In 2010, a company called Freight Farms set out to create mobile hydroponic farms that would sit inside up-cycled shipping containers – the idea being the ability to create a local produce ecosystem anywhere in the world.
It was just about the time the 'internet of things' began to gain traction as a term, and it was clear connected sensors would have a big part to play. The company tells ComputerworldUK why it picked LogMeIn's Xively to manage that connectivity.
"When we started, our whole goal was to grow food anywhere, and figure out a way that people without access to land and a lot of pre-existing agriculture knowledge could jump in the game," says co-founder Jon Friedman. "We started looking at shipping containers as the main method to get that controlled environment to operate anywhere in the world.
"And as we started going down that path, we realised there were a lot of technical parts in the system we could get really smart about, that we could automate, and that we could develop a lot of different products around."
The final project is called the Leafy Green Machine (LGM), starting at $82,000, a pre-assembled hydroponic farm built inside an upcycled freight container. According to the company, the LGM can produce yields at commercial scale in any climate and any season.
But when Freight Farms was first setting up, the internet of things system in place – sensors and so on to detect humidity, CO2 levels, etc. – had been put together in-house. That's when the team noticed Xively, a 'connected product management' platform designed to provide a simple way to monitor and control connected assets. Through APIs, it can be integrated directly into customer software, in this case the company's own 'Farmhand' suite.
Freight Farms looked at some other IoT platforms to simplify the management of its connected devices both in-house and for customers. Amazon Web Services' IoT wasn't complete enough at the time, according to Friedman, and with ThingWorks the company would have had to host its own java instance on AWS.
"We settled on Xively for its simplicity, it was simple to use but with a pretty rich feature set," Friedman says, adding that the open source protocols it operates on and an approach that allows the business to interact with it through APIs were also a major factor. "It's been really straightforward."
"When we first met Xively we had been using our own version of an IoT system, and we put it together with all sorts of sensors and hardware," says Friedman. "Where it be something from an electrical store to a current sensor – the type of things that you pick and choose from different industries, to bring together to make it a really intelligent climate monitoring system.
"But as you do that, you realise how hairbrained your product management has become, and now only one person can really tell you how it runs – the greenhouse manager."
Friedman and the team began to think that as the system grew it could become unwieldy – and besides, customers might have issues or just want to develop on top of it.
"We needed some kind of framework to bring this all in and have it all make sense," Friedman says. "We want to scale this, and we need to bring it into a platform so we can actually see what's going on, manage it securely, and build modules off of it."
Director of farm technology Kyle Seaman adds that there are 10 climate sensors and 40 other pieces of equipment that are continually tracked and monitored. "By the time we started working with Xively we already had 40 of our containers in the wild," he says. "So we had to think about backwards compatibility as well."
"Where Xively fits with us is they're really the connectivity layer," Seaman explains. "Every farm was automated but they were talking to no one, it was juts local processing. When we rolled out with Xively we are able to bring everyone on, our previous farmers and all new farmers. Really what's unlocked is instead of having 80 farms working on their own and running old firmware, we had these farms connected across our network that we're now continually learning from."
And that translates to real-world benefits not just for Freight Farms to understand how best to optimise its Leafy Green Machines, but also to the farmers themselves.
"If we see common behaviours from farmers, maybe specifically tuning things, we have the ability to automate that now," Friedman says. "Xively changes how you interact with these farms: we can build a better climate for growing, whether it's basil or kale, and give peace of mind by knowing if a pump happens to run too long or there's an issue with the plumbing line, we can kill stuff remotely and save a lot of damage to the farms."