Technology has long been able to provide insights into what is happening at very specific locations on farms, using moisture probes or water rain sensors and a range of other devices. It’s also long been able to shed light on high level trends, using tools such as satellites. CSIRO’s Dr David Henry says it is the space in the middle where artificial intelligence (AI) will have a huge role to play in agriculture, using sophisticated modelling to integrate data to deliver locally-specific insights, at scale.
“Technology like weather stations is very specific to its exact location, while satellite data is often used to identify broad trends beyond paddock-level detail,” says Henry, a Principal Research Scientist and Research Leader of the Digital Agriculture initiative within CSIRO’s Agriculture & Food Business Unit.
“We can’t put a sensor everywhere, and we can’t get very granular data from satellites, but AI can help integrate data stacks to give us insights that are at scale, but locally specific, and most importantly, actionable at the right scale for the decision,” says Henry.
“A soil probe can send data from a particular location many times a day. A satellite might pass over that same location every five or more days. Artificial intelligence can look for relationships between what the sensor is showing and what’s embedded in the satellite data, and use that insight to reach reliable conclusions about what is happening in other locations and even into the future.”
Henry believes developing robust, locally-specific insights at scale will deliver positive outcomes that go beyond the farmgate and across the entire agri supply chain, amplifying the potential of technology at both ends of the spectrum.
“There is opportunity for co-benefits from AI in agriculture. It will allow us to increase productivity at the farm level, improve sustainability across the sector, and provide consumer confidence about our methods of production. But first we have to work out how to best integrate farm level data with regional or global data to deliver scalable solutions.”
Harnessing the power of AI for improved forecasting
The other key benefit AI may deliver is the ability to make predictions, Henry says.
“Data is interesting but a probe or sensor only tells us what is happening right now. AI may enable us to use that data to work out what is likely to happen next week or in 30 years. This will be of enormous benefit in Australia, where climate variability and climate change is one of our greatest challenges,” says Henry “Being able to mine and analyse data and preview future states will help us see where yields might drop over time or understand what our water storage might look like, and those things can help us plan. AI can also give us predictions about an entire farming region, which will be useful for policy makers, investors and lenders.”
While AI is able to derive insights from complex and disparate data of all sorts of scales, Henry cautions that having domain expertise is crucial to ensure the insights are meaningful and useable. CSIRO is involved in a number of programs, including a trial of a digital twin platform at the CSIRO research farm at Boorowa, near Canberra.
The project is a collaboration between CSIRO, Microsoft and Australian start-up Agronomeye. The new platform takes real-time data and puts it into the hands of farmers in a way that is easy to use and simple to interpret – so they can confidently make critical decisions, such as matching their sowing timetable to soil moisture profiles or planning water catchment systems based on natural flow patterns.
The digital twin is being used at CSIRO's Boorowa Agricultural Research Station to provide insights into soil condition, crop growth and farm management.
How AI will impact the agriculture sector over the next five years
Henry sees three key areas where AI development will be focused:
- Rapid development of shared data foundations and system interoperability to enable data owners to share their data securely and privately in ways that enable additional value to be derived;
- Predictive insights will start to take over from ‘right now’ insights. AI will tell us what things will look like three months down the track, or next week, or in 2050;
- Automation will become more widely available, with AI enabling the adaptation of automation tech to our diverse and uncertain environment.
What CommBank AI experts are seeing
AI Labs is one of many AI teams in CommBank. They are primarily focused on cutting edge AI applications, and how that can be used to benefit our customers.
Executive Manager Dr Luiz Pizzato says AI has the potential to create opportunity in every aspect of agriculture – a combined multimodal AI technology which includes image, video and sensor information can bring the next sweep of changes.
“Already we see sensors in all kinds of on-farm applications, from soil monitoring to animal health checks and milk quality monitoring. AI can monitor remote produce storage sheds and ensure optimal storage conditions are met for long-term storage and it can enable drones to conduct inspections over huge properties.
“I expect to see a lot more use of visual technology, which is already in use in other industries. For example, robots are picking fruits and ensuring its quality, and being able to learn to refine choices over time. This kind of capability will be extremely valuable in an environment of extremely tight labour supply.”
Dr David Henry is team leader of the Climate Adaptation and Thresholds team at CSIRO, a Principal Research Scientist based in Melbourne, and is a Research Leader in Digital Agriculture. His focus is on the development and application of precision and remote digital technologies, and how to capture value from new technologies, data and information in industry. His current work is both national and overseas and some may recognise his name as creator of “Pastures from Space” and early champion of virtual fencing. He has worked extensively with start-ups and at the public/private co-benefits interface, particularly focussed on bringing domain expertise together with cutting edge data science to drive impact. You can follow his research on ResearchGate and LinkedIn.
Dr Luiz Pizzato is Executive Manager, AI Labs at CommBank. AI Labs is a team that focuses on improving the financial wellbeing of the bank’s clients and society while bringing innovations in AI/ML. Some of the work led by his team involved the detection of financial abuse and weather data modelling. Luiz is a Data Scientist and Entrepreneur with more than 20 years’ experience in artificial intelligence. He has a PhD and a MSc in Computer Science both with focus on natural language processing and information retrieval. He also has expertise in recommender system, machine learning and data mining.
Luiz is an academic at heart with a strong commercially-focused mind. Before joining CommBank, he consulted with clients across a wide range of industries through his private consulting practice and at Accenture where he led an expert team of data scientists and full stack engineers at the Accenture Liquid Studio in solving very hard problems.