Agtech success doesn’t come easy. Despite a diverse farming environment, world class research and some of the globe’s most innovative farmers, Australian agtech organisations have typically struggled to capitalise on their ecosystem.
One notable exception is Agerris, a company that develops machine learning powered autonomous farming robots. Over a decade in the making, it was spun out of the University of Sydney’s Australian Centre for Field Robotics, raising $6.5 million in funding earlier this year.
The Agerris vehicles run on solar power autonomously weeding and spraying crops, and the company says it is also working on bots that can harvest produce. The 300kg robots run on battery power without a constant network connection, with the AI running onboard in real time in some bots.
In ideal weather and terrain conditions the bots can operate 24/7, only stopping to update AI algorithms and machine learning models. The bots running today are the result of decades of research and development and troves of historical data.
Agerris founder, professor Salah Sukkarieh says the success is the result of a confluence of technologies but machine learning had crucially provided researchers with “provability” for their models.
“You could actually now start to throw statistical data at it and it would spit out models but you could have performance bounds around those models,” Sukkarieh told Which-50.
He says the early funding will be used to transition the production of farming robots out of controlled labs and get them into the hands of farmers, a point many agtech initiatives never see.
Agerris is targeting smaller farms rather than large corporate customers. More than half of all Australian farms have a turnover of less $200,000 and more than 70 per cent do not employ workers, according to figures from the ABS.
Small farms are a more challenging market to crack but potentially more impactful, according to Sukkarieh. “It’s much harder but it impacts more people and may have a bit more of an impact on changes to the ag scene.”
The commercialisation hump
Despite the nascent nature of the technology the industry produces, agtech can move slowly at times. In the case of Agerris, its funding is the culmination of years of work and the confluence of several technology advances, according to Sukkarie.
“Agtech is a unique area because you’re talking about agronomy and you’re talking about biology, and then you’re trying to throw computer science engineers into it and robotics engineers into it.
“So you can never come along and go ‘I’m going to start up a company with a co-founder that might know something about plants’ and off I go from there and try and get some [investment] money.”
Sukkarieh, has spent two decades working in field robotics. In 2017, he was awarded the CSIRO Eureka Prize for Leadership in Innovation and Science, for his work in developing autonomous farming robots, among other achievements. The commercial fruits of his labour are a long time coming but the wait is not unusual in agtech, he says.
“[Agtech] is something that has to be based on science and some IP and hence spends a long time in the lab. While it’s in the lab your job and objective is to lock in some really cool innovation. Then there’s the selling of that to some VC that wants to give you some money.
“I think it has to be more than the technology then. Especially in this space. We are talking about low margins and you are talking about a significant capital investment to start with. So you’ve got to wait for many different influences.”
When the influences align and the opportunity is more than just the technology investment is waiting, Sukkarieh said.
“When they all come together then the VC investment eyes kind of glow because then you can say it’s a bet on the team and on the technology working in that particular space.
“Yes our story is successful but it’s been 10 years in the making. It’s been in the lab for that long to generate that amount of interest.”
Another significant factor is the perceptions of farmers, which Sukkarieh says have shifted from “that’s not a tractor, that won’t last long on my farm” to “bring on the technology”.
“When I was presenting Ag robotics to a conference eight years ago I was almost booed of stage. Not because [they thought] ‘robots were coming’ but because ‘that’s crap, it’s never going to work on a farm’.”
Now, after years of struggling with labour costs and availability farmers are much more receptive to autonomous bots, Sukkarieh said. And while expectations around bots are still ahead of capabilities the gap is closing quickly.
Agtech is expected to play an important role in mitigating a possible global food shortage — the global agriculture industry needs to increase food production by 70 per cent by 2050 to meet demand, according to the UN Food and Agriculture Organisation.
Sukkarieh says agtech is important but only part of the solution to more sustainable growing. He says there is lots of work to be done in plant genetics, farmers’ perceptions of what to grow and where to grow it, as well as addressing climate change.
However, near constant autonomous farming could be a breakthrough moment and one Australia could help lead.
“You’re really now doing 24/7 agronomy on a farm. That’s kind of the real aim of all this [agriculture technology]. And that’s just going to mean better productivity on farm,” Sukkarieh said.
“Australia, I would say, leads because we have such a significant backdrop of infrastructure, automation, aviation automation, logistics automation, mining automation … There’s a lot of knowledge there and passing that across into the ag environment can happen quite easily. [There are] lots of innovative farmers.”