Meet the AgTech innovators with their fingers on the pulse
From little things big things grow and for Anna Falkiner and Andrew Hannon it’s the tiny lentil that has driven their big ambition to modernise Australia’s $35 billion grain industry using Artificial Intelligence (AI).
Their Stone & Chalk ag tech start-up Cropify is combining agricultural experience with Artificial Intelligence and machine learning expertise to develop technology they believe will improve the classification of pulses and grains and reduce conflict between sellers and buyers around quality, from paddock right through to the export and importing customers.
Presently most grain assessment is done subjectively by eye, but with his years of insight in the grain industry, Hannon had identified the need to provide an objective, accurate, faster and repeatable means to assess grains.
“There was a big issue relating to disputes and dispute resolution that arises out of the subjective nature of classification which is traditionally performed by human classifiers,” Hannon says. “Our objective is to be accurate and repeatable, eliminate the subjective testing and increase confidence along the entire supply chain.
“We have gained confidence in the industry and we will soon see industry adoption.”
Their software – developed with the support of the Australian Institute for Machine Learning, also based at Lot Fourteen – uses high-resolution imagery to assess an industry-standard sample of lentils in 90 seconds compared with the current industry average for pulses of about 24 minutes.
The Cropify technology can pick up defects in the lentils as well as identify weed seeds that could contaminate shipments. It also provides a solution to current shortage of grain classifiers in regional areas.
“Being able to do the whole sample size, in line with Grain Trade Australia standards and do it repeatedly and accurately and quickly is going to be a game changer, delivering efficiencies across the grain supply chain” Hannon says.
“Then there’s the environmental benefit, you haven’t got a line of trucks burning diesel waiting for two hours to have their product tested.
“There’s also presently a huge plastics issue because presently every one of the 4.4 million grain and pulse samples done in Australia has to be stored for 12 months in single-use plastic bags whereas our technology will store the sample image in a database.”
SA is Australia’s largest lentil producer, and pulses were chosen as the testbed for the technology, because they are notoriously difficult to classify, Falkiner says.
“We’ve kept our scope narrow so we can demonstrate to industry that we can solve the challenges they face in order to create demand and bring on investment,” she says.
The benefit to farmers is through more accurate classification they can have better control over how, and where, they sell their product. Test results can be shared with stakeholders in real time, allowing for more informed marketing and storage decisions.
The company is undertaking a $2 million seed funding round to accelerate their growth and is aiming to do first commercial trials of the prototype in the first-half of 2024 in South Australia.
“We’ve got lots of industry interest in Australia and internationally, potential customers are all really keen. The accuracy level that our prototype is operating at is really generating a lot of positive interest,” Falkiner says.
The pair have largely self-funded their enterprise since starting the business in 2019, but have received grant funding from Primary Industries and Resources SA through the inaugural AgTech Growth Fund.
“We’ve been fortunate to also have support from the Australian Institute of Machine Learning (AIML) that’s allowed us to get to a point that we are today,” Hannon says.
“The proof-of-concept work completed by AIML was instrumental in confirming we were on the right path to solving this global issue for the grains industry. Without this foundation work our path to market could have been a lot slower.
“We also have grain growers who have come on as investors and we have international parties interested as well.”
Falkiner and Hannon predict their technology alone could potentially eliminate 138 million tonnes of CO2 emissions and 22 million tonnes of plastic annually.
Falkiner says one of the biggest challenges of starting a business was trying to do everything on their own.
“With our first iteration of our hardware we were literally buying LED panels off the internet and ripping them apart and doing the tech ourselves, then trying to find people who could help us,” she says.
"*" indicates required fields