It started as a project to answer the blue-sky question of 'Can robots be used to map fruit trees?' and has blossomed into something big.

James Underwood was inspired by a project investigating whether or not mango, avocado and macadamia trees could be mapped using technology to answer a fundamentally different question: could it be done on a commercial scale?

"How can we affordably map every flower and fruit in the whole country … really required a whole new engineering approach," Dr Underwood said.
"For the past several years I've been talking at industry conferences around Australia and people would always ask when is it possible to do this commercially, and for many years I've been saying it's about two years off.

"So I'm pretty pleased to be able to finally say that this is now commercially available."

Stories of new research and technology in the agricultural industry are common, but it is somewhat rarer to see that research turn into a practical on-farm option for growers.

Along with co-founder Steven Scheding, a robotics and automation professional with experience in the mining industry and artificial intelligence, last year Dr Underwood launched Green Atlas to make his all-new mapping technology available.

Using a combination of satellite imagery and a sophisticated, specialised camera mounted on a quad bike, the Cartographer can count the fruit on 6,000 trees in the time it takes a person to count just one tree.

"It's a process of moving away from driving around in your ute and looking at the trees and maybe making some notes on a pad, maybe taking a few photographs, to instead having a much more streamlined digital workflow," he said.
"It's all scanned, recorded, available for inspection for comparison from one year to the next.

"So it's still driving around looking at the trees, but it's much more of a meticulous, digital, thorough kind of way of doing that."

Digitising orchard management
Dr Underwood said no matter how well managed an orchard was, there was always variability between trees that made predicting expected flowering and fruiting difficult.

"Mother nature causes that variability at some level — it's impossible to avoid — but that variability can be managed," he said.
"The data that we provide gives visibility on this and essentially allows growers to see exactly what's going on all laid out in front of them.

"It really simplifies precision management, especially when you have a lot of hectares under management."

He said growers would still use their own expertise to judge how to manage the trees, but the digital tools gave them more information to decide on their program, and could perform tasks such as counting fruit faster, cheaper, and more accurately.

"At the start of the apple season during flowering the data we provide helps them plan for spray thinning and flower pruning, which is a necessary part of how they manage the crop load and get that uniform distribution that leads to good quality fruit," he said.

"Our data gives a more accurate stocktake or yield estimate, which is useful for reporting financials.

"It's one of those things that the whole supply chain can really benefit from … from farmers at the start of the season managing the crop to yield targets through to logistics — like ordering the right number of cardboard cartons to pack the fruit."

Dr Underwood said it was rewarding to deliver new technology with real-world benefits to farmers.

"We chose to start applying it to apples and to apple growers and this is because it's where the initial use case and economics made the most immediate sense," he said.

"Our plan this year as we scale up is to address other fruit and nut varieties as well.

"We've actually had quite a bit of international interest as well, which speaks to the global demand for solutions in this space, but we're keen to keep deploying in Australia to start off with."

He said the company was currently raising capital to scale up the technology to provide services to other crops.