Transforming the agricultural industry with machine learning
Adam Neilson, Chief Technology Officer at Wefarm discusses the ways in which machine learning can transform the African agricultural industry.
Ever since Fritz Lang's Metropolis was first shown in the cinemas of 1927, the film industry has been forecasting how technology of the future would transform humanity. Fast forward to current day and we may not have flying cars or replica people mining in off planet worlds, but we do have something that I believe in the long run will be far more important to the future survival of our species.
Over the last few years, machine learning (ML) has steadily rolled across the “hype cycle” from the "peak of inflated expectations" to officially entering the mainstream, and is now beginning to quietly revolutionise every aspect of our lives. For us consumers, it's now so deeply embedded within so many of the everyday products and services that we interact with it's almost invisible. From the post-processing on the photos we take, to the search results within our browsers, machine learning has become omnipresent.
Perhaps more importantly from the perspective of humanity is how ML has begun reaching to the core of global industries. Optimising processes for manufacturing, health, logistics, industrial design, it's deeply ingraining smart design and routing efficiencies that will on-the-whole cascade benefits down to us all.
The population of this planet is not getting any smaller. The United Nations has predicted there'll be nearly ten billion of us humans by 2050 and 11.2 billion within the 50 years following that. Clearly that's a huge increase in the number of mouths to feed, but then add new challenges presented by climate change (we already know many crops in Africa and Asia are now threatened by warmer temperatures and pests like Fall Army Worm) and it’s clear in order to do so we need an industry well equipped with adequate knowledge and tools to adapt quickly.
Our belief at Wefarm is that a lot of the information, tools and services required to help make farmers more successful already exist––it’s the access to them that is lacking. Currently small-scale farms are numerous: 1 billion across the world who grow 70% of our food. But half of these farms don't have easy access to knowledge and advice, whether it’s choosing what to grow based on market conditions, how to effectively deal with a pest, or finding the right fertilizer to maximise yield. Cut off from "the information superhighway" these farmers can't ingrain smart design and improved efficiencies on their farms beyond the advice of local peers.
Wefarm is helping overcome this challenge by building a trusted community for farmers, where our members are able to gain and share access to the resources needed to be more successful. In order to do that, we need to work with the technology at their disposal, not against it. As much as we may wish everyone had access to the internet, the truth is they don’t, and won’t anytime soon. To enact serious change we have to be able to operate with what rural farmers do have access to, in our case SMS. We also need to respect the huge levels of variations and nuances in how we as humans interact with each other, especially when multiple languages and local phrases are introduced. For anyone who may be aware of a famous comedy sketch by the Two Ronnies about a customer going to a hardware store and asking for what sounds like "four candles" but after multiple mistakes the shopkeeper discovers he wants "fork handles", you’ll realise it doesn’t take much for us to misunderstand each other. So take a moment to consider the scale of the challenge in getting a machine to understand these differences.
"Artificial Intelligence" has always meant to augment human capability and help us reach ever greater heights. Well now it's time for AI to rise to that challenge and shine. It’s been impressive to see DeepMind's AlphaGo and its historic victory over its human opponent, but what we need now is to apply these incredible technologies to solving serious challenges on a global scale, and put the power of them in the hands of the people who will grow the food we all eat.
Here at Wefarm we are doing just that, we’re using ML in conjunction with our leading Natural Language Processing libraries to help drive relevant connections even when we only have the limited content of an SMS to work with. In fact, we now have 1.9 million farmers across Uganda, Kenya and Tanzania on the network, who have collectively asked more than 4.6 million questions, and received back over 10.6 million answers thanks to ML. Impressively, as we have continued to learn, those answers are coming back on average within 6 minutes of the questions being asked. But our ML doesn’t stop there, the data we collect from those questions and answers can help us map wider trends and issues at a macro level, as well as connect farmers with relevant retailers who can provide tools and services to help overcome challenges they’re likely to face. To help with this we launched Wefarm Marketplace 9 months ago, and have seen it enable $1.5 million of sales of products to this end already.
It's been a thoroughly rewarding and humbling experience to work on this project. We're just getting started, and thanks to our recent Series A backing of $13 million, we’re planning to keep moving fast. Because as I said earlier, the world’s not getting any smaller, and the challenges facing farmers are not getting any easier, but together, as humans and machines, we’re confident we have a way forwards. The invisible power of ML in agriculture may not always feel as shiny as flying cars or replica people, but it sure feels needed, and that seems far more exciting to me.
For more information on business topics in the Middle East and Africa, please take a look at the latest edition of Business Chief MEA.