Wine research: Drones remove guesswork on taint

VIGNERONS could have access to drone-powered technology to detect smoke taint in grapes this fire season.

Preliminary models will be presented at a conference in Argentina at the end of the month before being trialled in South Australia in December-January.

Developers are hoping to have something available for use in vineyards by March.

“The model will recognise and map which parts have been contaminated and which parts have not,” University of Melbourne researcher Sigfredo Fuentes said.

Smoke taint can be a serious issue for grape growers who have had smoke in their vineyards due to bushfires.

It affects the quality of wine, often forcing it to be discarded.

Dr Fuentes said he developed the model — the algorithms that can interpret information collected by sensors mounted on drones — because there was nothing like it available.

He said the model would help growers separate out parcels of grapes that might be more affected by smoke than others.

“The program assesses pixel by pixel and transforms the data you obtain into a meaningful parameter, for example, contamination or no contamination, or low, medium and high contamination,” he said.

“Growers will have information so they can say ‘I will harvest all that because it has not been contaminated and the rest I can mix’.”

Dr Fuentes said all the field work for the project had so far been done in China due to lower costs but trials would be conducted in partnership with the University of Adelaide to test precision in local conditions.

“We are getting 80-85 per cent accuracy but it (should be) more than enough,” he said.

Other research being conducted in China by Dr Fuentes includes generating plant water status indices in grapevines to predict yields using tension sensors and identifying high quality parcels of grapes within vineyards.

Originally published as Drones remove guesswork on taint