John Mawson and his colleagues are building an orchard – a digital apple orchard – which will bear fruit of a very different kind.
It’s part of the work New Zealand scientists from Plant & Food Research are pushing ahead with to help solve one of the planet’s most pressing problems: how the world will feed itself by 2050 with an estimated population of nine billion, faced with issues like climate change and the major impacts it brings.
Mawson is leader of the Plant & Food Research Growing Futures Direction Digital Horticultural Systems. If that’s a bit of a mouthful, so is the task he and his team are biting off.
They are using a technique known as “digital twinning” to replicate a real orchard. Through a mix of tree crop modelling, data science, new sensing technologies and artificial intelligence, and good old-fashioned hard work, they are creating a virtual orchard growing on the back of an immense amount of data gathered from the real world and a real orchard.
It’s groundbreaking work. A few other organisations are developing digital twins for other specific horticultural crops, but no-one is tackling this challenge for perennial fruit crops with the same wide view as Plant & Food Research’s programme, which is looking at all steps in the journey of fruit from orchard to consumer.
“Digital twinning simply means we are making a digital construct of a physical system – in this case an orchard,” Mawson says. “We build mathematical models which will predict the future behaviour of that physical system in response to scenarios that affect it, for example, a change in management system, an extreme weather event, or climate change.
“Those models are constantly updated according to changes occurring in the real world, in that physical orchard, and the model predictions are then fed back to actively manage the orchard to optimise its performance.”
It’s a painstaking task. Thousands upon thousands of images and other data from other sensors are uploaded to the digital twin, closely examining all the factors that affect fruit production. The eventual aim is to reliably predict which crops and growing systems are likely to perform best in the future.
“Ultimately, we will devise better ways of producing crops of better quality, with higher productivity – and producing fruit or crops more resilient to change.”
The data is not just gathered from the trees. It also encompasses factors like weather, orchard management, pests, diseases and unexpected events of the kind that have been plaguing New Zealand recently. It extends across the whole supply chain and tracks what happens to the fruit when it is picked, graded, stored and distributed and any potential issues that can arise in taking a piece of fruit from orchard to consumer.
Mawson compares the constant updating and assimilation of changes into a digital environment using AI to the sophisticated digital data emanating from F1 cars: “The F1 cars have as many as 200 sensors measuring everything from temperatures to air flows to tyre pressures – all those things which affect the car’s performance that can be tweaked to get better results. We’re doing something very similar – but seek to predict orchard performance for both the current season and multiple years, so we can make predictions of what might happen over generations.”
The need for research like this is undeniable; globally, the rising population and other major issues like climate change and food security bring social, environmental and political upset – and the pressure on food systems will only worsen unless science can find ways to lessen the impact.
“Plant & Food Research’s mission is to develop more sustainable food production to the environmental and social benefit of New Zealand,” Mawson says, “and digital twinning is about helping our decision-making to be smarter and more timely.”
Understanding the power of new technologies, such as AI, is part of Plant & Food Research’s remit. They’re also looking at other new advances in lab- and field-based science that could support food production like open ocean aquaculture, lab-based meat production, vertical farming, gene editing and enhanced processing techniques so that the by-products of food production, like fish skin for example, can be optimised into products other than food to avoid wastage.
Plant & Food Research Chief Scientist Richard Newcomb says New Zealand already feeds 50 million people from a country of only 5 million: “There are more opportunities for us and we could have a greater role in feeding people, not just directly with the food we produce but also with technologies and food production systems; we could help other countries so they could produce more food as well.
“Technology is moving faster than ever before, and one vital role Plant & Food Research can play is understanding how these could be used to benefit Aotearoa and keep all the options on the table. We see gene technologies, for example, as a potential game-changer for food systems; it has a huge range of applications – especially now we have gene editing technology that is much more precise than the version 1.0 genetic modification technology.
“It could do all manner of things in terms of making our crops more sustainable, reducing the need to fertilise them, making them more climate change resilient, increasing nutrition and dealing with pests and pathogen problems that we have coming at us.
“Understanding the science is only the first step for many of these advancements and in some ways the most black and white. Applying them in a way that makes sense for the nation is a much more complex discussion.”
Another example under investigation is lab-grown foods. Consumers in Singapore can now buy lab-cultivated chicken products, with the USA following after recent regulatory approvals earlier this year. Plant & Food Research is looking at how this technology can be applied to Aotearoa’s fish species, for use as a research tool and as a way of diversifying seafood production.
“In some cases, the research is a science tool first but there is often a practical application that follows in due course,” says Newcomb. “We recognise the future will be far more volatile so we are trying to be smarter in terms of planning to combat that volatility – and we are working hard to reduce what we call ‘time to value’ – when the science produces something practical the food sector can use.”
For more information: plantandfood.co.nz