Digital twins of orchards created to boost food production

University of Queensland scientists are creating computerised models known as ‘digital twins’ of mango and macadamia orchards to help boost food production. 

“It’s an example of how computers are changing the industry,” said Professor Neena Mitter, Director of the Centre of Horticultural Science at the Queensland Alliance for Agriculture and Food Innovation (QAAFI).

Digital twins provide a real-time virtual representation of a physical object — like a orchard — that can be studied digitally. Professor Mitter said digital technologies could help accelerate innovation to make food production more productive, resilient and sustainable.

“Developing a digital model for an orchard with slow-growing crops like mango and macadamia enables us to run virtual experiments at a scale and speed never before possible,” Professor Mitter said.

Lead researcher Dr Liqi Han said the technology would particularly benefit slow-growing crops like fruit trees.

“The digital modelling provides untapped opportunities for users to rapidly trial new ideas and acquire a reliable indicator of how to best optimise production systems,” Dr Han said. “We call this technology ‘DigiHort’, short for Digital Horticulture.”

The computer simulations can be a conceptual design of an orchard that doesn’t yet exist, a digital twin or detailed replica of an existing orchard, or a digital variant, where changes are made to a digital twin.

“All three forms can be integrated with environmental and management simulators,” Dr Han said. “For example, this might include sunlight and chemical spray simulations to allow for evaluation and optimisation of orchard management practice.”

Virtual trials start with the design, with software users able to decide where in a landscape to plant trees, the density of the canopies and the configuration of the rows.

Users then consider how the trees are maintained, wielding virtual pruners and testing the impact of different — and even unconventional — tree-training systems. This innovation is based on new LiDAR scanning technology applications undertaken with industry partner Riegl Australia and state government research stations in Queensland, Western Australia and the Northern Territory.

Dr Liqi Han with the terrestrial LiDAR scanner in the field.

It relies on high-performance computing (HPC), which allows Dr Han to run extremely fast virtual experiments without loss of accuracy.

“These days, we talk more and more about precision agriculture,” Dr Han said.

“We enhance precision by looking at the details, such as how much light can be captured by each leaf or fruit, or the distribution of sprayed chemicals across the canopy. We can accumulate small benefits into big benefits or prevent big losses from occurring. And we’ve found that small differences can have a big impact.”

The DigiHort platform was designed as a decision-support service for industry and will be accessible via the internet.

Top image by Dr Liqi Han. Image credit: University of Queensland.