How did you come to start Connecterra?
Previously I’d been working for Microsoft. I’d written a paper on IoT for them back in 2007, way before the technology and connectivity had taken off to support IoT solutions as we know them today. I was interested in sensors but I was left with questions about what to do with data. This is 2007-2014 and machine learning was only just coming along.
I wanted a new challenge and I wanted to solve a real problem. So for inspiration, I turned to the list the UN has published on major issues the planet is facing. In tandem with this, I had been living in Dubai and returned to Europe. To ring the changes I chose a place to live that was fairly remote. A rural island in xx where there are more cows than people.
I was thinking about these big problems and I had lots of cows outside my window. So I guess it was inevitable that I ended up focusing on food production – dairy food production in particular.
So what is the challenge you’re tackling?
The big picture is this: by 2050 there will need to be a 60% improvement in global food productivity, in order to feed a population of 10Bn people. There are currently 114 million dairy farmers in the world and 570 million people associated with the dairy farming industry. 10% of the land is used for dairy farming and 14% of the global food trade is from dairy. But the industry isn’t particularly efficient and production methods simply aren’t sustainable. Plus your average dairy farmer is c57 years old. Possibly not the primary target you’d imagine for deep tech innovation…but that’s where we decided to focus. Plenty of technology companies swerve the agricultural market – but this turned out to be to our advantage, as it’s created a clear opportunity for Connecterra to lead.
How did you find the experience of founding your own company?
Right at the outset, I set myself a challenge. Within one year from my start date, I wanted to pitch a really cool idea at Web Summit. Not only did we achieve that target: we won. So Connecterra – which means connected earth – was on the path to growth. By that stage – in 2015 - we had a solid MVP and we were already talking to Danone. But winning Web Summit put development into overdrive. 2 years later, we were working with several major industry players and we were building quite a reputation.
How did you focus your efforts to secure success for your company?
First of all, I need to point out that our whole team at Connecterra is made up of technologists. And right from the outset, we were clear that it wasn’t just data we were after. For me the more raw data you have the less likely you are to make any sense of it.
Instead, our focus was on insights – so we were looking at machine learning. But you’ve got to remember that dairy farming is a people business. We didn’t just need insights about the cows, we needed insights about the farmers – to harness the power of their experience and expertise as well. So we set about building a solution that combined the number crunching capabilities of AI with the human ability to ask the right questions.
What we developed was IDA – the dairy farmer’s assistant. It’s like Siri or Alexa, but specifically for dairy farmers. It’s built on a wealth of deep learning but delivered in simple terms. We’ve got IDA to a point where we can put the technology into the hands of a non-expert worker on a dairy farm – and if they simply follow the instructions IDA provides, it will increase the efficiency of the farm by 30%. This isn’t a pie in the sky number - we have hard data to prove it.
So we were working hard on the technology, building the data and insights, making IDA easy to use (and branding it…to make it easy to sell). Plus we had to rapidly grow the team.
So how does IDA work?
At our core, we’re an analytics and insights company that serves farmers and their industry partners. But we have hardware and software. The cows we monitor all wear a device – which comes with its own challenges.
We had to create low-powered devices with long battery life. Right now the battery lasts five years and the range is 1km. There is no operating system – so this is intelligence at the ‘Edge’ and it’s been sufficiently exciting to get some interest from Google.
The data effectively goes to a ‘black box’ and the AI and ML kick in to produce insights. At that point, the data disappears and what we’ve got is the ability to advise, predict and guide farmers on things like cow ovulation, health and feeding habits. In effect, we’ve got the equivalent of thousands of years’ of dairy farming know-how captured within our data set.
It helps the individual farmer to improve productivity and manage the herd. But the ‘big data’ we’re building also allows us to make predictions with a broader impact.
If IDA covers 20-40% of the cows in a region, we can predict productivity for the entire geography – which has important implications for things like supply chain and logistics as well as for consumer-facing retail.
So what does the future hold for Connecterra and IDA?
The end game has to be to make farming more efficient – so we can maintain protein input into the food chain more sustainably.
We’ve built a robust business model that delivers payback to the farmer in as little as 60 days, so as we look ahead we’re aiming to scale by broadening our customer base amongst farmers. But we’re also getting a lot more engagement with corporate customers who can harness the insight were generating into the supply chain.
What insights have you gathered that you would share with other founders?
One thing that surprised me as we got involved in this space was that there are few platforms already in existence. In order to achieve the ‘ground to cloud’ coverage that we wanted, we’ve had to build a lot of things from scratch. When you decide to get involved in technology that is genuinely ground-breaking, best be prepared to get stuck in!
My other observation is that developing hardware is a task that is not for the faint-hearted. We’ve faced a lot of challenges on our journey so far – but with the right kind of problem-solving approach and engineering talent, none have been insurmountable. In fact, our hardware has come on leaps and bounds and is getting a lot of positive attention in the media.
One further insight is not to make any assumptions. I mentioned earlier that some might perceive the farming community as laggards in terms of technology. But really that couldn’t be any further from the truth. When presented with insights from data in ways that are accessible and practical, we’ve found farmers to be an extremely responsive community of technology adopters.