Three ways AI is transforming agriculture and food
The University of Melbourne is charting the way as a world pioneer in the application of artificial intelligence (AI) in food science and agriculture. Significant developments are being enabled by drones, remote sensing and machine learning, helping us to save wine from bushfire smoke, measuring our unconscious responses to food and creating smart farms and vineyards.
Associate Professor Sigfredo Fuentes is group leader for Digital Agriculture, Food and Wine, at the School of Agriculture and Food.
Associate Professor Fuentes says, “The main problem producers have at the moment is how to assess the impacts of climate change. Artificial intelligence including machine learning, robotics, drone technology and sensor networks can do a really good job collecting large amounts of complex data within the field. These technologies can even go automatically through a plant-by-plant assessment, so there is a really big revolution on steroids in the next couple of years with the development of 5G technology.”
For farmers, more accessible and reliable data gives them an advantage. As more data is entered into systems applying machine learning, they become increasingly accurate, helping growers to assess how healthy crops are, how to save and retarget water, and pinpoint nutrient levels.
AI application #1 Saving wine from bushfires
Bushfires in wine-producing areas such as south Australia, California, Europe, central Chile, Greece and South Africa have the potential to ruin a whole vintage.
If grapevines are exposed to smoke and become contaminated, tools like infrared thermal image, near infrared spectroscopy analysis and machine learning modelling developed by Associate Professor Fuentes’ group could be used to spot smoke damage with unprecedented speed and precision, allowing growers to separate out damaged fruit. He says, “If grape growers and winemakers could assess in almost real-time which grapevine plants are smoke-contaminated and the level of smoke-taint present in grapes, as well as the potential wine made from them, it would be a game-changer for the industry.”
The potential of drone technology is impressive, with the ability to survey 2000 hectares in 1.5–2 hours, scanning at a plant-to-plant level and producing maps overnight detailing the bushfire contamination of grapes.
#2 Better cattle welfare in heatwaves
In our changing climate, heat stress is becoming an increasing problem for animal welfare, productivity and quality. Our robotic dairy at the Dookie campus collects data from each cow. Health and productivity measurements are taken when a cow is milked, and the data is used to develop a machine learning model that uses feed and weather information to predict milk outputs related to volume and quality. Furthermore, a recent study on sheep at Dookie was able to produce AI models to assess heat stress based on biometrics, such as heart rate, skin temperature and respiration rate of animals from an integrated visible and infrared thermal camera automatically. This application open the doors to monitor live animal in transport for exporting and also heat stress of animals in the farm.
#3 Assessing our tastes using data and AI
Beyond the farm and vineyard, technology such as remote sensing techniques are being used to measure how we respond to new food. In the search to create produce for increasingly demanding consumers, these techniques assess physiological and emotional changes: face recognition, emotional response, heart rate and respiratory rate.
Associate Professor Fuentes says, “Those physiological parameters are related to the autonomic nervous system, which is our automatic response to food and beverages. So, when you drink something, or you taste something, the unconscious response is around 80 per cent of the information that related to preference and liking. With biometrics, now we are assessing what you like and how much you like a certain product. We can do all that, while you are tasting or drinking something, without asking you any questions.”
Young farmers are now starting a tech revolution in agriculture. Barriers to the uptake of technology are disappearing as a new generation of trained graduates travel from the city to the farm to apply new and emerging technologies.