As a grapegrower, whether on this hemisphere or the other, life is getting increasingly difficult. A changing climate, increased input costs, the ongoing trade nightmare between Australia and China and labour shortages due to COVID-19 travel limitations are all putting growers in the hot seat.
To account for climate change in the vineyard – which is ultimately where the viability of wine as an industry lies – new technologies are being developed and implemented in order to keep growers pushing through the quagmire.
Wine Australia, in its report titled, ‘Driving Adoption of Agrifood Technology in the Australian Wine Industry’, has identified the top four industry challenges growers currently face.
If accurate yield forecasting estimates are incorrect, for example, growers could incur significant logistical and commercial run-on challenges for their winemaker clients.
Industry research shows yield estimates can vary by 30% on average. New technology from a UK-based remote sensing and machine learning specialist agtech supplier, Deep Planet, aims to solve this challenge for growers the world over.
Over the last three growing seasons, Deep Planet has worked with small and large Australian growers to develop and commercialise its Vinesignal product, which provides a full suite of vineyard health, soil moisture and hands-off yield monitoring and prediction services.
Co-founder David Carter said, “Our world leading machine learning and GIS expert team have been working diligently”.
“We have support from the European Space Agency and plant scientists and biologists developing our technology. We’ve ensured our technology can solve industry challenges in the most practical and cost-effective way.
“Launching commercially for the 2021 vintage in Australia was perfect timing for us. On a personal note, coming from the Hunter region, it’s exciting to be rolling this out in Australia first.”
Deep Planet is confident of providing hands-off yield estimates with average accuracy over 90% for Australia’s top varietals.
Extreme unforeseen weather events aside, the accuracy increases from budburst through to two to four weeks before harvest, according to the agtech company.
“There is no need for any infield sensors or infrastructure, costly aircraft fly overs or drones. We use proprietary identified satellite bandwidths, historical data, and agronomy models to provide regular, accurate forecasts to our clients,” added Carter.
“This saves them time and money. They can better manage relationships with their customers and maximise commercial returns.”
There are several barriers to agtech adoption that suppliers like Deep Planet need to overcome. Top of that agenda is confidence in the technology and showing its value proposition.
To address this, Deep Planet is inviting growers and wineries to join its Yield Challenge initiative.
“We know that for the industry to adopt new agtech, suppliers must show it works,” said Carter.
“We’ve been working with several regional grape and wine bodies and speaking with clients and growers across Australia.
“These discussions have resulted in this unique opportunity for growers to take part in our Australian Yield Challenge. Growers submit their historical yield data, and we compare that with our machine learning predicted results.
“The 2020 vintage was difficult. We’ve seen up to 500% variation between grower predictions compared with the Vinesignal AI model prediction and subsequent actuals.
“This is an opportunity for the Australian grape and wine sector to benefit from being early adopters of the technology. We’re keen to partner with growers to co-create and guide the development of extra features to implement in the future,” continued Carter.
“Australia’s wine sector has an enviable reputation for innovation. This is evidenced by the increasing uptake of our technology by small, medium and large corporate growers and groups. These early adopters will continue to benefit as we add extra functionality to our technology.”
Are you a Daily Wine News subscriber? If not, click here to join our mailing list. It’s free!