Hunger is a persistent global challenge. The world needs to sustainably feed a growing population — an estimated 9.7 billion souls by 2050 — with a per-capita agricultural area (the area available to feed one person) that continues to shrink (from 0.367 hectares in the 1960s to 0.192 currently). On top of that, about one-third of food is wasted.
We are facing many obstacles, including limited arable land, environmental issues, and restricted availability of water and labor. For example, currently, more than half of the world’s land is used for agriculture; the remaining land is a mix of forest and suburbs/cities. Creating more farmland likely would require converting forest, which is a major climate and biodiversity concern.
Our team is tackling the issue of available labor. Row crop production is a $100 billion-plus industry in the United States alone. This type of production is heavily mechanized but still suffers from a scarcity of seasonal skilled labor.
We are focused on encouraging the adoption of robotics in producing commodity crops, such as corn, soybeans and wheat. Ag now uses the biggest machines possible to maximize productivity. These machines are expensive and may not offer the most energy-efficient solution.
We believe a shift to multiple, smaller autonomous vehicles could change the modern farming paradigm, providing the same capacity but with a smaller energy footprint. If these machines also autonomously monitor crop health and dispense nutrients, we can maintain or increase crop production while reducing environmental impacts.
This advance will require the development and integration of a number of elements, including mechanical systems, machine vision, satellite guidance, and onboard sensors — basically, the same issues facing the adoption of on-road and off-road autonomy.
Farm tractors traditionally perform many different operations, from tillage to applying fertilizer to pulling planters. Each operation has unique needs, and most proposed farm robotic platforms are purpose-built machines designed to handle a single operation. That approach allows manufacturers to address the technical dilemma, but it might not make business sense for farmers.
We’re looking to provide autonomous off-road vehicles with multi-operation flexibility by replicating the supervisory and decision-making role of the operator. This is a big challenge because equipment operators have to complete multiple complex operations simultaneously.
Our team is working on several solutions to the problem. We’re performing energy and techno-economical modeling to understand how multiple, smaller autonomous machines, rather than a single large machine, affect energy use and profitability. In addition, we’re working on machine vision to aid in automation and using simulation software to evaluate different technologies and algorithms. We’re also developing navigation algorithms for in-field operations.
Currently, we’re collaborating on a project sponsored by John Deere with Professor Gregory Shaver from Purdue’s School of Mechanical Engineering and Tony Vyn, the Henry A. Wallace Chair in Crop Sciences and Professor of Agronomy in the College of Agriculture. The goal is to develop an automated unloading strategy for combine harvesters; a provisional patent has been submitted.
Many challenges remain, but we are working to overcome them in order to guide industry toward big-picture solutions that are environmentally and economically sustainable. We don’t want automated machines to be cost-prohibitive for small and mid-sized farms. That would keep a portion of farms from realizing the benefits of automated machines and would not help sustainably increase food production on those acres. It also could cause profitability to decrease on this land, further accelerating the demise of small and mid-sized farms.
My hope is that automated machines will better help farms of all sizes feed a growing population in a way that is both environmentally and economically sustainable.
John T. Evans IV, PhD
Department of Agricultural and Biological Engineering
College of Engineering, Purdue University