Co-Founder and CTO of Prospera Technologies, leading the company’s vision to transform the way food is grown using data science and AI.
The human race has come a long way in our ability to produce food at scale. Historian and author Yuval Noah Harari refers to it in his book Sapiens as “an agricultural revolution,” using wheat as an example.
Ten thousand years ago, wheat was a wild grass that grew in a relatively small region in the Middle East. Today, wheat can be considered one of the most successful plants in history, according to the evolutionary criteria of survival and reproduction. In regions where wheat never existed, such as the Great Plains of North America, you can drive for hundreds of miles without seeing anything else but wheat fields. Worldwide, wheat now covers over 2 million square kilometers of the world’s surface, nearly 10 times the size of Britain.
In spite of these advances, even when technology has allowed us to improve and streamline food production, more than 800 million people around the world remain undernourished. Maybe it is the turn of AI-driven agriculture, autonomous robots and better insights to help close this gap and eradicate famine. With the world’s population growing at a vertiginous rate, this might be our generation’s biggest challenge.
Artificial intelligence, predictive analytics and other technologies are quickly becoming essential tools for growth across multiple industries. Agtech is no different. What’s more, in spite of the economic slowdown in 2020 due to Covid, agtech investment is booming. According to Crunchbase data, 2020 is poised to top 2019’s $4 billion VC funding for agtech startups worldwide.
Are we on the cusp of a second agricultural revolution? And more importantly, are food growers equipped to embrace and leverage all these new technologies and tools being developed for them?
AI, machine learning and robotics: A digital revolution?
Digitization of data collection processes in the crop fields will be at the heart of the next agricultural revolution. But this will take time, especially when taking into account that most farmers or agronomists are still using pen and paper to collect data.
This is bound to change due to the increasing amount of useful data points that are emerging from all types of tools. Sensors, cameras, robots, drones and other artifacts deliver an increasing amount of data that is exponentially richer in quality and quantity. What’s more, the data brings insights in real time, with devices that are on the ground 24/7. However, other challenges such as data orchestration emerge.
With so many sources of data and insights, how can we expect food growers to streamline all the different streams of data and create a single source of truth to help them inform their decision-making process?
A recent report by USDA’s Economic Research Service on Agricultural Resources and Environmental Indicators highlighted the introduction of precision agriculture technologies. Variable-rate technologies (VRTs) enable farmers to make customized land management decisions to optimize the use of seeds, fertilizers and pesticides. This allows farmers to manage inputs foot by foot, with an unprecedented level of precision. The report showed that by 2016, between 15% and 40% of farms in the U.S already used variable-rate application equipment.
Will the future role of farmers resemble more that of a data analyst?
Having worked with farmers and food growers across both open fields and greenhouses over the past five years, I’m amazed by the strides the whole agriculture industry is making toward automation. It wouldn’t be crazy to predict that today’s manual processes in fields and greenhouses could be automated within a decade. From autonomous robots that will replace humans in processes such as picking strawberries or apples to AI-driven sensors that will inform farmers (or machines) when is the optimal time to fertilize, irrigate, plant or harvest.
The decision-making process will be greatly influenced by these technologies that provide an unprecedented level of data and insights. The management and orchestration of these new technologies will be a challenge. Because of this, simplicity and usability must be at the top of the agenda for those of us who develop software and hardware.
Precision agriculture and automatization of food production are priorities for any food grower and, to an extent, for any stakeholder within the food industry. The technologies to achieve this new agricultural revolution are within our reach. In fact, it can be argued that this upcoming agricultural revolution is only possible thanks to two other technological breakthroughs: deep learning algorithms and cheaper Internet of Things (IoT) hardware.
In the book Prediction Machines: The Simple Economics of Artificial Intelligence, three economists from the University of Toronto demystify AI by examining it through the lens of economic theory. In essence, the book explains how the use and expansion of prediction machines are becoming as cheap and widespread as the expansion of the use of electricity or cars during the early parts of the last century.
It’s now up to both the agtech industry and farmers to learn from each other and reimagine the future of agriculture. The judgment of expert farmers and agronomists will become more valuable to augment the input of artificial intelligence. However, jobs will have to be redesigned and workflows altered. It is, after all, a revolution.