Agribusiness Report: The new inventions that could make agri greener and more profitable

Cropsy's co-founders, Ali Alomari and Leila Deljkovic with one of the units that monitors vine health.

Horticulturalists were well catered for at Fieldays with a range of different innovations on show to save them time and money. Many applied new and developing technology to tasks which have proven time-consuming and labour intensive in the orchard or glasshouse, mirroring some of the innovations which have long been used in the post-harvest and shipping stages of different crops' journey to markets domestically and around the world.

Auckland-based Cropsy Technologies won both the Early Stage Award and the Young Innovator's Award. Co-founders Leila Deljkovic and Ali Alomari, attended Auckland University, graduating in 2018 as computer system engineers. But seeing a gap in the information vineyards had to work with when it came to how their crops were performing they set up Cropsy with Callaghan Innovation funding and now employ six staff, including two additional co-founders Rory Buchanan and Winston Su.

"We're poised for growth. Our mission is to deliver our tech to growers globally," Ali said.

The company's unique hardware means it can deploy high-accuracy artificial intelligence (AI) in real-time to let growers monitor every plant. And with hardware that preserves real-life colour and texture in images that other technology isn't able to produce, a data-rich map can show growers information which was impossible to achieve before; locating disease clusters, fruit counts, dying or missing plants, just as a start. The compact vision system with machine learning that looks out for anything that's wrong with plants can easily be attached to the front of a tractor. So crop losses can be minimised, estimated yield information can be supplied and any replanting required can be carried out with precision.

"We're creating technology that's accessible to growers and easy to use, which is a big deal. It breaks down barriers to adoption which so many other technologies that exist now have encountered," Leila said.

"And the technology is still evolving."

Cropsy has completed its first season of trials with vineyards in Marlborough and Hawke's Bay and looking at the possibilities for expansion in the pipfruit and kiwifruit industries where pruning, for example, could easily be optimised. "Imagine if PSA could have been detected in kiwifruit before it spread so rapidly?"

Cropsy is seeking pilot partners for commercial use of the technology next season.

PlantTech was showing off its application of AI to the kiwifruit industry in order that growers and marketer Zespri are able to make better decisions about the fruit's crop. Zespri usually undertakes two weeks of surveys in January and February to estimate the coming season's kiwifruit crop. But there have been limitations as the small sample collected isn't representative of fruit size through the whole orchard.

PlantTech's chief executive Dr Mark Begbie with a tablet on which kiwifruit growers will be able to check the firmness profile of their crop.

PlantTech saw three challenges which needed to be addressed; identifying individual kiwifruit and measuring their size, estimating distance from the camera to work out exact dimensions and estimating the weight of each based on the two-dimensional images. By applying deep learning, the exact boundaries of each fruit in the images could be worked out, and with two cameras in operation fruit position could be triangulated so 3D positioning was possible. Then an AI model was used to learn how to estimate kiwifruit weight from the images produced. The system was used on both Green and SunGold varieties of kiwifruit and validated by comparison with actual fruit weight where there was only a mean average error of seven grams, or six per cent.

So far the system has been used on over 20 million pieces of fruit with images processed in real-time. That means Zespri could scan one hectare of kiwifruit orchards an hour, data could be processed overnight and reliable size profile information could be produced the next morning. As well as better planning for accurate packing, coolstore capacity and shipping it would significantly reduce the risk of under-delivery or over-supply to markets.

PlantTech's entry in the Innovation Awards was a new way the measure kiwifruit firmness to determine harvest time. In the past using a sample of fruit from an orchard block to see whether pickers could move in, meant it was destroyed. But the company saw that advanced machine learning could be used to translate an acoustic signal from an eFast prototype developed by Eurofins to a firmness index so this could be avoided. And after trial work, it believes there's also the potential to measure fruit after harvest such as where there could be commercial value in preserving fruit condition.

The company is now looking at how to get the technology out to growers who it believes in the future will be able to use an Ipad to calculate the size of their crop as it develops.

It's also involved in work with New Zealand Avocado gathering rich data sets in order to estimate crop size and detect early signs of pest and disease attack. It's partnered with Scion using similar technology in the forestry sector and is at the early stages of working with tomato and capsicum growers to add extra value through early disease detection.
Dr Mark Begbie, the Tauranga-based company's chief executive, believes there could be benefits across the whole agricultural sector from the technology with so-called unproductive land being mapped and recognised for its uses for water retention and retaining biodiversity, he believes.

"We're starting to see it can deliver value to the wider sustainability scenario."

A father and son engineering combination has come up with Tractos, a retrofitted unit which allows mowing and spraying to be carried out without a driver.

Dave Walters is a Christchurch mechanical engineer who has been involved in design development work for the last 30 years. His son, Sean, is an electrical engineer based in Auckland and together under company name Machines at Work they were looking for innovative growers to be part of their early adopter programme. The retrofitted Tractos unit sees the orchard in 3D whilst moving, reacting to and avoiding obstacles such as fences or canopy frames. It's been developed over the last eight years after they initially looked at designing a driverless pasture measuring machine for use by dairy farmers.
"But then we saw it was better for orchards with the repetitive tasks they have to carry out," Dave said.

"People are definitely looking for a solution and with retrofitting it's at the right price."

A Cambridge blueberry grower has been developing the technology further with them using it on a mower on his 40 hectare orchard over the last year. The unit could also be used on a sprayer, giving the added advantage of keeping workers clear of any spray residue.

BA Pumps and Sprayers, which is based in Cambridge, was showing off the BA Smart Sprayer, powered by Smart-Apply® which provides a more efficient spray application and more data for growers. It was developed and field-tested at the USDA National Institute of Food and Agriculture over the past decade, where it's been shown to reduce spray consumption and reduce up to 87 per cent of airborne drift, a big environmental benefit.

The system uses LIDAR sensor technology to detect specific architecture of the vine canopy, recording the height, width, spacing and density of each plant it passes. It can be retrofitted to most vineyard sprayers. Key differences to other systems are no uploading of vineyard maps is required and the data collected is available in real-time.
"We need a more sustainable approach so let's do it smarter," said operations manager, John Dixon.

Vineyard trials by Pernod Ricard showed input savings and better data collection via a user-friendly, tablet-based control platform. There could be future uses as a scouting and reporting tool and there's already been a lot of interest from Hawke's Bay pipfruit growers.

A kiwifruit survey robot has been developed in a collaboration between Pollin8 founder, Nick Pickering, a researcher and lecturer at Waikato University, and Zespri. He has a background in avionics and said the technology involved is now affordable enough to be able to use in orchards. A prototype of an autonomous machine is planned within the next six months, certified to aviation standards. It will able to navigate its way safely through an orchard due to an onboard camera. Sensors will be able to count flower numbers in spring as well collecting data to build leaf index information, determine fruit size and potentially give an early indication of any pests and diseases.

"The labour shortage is a huge driver for innovation, and it's only going to get worse," he said.

Growers wanted to improve their margins and save money.

And with a new generation of growers staying on orchards and getting involved in their parents' business they were more likely to adopt new technology.

Early on the first day of Fieldays, five growers had already signed up to trial the unit which is at present being towed through orchards while further fine-tuning takes place.

"There will also be the choice of adding more sensors on the one platform," said Mark Graham, Zespri's innovation leader.

Nick Pickering, right, and Mark Graham with the Zespri kiwifruit survey robot.

This information could all be sent through to a kiwifruit grower on a map of their orchard, allowing them to send staff to a particular area, for example, to treat pests only where they were in large numbers.

On its first visit to Fieldays as part of Amazon Web Services (AWS) stand was Auckland company, WayBeyond. It split away earlier this year from Autogrow, where it was incubated as a start-up and now specialises in data, AI and plant science, concentrating on tomato growers.

It provides easy to use measuring and data storage solutions via FarmRoad, its farm management platform which lets glasshouse growers better understand the microclimate and what's happening with their plants.

Its Yield Prediction tool uses AI to get up to 95 per cent accurate tomato crops predictions one to six weeks from harvest. And its mobile Crop Registration app digitises what is for tomato growers a time-consuming manual data process of measuring 10 plants per area in a glasshouse every week using a tape measure, calipers and a clipboard to work out average plant and leaf length, stem thickness, flower numbers and fruit sets.

Interest in WayBeyond's innovation has so far come from large growers overseas but it hopes that in the future the local horticulture industry will be able to benefit fully from the available new technology.

Ideas that cross the border

The Growth and Scale Innovation Award was taken out by IGS Limited, which is based in Scotland and the United States so couldn't attend Fieldays because of border restrictions.
It's developed growth towers, a vertical farming system which uses the internet of things and a three-tier intelligent system to control all aspects of plants' growing environment. It promises only half the energy usage of traditional covered crop growing systems as well as using 80 per cent less labour.

Amongst other companies represented at Fieldays by the British High Commission was Inmarsat which services the New Zealand market from Australia. There it's combined with Farmbot Monitoring Systems, to supply rugged self-powered two-way connection devices in remote locations. Hardware, sensors, satellite and cellular connectivity are all combined in an end-to-end solution.

Keenan nutritionist, Alejandra Vergara, with the InTouch system on the back of one of its feed wagons.

The corrosion resistant unit doesn't require a technician for installation and its dashboard can be used by a number of people at once to perform tasks such as monitoring water levels, finding pump issues or keeping an eye on diesel tank levels. Staff can also use it to send an immediate distress signal.

At the Enterprise Ireland stand innovations ranged from the small to the large with both available in this country.

Tailpainter was launched at the end of last year by Agrify Solutions. Its invention came about when an Irish dairy farmer mentioned his problems with tail painting to his brother in the corporate world. Together they came up with the lightweight, adjustable tool which can pivot up, down, left and right for the most comfortable position for the operator. The company's Daisy Paint is low cost, water-based, has a bittering agent included so heifers won't lick it off and the packaging is recyclable. It's distributed by Fortis and available either online or through Farmlands stores.

Misty Manu from Fortis with the Agrify Tailpainter.

Keenan, who have sold their feed wagons in New Zealand for over 20 years, had their computerised InTouch mixing system on show.

It allows farmers to change their cows' diet by logging in and selecting one of a choice of 200 different feed rations with the machine automatically adding the amount of water and silage required. Data is drawn on from one million cows on almost 10,000 farms in 25 different countries to determine the most efficient diet which is then mixed in the wagon.

The system complements pasture feeding, allowing farmers to ensure the best diet for their cows at any time during the year when feeding supplements is required. Around 60 New Zealand farmers are currently using the system.

Robot harvest tip of the crops

A robotic asparagus harvester which has been under development for some years by Waikato University is being eagerly awaited by growers struggling with labour shortages.

It was first thought up five years ago with the research work gearing up two years later. A PhD student is about to submit a thesis on the work shortly.

Hin Lim, from the university's school of engineering, said the unit had been developed with Tauranga company, Robotics Plus. In 2019 the first version of the machine was trialled in California in the off-season for the crop's production here. Then over the last two years Waikato asparagus grower, Lewis Farms, producers of Tender Tips Asparagus, as well as the New Zealand Asparagus Council have been involved in local field trials at Newstead and Matamata.

"The growers were happy with it but we wanted to make sure the vision system worked well," he said.

A unit under the machine detects individual asparagus spears, activating a knife to cut them to be collected on a conveyer belt. The first arm used to cut the spears was found to operate too slowly so a new one was fitted to speed the action up. And it was also found that the technology worked much better with a flat rather than mounded growing bed, meaning some flattening out was required before harvest.

Hin Lim from Waikato University's school of engineering with the robotic asparagus harvester.

It's hoped the machine will be commercially available within a couple of years.

"We want it to be self-driving so that it will slow down where there's a dense part of the crop and speed up where there's not," he said.

"Growers being able to send it out to harvest their paddocks by itself is the plan."In the future, the same technology might be able to be applied to harvesting broccoli.