machine learning technologies for analysis of hyperspectral data and aerial hyperspectral hardware. This collects images of a scene with over 400 narrow spectral bands for each pixel, rather than the three that are in a normal colour image.
But the focus is not all on the high end of high-tech. The researchers will look at what can be achieved by working differently with, for example, conventional cameras or sensors such as LiDAR, which are beginning to find a place in the wine sector. Dr Edwards is working with wineries in the Coonawarra to see if LiDAR can be used to measure the porosity of the canopy in conjunction with its size to assist in canopy thinning activities.
Before things got started the CSIRO team held a sector workshop to ask where people thought digital tools would be most useful and what their most pressing needs were. Top of the list was yield estimation, closely followed by canopy management and condition and fruit composition.
Dr Edwards wasn’t surprised. ‘One of the things that’s interesting with the wine sector is the emphasis on the composition of the fruit’, he said. ‘There’s no other agricultural industry where the composition drives such a massive diversity in the value of the crop.’
The research will be wide ranging. There are things growers do now that they’d like to be able to do better, more efficiently or on a bigger scale, and the researchers have a fair idea about how sensors could be applied. But at another level there are things they can potentially pick up with the sensors that they can then apply in completely new ways that hadn’t been thought of before.
The bottom line (or at least one of them) is practicality. There is no point in developing a piece of equipment, no matter how well it works, if the purchase cost is higher than the value to the purchaser. Sensors worth a few hundred dollars you can strap onto a tractor have much more immediate appeal.
At the same time, it’s important to know what is possible down the track. ‘The cost of everything to do with sensors and computing is coming down quickly, so in a few years what was once $100,000 may only be
$20,000’, Dr Edwards said.
New and developing technologies have the potential to monitor all aspects of vineyard activity.
A future vineyard may possess automated imaging capabilities that generate a 3D model of the vine canopy, highlighting the differences from the desired structure and how to improve fruit composition through canopy management. The same imaging may provide whole of vineyard data on vine nutrition or early warning of mildew infection, allowing proactive management on a rapid timescale.