Efforts to blend artificial intelligence and farming are coming to fruition, with the market predicted to grow at a compound annual growth rate (CAGR) of 22.5% between 2017 and 2025, to $2.62 billion. Recent developments on the frontier of AI in agriculture have resulted in innovative new technological applications, like the installation of Internet of Things devices in farms to generate vast amounts of new data. Two key yet contrasting players are the privately held company Cargill and the global initiative CGIAR.
Cargill is the largest privately held company in America, and it has started combining AI with its shrimp farming and cattle ranching divisions to reduce waste. It has made use of the same machine learning techniques utilized in high-frequency trading bots, speech-processing virtual assistants, and self-driving cars. Cargill uses these technologies to allow farmers to scan cow faces and listen to the sounds made by shrimp as they eat. By monitoring the sounds that shrimp make, farmers can discern whether they have given them enough food. If they've given them too little, then they will adjust for the next round of feeding. But if they have given them too much, then they will know to give them less feed next time, thereby reducing the chances of future waste. Likewise, monitoring dairy cows' behaviors and diets may allow farmers to navigate around such things as pregnancy rotation schedules, to discern optimal insemination periods and ultimately result in better milk output.
The farmers adventurous enough to risk incorporating these new technologies are naturally concerned about not mismanaging or breaking their new tools. These farmers are turning to Cargill for help in translating to them how the new technologies work, and ways they can be seamlessly incorporated into the farms' existing work systems.
Elsewhere in the US, similar efforts to mobilize AI for the benefit of farmers are underway from the CGIAR Platform for Big Data in Agriculture. The CGIAR team seems to use big data and computational analysis to increase farmers' efficiency and reduce the basic risks inherent in farming. CGIAR hopes to provide a better repository for researchers to use, enabling them to share and manage agricultural data. The goal of CGIAR is eventually to be able to integrate data from farms around the world into algorithms that generate insights of practical value to the farmers. For example, algorithms to create probabilistic models for seasonal forecasting would provide a huge advantage to smallholder farmers, who collectively produce approximately 70% of the world's food supply while individually working on farms under one hectare large. Such a model can allow them to know half a year in advance which crops will be suitable for the upcoming season, the ideal planting time, and the best ways to optimize farm management based on weather patterns.
Unlike Cargill, CGIAR focuses much of its efforts on helping farmers in the developing world, who face basic obstacles like a lack of irrigation. There have been recent efforts to forge a collaboration between CGIAR and the private sector in a move that may bring humanity closer to ending world hunger. Scientists at CGIAR have built data-driven models describing various crop production systems that take into consideration the crops' management, environmental conditions, and ensuing harvests. Big data can then step in with specific details collected from the thousands of plantings under varying conditions, thereby enabling scientists to analyze the relationships between the circumstantial variables and the resulting harvest.
Still, challenges arise with the sheer number of variables present - including minerals, cultivation, germination, irrigation, soil nutrients, pests and disease, to name a few - that must be evaluated in order to generate a complete portrait of a farm's internal system.