The rural edge
The world needs to increase food production by up to 70% in less than 30 years. It’s a significant challenge which, if not overcome, could push millions into famine. The answer involves IoT and, as Melanie Mingas finds out, a vast web of edge infrastructure that many other industries could tap
Humanity faces many challenges today – and population growth is just one of them. In fact, over the coming 28 years the total number of humans on the planet is on track to reach – and exceed – nine billion, driving up our requirements for food by 60-70%, according to current estimates.
Compounding this issue, the total land available for farming is decreasing, rural workforces are dwindling and the cattle some of us eat will also need feeding.
It isn’t the first time agriculture has faced such issues, but the arsenal of tools that emerged over the past 50 years are nearing the end of their useful life. Now, the answer lies in data and connectivity, which will become integral to the precision farming that the future demands.
In North America the Rural Cloud Initiative (RCI) is one of the bodies exploring these issues. It is a coalition of rural telcos and tech providers – and it is bringing digital transformation to rural US.
“We need to more than double food production in 28 crop cycles – that isn’t a long time,” says Nancy Shemwell, COO of coalition member Trilogy Networks. “There is only one shot per cycle and it’s complex work.”
Trilogy takes the necessary cloud capability to the rural edge. And when the Federal Communications Commission (FCC) was looking for somebody to head its US$1 billion Precision Agriculture Task Force last year, it recruited Trilogy CEO George Woodward.
“One of our initial commercial sites was at Hurst Greenery in Missouri, where they have a series of greenhouses and grow crops,” Shemwell adds. “Putting in a couple of simple applications allows them to turn heat on and off, control fans, water mixing, and so on. They believe that will drive a 10% productivity boost and cost savings with just a couple of simple applications. That’s the power of these technologies – even simple technologies.”
Hurst Greenery was essentially the pilot for Trilogy’s Farm of the Future. For its edge delivery platform Trilogy deploys its nationwide LinX Network and ConEx regional edge cloud platform, with the architecture powering a variety of advanced precision agriculture applications – sensors, monitoring, satellite mapping, drones and robots, to name a few. Phase one was completed in October 2020 and, facilitating the creation of a wider ecosystem, Trilogy is now looking for more telco and data centre partners.
CTO and co-founder Venky Swaminathan says: “You need robust wireless connectivity on the farm.
“Then, how do you connect that from the cloud? And by that, I mean distributed cloud, depending on where the application is. For us it means anywhere from the farm to the network edge, the telco edge, or to a hyperscale cloud,” he adds.
In Trilogy’s words, the network it is creating will be the largest edge network in the country, and to build it, Swaminathan says, you need compute and storage capacity “everywhere in the network from the farm to the network edge, the telco edge and to the far-away cloud”. Then, of course, you need data centres, too.
“We are partnering with the likes of Vapor IO, EdgePresence, EdgeMicro, which have these micro data centre structures that can complement an existing telco facility – it could be at a cell site, a central office, wherever you need to add compute capacity,” he explains.
While projects such as these make deployments sound simple, getting the required infrastructure into remote locations turns humanity’s agricultural conundrum into a digital divide challenge.
In Australia – where government figures state the agriculture industry drives up to 12% of GDP and 13% of export income – digital tools promise to assist with everything from drought to monitoring wildfires. However, there are major challenges in getting connectivity, and even power, to the regions.
Jonathan Eaves, founder and CEO of Edge Centres, plans to deploy 20 solar-powered rural edge facilities by the end of 2022 as the first phase of an edge network build that aligns with the same ecosystem philosophy seen in the US – and agtech is one of the many sectors set to benefit.
“It means anybody – a farmer, for example – anywhere in the country can build a far-edge facility, have it solar powered, for what I would call reasonable cost, and power a community,” Eaves explains.
While the plan sounds simple enough, there are many challenges. For example, it’s a six-month lead time to secure a direct electrical supply in Australia, which can blow out timelines on projects that depend on rapid deployment. It’s the reason Edge Centres is “solar prime”, allowing it to deliver services without delay.
“I’m not going to do 1MW with solar because the battery and panel requirements are huge,” Eaves says. However, with each facility on a 12-metre footprint, the 16 cabinets become 64 racks at 1kW each, plus 25kW for operations.
“It’s just under 100kW – that’s doable. So we have now created a footprint that is replicated in every location, and it’s purely off solar and batteries.”
The Grafton facility is now connected to the grid, which is used for top-up and back-up rather than as a primary source, and the same model will be used for subsequent deployments.
Another challenge in the regions is the price of connectivity. IP transit in Sydney averages A$3 (US$2.18) per megabit but it doubles at each subsequent tier.
“Region one is double, so A$6. Region two is A$12, region three is A$24 – all for exactly the same thing as in Sydney. If you are in agriculture it’s four times as much, minimum, to have that same level of connectivity as it would be for your metropolis counterparts,” Eaves explains.
“The first part was just to level the playing field. I charge A$3.50 anywhere in my network, so that’s Cairns to Melbourne,” Eaves says.
Solving another problem, Edge Centres lays its own fibre, hosts masts for Everest and, in time, is planning to deploy a LoRaWAN network, with access charged on a per kilobit, per minute basis.
“We will start creating the services to support the solutions that are coming even before they arrive, because with LoRaWAN we have 100km ranges – because we have our own fibre in the ground, so we can effectively cover a larger landmass. If you want to send a LoRaWAN ping every second for 60 seconds every day of the week, it will be X. If you want to send one ping a day just as a heartbeat, it will be Y,” Eaves says.
As with everything in connectivity, speed is key; but the economics of deploying low-latency infrastructure is often prohibitive.
Bluebird Network is tackling this through its deployment of fibre, wireless connections and points of present (PoPs). Chief revenue officer Trent Anderson says: “There’s more and more computation inside these $1 million tractors – for navigation, fuel, irrigation, fertilisation – and they are tapping into huge databases to learn about moisture, topography and crop readiness and so on.
“It’s about maximising money going into the farm and pulling it back out as profit. It used to be a roll of the dice; these technologies are about making that equation tighter,” he adds.
In the future, Anderson says, the cost of related fibre builds is likely to be covered via private funding and public-private partnerships. However, more inventive ways to connect will also emerge.
“The future is likely a hybrid of connectivity from the fibre or cell tower and the IoT of equipment communicating with each other directly on the farm.
“What’s most likely is a well-co-ordinated effort with the large carriers having maybe microwave or an old technology, something analogue-based, still in play in those communities – for example 3G, legacy. Also, with public funding being thrown into the mix, there will be a need for backhaul, and we will help enable distribution as other builders take advantage of grant dollars.”
With the wide-scale adoption of agtech solutions, farming joins a rapidly growing list of forward-thinking industries leveraging the benefits of the Fourth Industrial Revolution. Think banking and logistics: both already have bespoke infrastructure and dedicated connectivity to improve latency and security.
“Whenever you have an area that is underdeveloped technologically, it means the next generation of technology can have an even greater impact,” says Loren Long, chief development officer and co-founder of DartPoints.
For a number of years, he says, agriculture and its supporting sectors have been “pushing the boundaries of technology but without the telecoms or interconnect standpoint”.
Forced to go it alone, Long adds: “They either have to put the onus on the farmers to make a large investment in digital infrastructure to support the applications, or they have to build an application that requires so little bandwidth it holds itself down.
“To have a DartPoints with an internet exchange, high-powered compute and hybrid cloud, and you bring all those resources to that market, it means the innovative impact can be even greater. It unleashes a whole new era of technology – especially when it becomes more ubiquitous,” he says.
While the emergence of agtech underpins global food security, it could also become a trajectory for wider network coverage, driving further digitalisation of other business that operate in – or pass through – the most remote locations.
By way of example, DartPoints chief strategy officer and co-founder Hugh Carspecken explains: “Don’t think for a second that the automotive industry is not coming – and they will be resembling agtech very closely.
“We always think about the tier 1 markets, but Americans go on long-haul drives and they hit the rural markets. In the automotive industry, Daimler Porsche are already doing it – they are ultimately building business models that will also require this infrastructure. So you’ve got everyone now able to subsidise the infrastructure.”
As for how that infrastructure will cover the necessary locations, Eaves sees a triple-layered approach.
“You will have hyperscale, edge and far edge,” he says.
“I see far edge, little two-rack pods, further out again in places like the Broken Hills, Broom, Alice Springs, at the far edge end-points where you know you don’t need a lot, but you still need that connectivity capture,” he adds.