Talking to plants can get you funny looks but data science company Autogrow is developing artificial intelligence so the world's indoor commercial growers can do the next best thing – talk to an avatar about the health and progress of their crops.
Not content with lifting export destinations for its high-tech automated growing systems from 14 to 45 countries in the past three years, the Auckland-based agtech company is setting up shop in America's fresh produce-growing heart, California.
It is also moving into artificial intelligence (AI) to help feed growing populations and improve the profitability of under-pressure indoor growers and farmers.
Chief executive, director and shareholder Darryn Keiller will by Christmas shift house to California's Bay Area, joining four Autogrow staff already working in the West Coast state.
The move will help the company have its finger on the pulse of America's US$20 billion ($30.7b) vegetable production market and the US$6-8b berry sector.
Autogrow estimates there are about 20,000 forms of indoor farming in the US, from protected cropping to greenhouse operations that can span up to 30 hectares and urban farms, with California historically producing fresh produce all year round for other states that can have winter climate for up to eight months of the year. Much of the US berry crop is grown under canopies.
The company will also tap the booming medicinal cannabis growing market, which Keiller likens to a gold rush.
"It's a very dynamic market at the moment with a lot of investment being made into expansion and upgrading of (indoor) farms," says Keiller, who is driving a US$10 million capital raise for Autogrow's next growth phase.
It's the pre-profit firm's first significant capital raising and Keiller's looking for US investors, though he wouldn't say no to Asian and New Zealand money.
Autogrow is pioneering new technologies, but is actually 25 years old and was founded by the late Auckland hobby grower, inventor and electronics lecturer Jeff Broad.
It reinvented itself under Keiller's guidance in 2015.
After subsequent capital raising, Autogrow is today 62 per cent owned by a Paihia-based private investor, with Jeff Broad's sons Donovan and Kevin each holding stakes of around 13 per cent.
Literally founded in a garage, Autogrow wasn't particularly profitable when Keiller, a branding and marketing high-flyer, came on board as a consultant.
That was only because the company hadn't commercialised its indoor farming technology, he says.
But Broad had built an export market, and had entered the US market in the early 2000s through a distributor.
A strength of Broad's company was hydroponic production; another was managing plant environments.
Broad's ambition was to produce healthy and nutritious crops, of good colour and size. That vision hasn't changed, says Keiller.
But Autogrow also wants to make growers and indoor farmers - working under relentless and costly environmental pressure - profitable.
"Autogrow leapt out of the box around 2011 when it developed Multigrow, which was launched in 2012 and is still sold today. Jeff was quite ill at the time and never got to see how important a tech development it was," he says.
Multigrow is a flexible system that can be configured to perform climate, control, fertigation, dosing, batching and irrigation.
"Through that system they were able to gather lots of data – I got there in mid-2014 and believed this wealth of data from farms was highly valuable, which has proved [true]."
"Agtech was an emerging field, I saw we could reinvent all that data into a high-tech company and take advantage of all the new technology that existed for AI and data science and analytics and machine vision. All the things I was familiar with," recalls Keiller.
"I couldn't believe the industry wasn't using these things. The whole indoor farming industry was in a bubble. I saw the opportunity to advance the technology and disrupt the industry with a fusion between plant biological science and data science and advanced technology - the three things that Autogrow now utilises."
These days Autogrow routinely collects about 40 trillion data points from its operations, Keiller says.
He said the data measures everything from temperature and humidity to the alkaline or acidity of water supply, weather factors, wind direction, and solar radiation levels.
"We track those data points and use them to inform the technology about what it needs to do to feed plants, water, control the temperature etc. We can predict what's going to happen with the fruit – whether it's healthy or not, when it's ready for harvest."
Traditional technologies don't have the "computational firepower" to process all this data, says Keiller.
"You would need a super-computer on every farm."
Autogrow is one of only six companies in the world using this type of tech in the controlled environment area, says Keiller.
But it would have needed to find US$100 million to distribute its technology without an established tech platform partner.
Four partners were considered: Google, IBM, Microsoft and Amazon Web Services.
Autogrow went with Amazon.
"The idea is you pull all that data up into the Cloud then it runs through Autogrow's data processing engine, our proprietary IP. We develop that tool that sits on Amazon and it allows us to deploy it to any farm anywhere in the world as long as they have Internet connectivity.
"We can take their data, process it, and return it as useful information to say what is happening with their crop and their farm production. That's very powerful."
Growers buy the hardware and subscribe to Autogrow's technologies suite, which is "open platform" meaning it is designed to integrate with other operating systems.
The next frontier for Autogrow is AI and as usual, it's not hanging about.
It aims to be the first agtech company in the world to commercialise voice technology that allows growers to ask an avatar about the state of a crop.
Using Amazon's voice technology, Alexa, growers will be able to talk to their systems and act on what the avatar reports.
And it won't be too long, reckons Keiller, before Autogrow's AI will be able to self-correct the problems and risks it identifies.
After all, if you're a Valentine's Day rose grower whose paycheck hangs on producing those blooms in perfect condition, or a berry grower dependent on the sugar levels of your fruit to stay in business, the least of your worries is funny looks.