January 01, 2017      

A close look at the emerging model for autonomous indoor farming reveals something surprising: The vertical farm model bears a striking resemblance to that of distributed computer networking.

Indoor farming (often referred to as vertical farming) can be as basic as traditional greenhouses operated with few or no autonomous systems. This article, however, looks at the current state and near-term future of highly robotic indoor farming that typically is organized vertically on shelves rather than horizontally over acres of soil.

Automated indoor farming’s mission differs very little from that of historical agriculture. Both seek to produce food while doing the following:

  • Cutting production, distribution, and environmental costs
  • Minimizing time to market
  • Expanding output variety
  • Addressing consumer demands
  • Standardizing high quality

“There’s enthusiasm for this among investors, but there’s also some jaundice,” said Nicola Kerslake, founder of Newbean Capital in Reno, Nev. “The last wave of ag tech met with resistance” from technologically conservative farmers, she said.

Research firm IBISWorld estimates that annual revenue for the industry is about $850 million and that it has grown at 5.3 percent a year between 2011 and 2016 — respectable, not phenomenal.

There are about 200 indoor farms around the world, with about 100 in Japan, 25 in the U.S., perhaps 50 in China, and 10 in the EU.

How a vertical farm yields benefits

Still, indoor agriculture advantages over the status quo are substantial. It virtually eliminates the uncertainty of weather-related challenges and enables 24-hour operations all year.

A vertical farm also allows farmers to grow anywhere they can get a facility, water, and electricity — notably within cities — and to rapidly scale operations to changing market demand.

There are also efforts under way to apply robotics to aquaculture for fish production.

Increasingly, indoor-farming systems monitor and control inputs for optimal production: climate, water, air, nutrients, and light. This is similar to precision agriculture, which often involves aerial drones for monitoring outdoor crops.

In addition, urban farmers can stress plants in ways that change their properties, including taste.

In highly autonomous facilities, software reacts to problems and farm needs, changing the input recipe in real time.

Here is where vertical farms are developing an information technology flavor.

OsmoBot networks plant monitors.

OsmoBot uses Bluetooth communications.

From farm to tablet

For example, a vendor named Osmo Systems sells monitoring packages that connect all segments of an indoor farm to a central information system. As many as 12 metrics for each segment can be uploaded on custom schedules for both real-time review and big data analysis.

Replace OsmoBot’s mention of farms above with “computers,” and you have defined a standard enterprise network monitoring system.

OsmoBot’s software also sends configurable alerts about problems, including the detection of viruses, to operator computers and smartphones.

Some farm owners are installing radio frequency identification tags for a view of operations that is unprecedented in its granularity — right down to the conditions of an individual tub of plants. It seems that indoor crops will be ready for the Internet of Things.

Considering this comparison for the first time, Robert Colangelo, founder and CEO of Green Sense Farms LLC, laughed with recognition.

Green Sense Farms logoPortage, Ind.-based Green Sense reportedly has the largest indoor farm in the U.S. Colangelo recently traveled to China as he prepared to expand there with Dutch partner HortiMaX BV.

Colangelo said he views his 30,000-sq.-ft. facility as a single autonomous machine capable of growing 1.5 million tons of harvested produce a year. It is largely capable of caring for plants and reacting to changes and problems on its own.

The only conventional robotic systems that vertical farm like Colangelo’s needs right now are those that can seed and those designed to bring in seedlings and move packaged produce out. In the near future, he anticipates deploying drones.

“It’s easier for a drone to check plants” growing in tall racks, Colangelo said, “than to have someone going up and down on a scissor lift all day.”

Too much automation?

Yet Colangelo has a conservative outlook when it comes to deploying automation. He insists that it is too early for the industry to use wall-to-wall robotics cost-effectively.

A prime example of over-automation might be Japanese indoor-farming firm Spread Co., he said. Executives have said that the company’s next lettuce farm will do almost everything autonomously but plant the seedbeds.

The factory, due to ship its first heads in fall 2017, reportedly will deploy articulated arms among other automation systems.

“The Japanese government is offering a lot of incentives for farmers to get into vertical” agriculture, said Colangelo. “With free government money, companies are investing pretty freely in robotics,” including industrial systems.

He said companies are better served by rigging facilities with sensors to document everything that occurs so that operations can be highly optimized.

Mike Betts, director of investments for crowdfunding platform AgFunder Inc., said Colangelo “is right on” when it comes to over-automation and facilities design. These are unique operations with unique challenges, according to Betts. AgFunder is considering becoming a venture fund outright.

The automated agriculture sector is going to be fairly fluid over the near term, according to Newbean’s Kerslake. She said it is likely to be dominated by “high-net-worth individuals investing in single-site farms designed to serve one community and larger institutional investors looking for larger farms and scalable opportunities.”

More on Agricultural Robotics: