Agricultural technology in Australia

A perspective from Tatura SmartFarm. By Anna Mouton.

Agriculture Victoria’s world-renowned Tatura SmartFarm is synonymous with horticultural innovation. Dr Mark O’Connell is a senior research scientist in horticultural production science at Tatura. He has more than three decades of experience in agriculture and currently leads several projects on orchard systems and precision management for stone fruit.

In December 2023, O’Connell organised and chaired the International Society for Horticultural Science’s International Symposium on Precision Management of Orchards and Vineyards, held at Tatura. He spoke to Fresh Quarterly about his research and the potential impact of new technologies.

Q. How are you using technology in your research?

A. It helps us to be more efficient in doing our measurements. We’ve changed the way we collect data. We weighed and counted things in the paddock and used pencil and paper. Nowadays, we’re using scanning and remote-sensing systems. Smart technologies allow us to collect more fruit and tree data.

One thing we do is mark trees or plots with NFC tags. When capturing data, whether measuring fruit size with digital Bluetooth callipers or measuring fruit colour with custom-made sensors, all that is linked to a tree or a sample by scanning the NFC tag.

We use a smartphone app as a data logger. The app records the data using Bluetooth technology and links it to the NFC tag.

Q. What are some applications of computer vision systems?

A. We’re working with an Australian company, Green Atlas. They have developed a platform called Cartographer that drives up and down the rows and scans the trees and fruit. Using machine learning systems, it can measure whatever you calibrate it for.

We’ve done a lot of work around apples, pears, and stone fruit, including counting buds or flowers, detecting fruit size, colour, and number, and detecting tree size and canopy architecture.

The machine takes four or five high-resolution scans or photographs per second. It’s fast — some people drive it at 15–20 kilometres per hour. The data is uploaded to the cloud, and within a few hours, you get a digital map of the orchard based on whatever parameter you selected.

We’re demonstrating it to the industry but also using it to measure our plots and experiments, especially to estimate yield and fruit quality—size and colour—near harvest.

A corporate almond grower in northwest Victoria is tackling the big problem of mummies left on trees after harvest. The unharvested nuts cause a disease risk. They’re using the Cartographer to scan orchards for mummies so they can send their troops out to those hot spots and manually remove the mummies.

People have also used it for weed detection. And it can come up with digital maps for variable-rate fertilisation, precision orchard spraying, or crop-load management customised for equipment such as the Darwin flower thinner, variable-rate sprayers, and fertiliser machinery.

Q. Are you investigating drone technology?

A. I’ve just started a precision-ag project with the summer-fruit [stone-fruit] industry and Hort Innovation [Australia’s equivalent of Hortgro Science but for the entire horticulture industry].

I plan to use a drone to scan fruit size and number and tree size. It’s been tried in apples around the world but never in stone fruit. Apples are probably an easier crop — you have big red apples, not much foliage, and wide row spacing — but we will test it on peaches, plums, nectarines, and apricots.

Variability in fruit number per tree is a huge factor that influences fruit size and marketable yield, so if we can manage that better and understand where the hot and cold spots in terms of fruit numbers are, we can potentially intervene within the season or know what to do better next year.

It links back to pruning, thinning, and all those sorts of things. We need a precision orchard map for variable management, and any way we can get that map other than manually measuring is a huge benefit.

Q. Which technologies could improve harvesting?

A. We’ve gone down the path of a new research project on narrow orchard systems, so a 2-metre tree row and a 30-cm narrow canopy with cordons and vertical leaders. That’s changing what machinery you can get down that row.

We’re aiming for high-production, potentially pedestrian orchards, so 2-metre-high trees, which remove the need for picking and pruning platforms. We’re getting ready for mechanisation and automation by developing a fruiting wall.

But we also have an Italian REVO picking platform, which we’ve started to instrument. Up to six people on this machine, two at each of three levels, pick straight off the tree into a conveyor. In winter, the gear comes off the top, and you can use it for pruning.

We’ve put a fluorescence spectrometer, reflectance spectrometer, optical cameras, and GPS and RFID readers on a conveyor for fruit-quality sensing and traceability. We’re trying to scan harvested fruit to get real-time information.

One of the biggest beefs of orchard managers is that people pick overmature or immature fruit or bruise the fruit, particularly apples. This technology could give the harvesting crew instantaneous feedback.

There’s also a lot of work around the world on robotic harvesting but it’s early days. It still needs a bit more research and investment.

Q. What drives ag-tech adoption by growers?

A. I think every grower would agree that tools or technologies that can increase labour efficiency and reduce production costs are a good thing.

Then, there’s variability in trees and fruit quality. An orchard is supposedly one management unit — one cultivar, rootstock, training system, and irrigation system — but there’s variability, and then you have humans on top of that. Anything that can measure and give you answers to make informed decisions is important.

Q. How do you see future adoption of ag tech in Australia?

A. Growers need confidence that whatever technology is offered will give them a return on their investment — that it’s reliable, robust, and tested. That all takes time and effort. It takes people like us to validate and endorse the technology.

There is snake oil out there. There are always those promises that aren’t delivered and products that aren’t robust. Technologies must be road-tested and scientifically proven. There are a few where we are very sceptical.

You can see why growers would be conservative, but horticulture still seems to have a surprisingly good appetite for adopting and evaluating new technologies. I’d argue that the industry is pretty innovative in general.

Apple has had the most research, and that’s where most of the early adopters of ag tech are. You could say that the apple industry is leading the way whereas some other industries have a little catching up to do.

Bonus content

RFID tags

An RFID (radio-frequency identification) tag is a small radio transponder. It reacts to an incoming radio pulse by transmitting data, typically an inventory number.

The tags can be passive — powered by radio waves — or battery-powered. Battery-powered tags can receive and transmit data over distances of 100 metres or more, whereas passive tags operate up to 25 metres.

Common applications of RFID tags include anti-shoplifting tags and pet microchips. Smart RFID tags incorporate sensors to log parameters such as temperature, humidity, and vibration.

Because they can be placed inside products or containers, RFID tags are widely used to track fruit throughout the value chain.

NFC tags

Near-field communication (NFC) technology is a subset of RFID technology. An NFC tag is a computer chip on a small sticker or embedded in plastic. The tag consists of memory to store data and a radio system to send and receive data.

Smartphones and other NFC-enabled devices connect to the tags over distances of a few centimetres. Tags have no intrinsic power sources — they are powered wirelessly by connecting devices.

Contactless payments are perhaps the most familiar NFC application, but it’s also used in identification cards and digital access tokens. Inexpensive, user-programmable NFC tags, including stickers, cards, badges and keychains, are available commercially.