Our world faces urgent, all-encompassing challenges. The good news is that we find ourselves in a truly exciting moment to solve these. A moment where our technological capabilities, including data analytics, artificial intelligence, and remote sensing, are converging with a deeper understanding of our biological systems, and providing unprecedented opportunities to address complex issues like climate change, food security and resilience.
At CiBO, we’ve built a software company powered by science which is focused on tackling these very challenges. We bring together the disciplines of biology and data analytics to unlock valuable insights and outcomes for agriculture that have previously been unattainable.
Our approach combines computational agronomy with advanced simulation, data generation and analysis. Plant biologists and agronomists contribute scientific understanding of agricultural systems — from plant and soil knowledge to the macro factors like weather which affect them, while software engineers and data scientists add their mathematical and statistical rigor to create a powerful, globally-scalable modeling and simulation capabilities that can answer complex questions for customers across the agricultural supply chain.
The platform we have built is as versatile as it is powerful, giving us the ability to simulate and test limitless scenarios of agricultural ecosystems for any crop, anywhere in the world, at any point in time. Being able to simulate at this scale is a breakthrough, made even more meaningful by connecting it to real-time data.
In the past, the art of prediction was limited to running historical data regressions — i.e. traditional data analytics — which amounts to trying to predict the future based on the past. We fundamentally believe that data science or machine learning alone cannot solve the problems facing our agricultural systems. These problems are simply too hairy and path-dependent on complex scientific interactions. You can’t “math” your way to a solution; you can’t run a regression to ask a “what if” question about the future of agricultural ecosystems on our planet.
Plants as Sensors
So, how do we do things differently? We’ve all seen the IBM commercial about how the Internet of Things can tell a farmer a lot about his crops before he has finished his morning coffee. To be honest, using low-cost sensors to monitor current growing conditions is the easy part.
But what if rather than deploying a bunch of sensors on a piece of land, we could instead make each plant that’s there work like a sensor? We can then observe a plant at any given point in time to understand what conditions — from moisture, to nitrogen application, to management practices — must have been true for the plant to grow as it has, and then use those derived inputs to power our modeling capabilities and understand what could be true in the future.
We can use this approach not only to predict crop yields or suggest management practices for farmers, but, even more excitingly, to finally solve some of agriculture’s trickiest questions, which have been unsolvable to date. We can work with agribusiness companies to streamline and turbocharge their R&D pipeline for developing new plant genetics. We can help big food companies overcome supply chain volatility, and have true visibility of their sustainability practices. We can partner with governments to develop effective food security strategies, and we can help donor organizations evaluate the effectiveness of international aid.
It’s not just talk anymore. With this lens into all that’s possible, we can find ways to optimize how we grow our world’s crops, so that we can develop a new agricultural system that feeds and nourishes people and keeps ecosystems healthy. We can move beyond only focusing on yield optimization, and look at how to improve soil health, sequester carbon, and keep nutrients from going to waterways.
Someday, we’ll have a virtual simulation of the entire agricultural system in our computers. Put another way: If we can imagine it, we can model it. If we can model it, we can test it. If we can test it, we can understand it. And if we can do all of this at scale — and we can — we’re betting big that we can transform it.