![Hone Ag managing director Peter Johnston at the GRDC Grains Research Update in Adelaide. Picture by Paula Thompson Hone Ag managing director Peter Johnston at the GRDC Grains Research Update in Adelaide. Picture by Paula Thompson](/images/transform/v1/crop/frm/215078332/39b1ec24-a579-4fff-a6ae-432e63ef899d.JPG/r38_123_3949_2534_w1200_h678_fmax.jpg)
Rapid on-farm sampling can enable production decisions to be optimised and help ensure farm enterprises maximise the economic benefits of the quality that they produce, according to Hone Ag managing director Peter Johnston.
Subscribe now for unlimited access to all our agricultural news
across the nation
or signup to continue reading
Mr Johnston, who started his career as a plant breeder, was a speaker at the recent Grains Research and Development Corporation Grains Research Update in Adelaide, taking part in the Ag Tech Startups Forum.
He said Australian farms had an ever-growing amount of on-farm storage due to a range of factors including increased domestic demand and more header throughput.
"With grain coming off so quickly, the first decision is, do I put it in the low, mid or high-protein bin?" he said.
"These decisions have a direct economic impact on the revenues of farmers."
Mr Johnston said in the past sampling required large instruments and highly-trained technicians, but Hone Ag provided a portable instrument for anyone to test anywhere, anytime.
"It's really hard to get a sample that's accurate out of the field and delivered to a lab in a timely manner from where you can make a decision," he said.
"We've inverted the problem of having to take the sample to the lab by actually taking the lab to the field.
"This has been enabled through advances in machine learning.
"What used to take three to four weeks, we're now doing in three to four hours."
Hone Create is a cloud-based machine learning engine that the company specifically designed and engineered to create complex models and calibrations.
Creating this enabled Hone Ag to design, engineer and produce a handheld spectrometer specifically for in-field agricultural applications.
It also allows the company to create models and calibrations extremely quickly from less samples, using the processing depth of the cloud, and to develop a self-learning validation process to ensure model performance.
Mr Johnston said the unit was designed to be ruggedised to suit agricultural production systems.
It also uses digital correction on the spectral curve.
"It's a bit like noise-cancelling headsets to get rid of background noise so we can really concentrate and focus in on the spectrum that we're actually looking at," he said.
He said economic analysis of Hone Ag showed positive results for producers.
"If you look at a 1500-tonne wheat producer, if you make a better decision on one-in-10 loads (with an upgrade), essentially you're getting a payback of $4250 a year," he said.
"It makes economic sense to manage your quality to make sure that you're not giving away the quality."