The Future Of Technology In Agriculture

A trip to a modern American farm employing precision agriculture techniques would likely impress even the most discerning gadget enthusiasts: self-driving tractors, mobile command centers of tablet computers constantly beaming information to the cloud for analysis and real-time decision making, vast arrays of sensors monitoring minute changes in the field, and livestock outfitted with movement trackers that rival the hippest wearable tech. But precision agriculture is much more than just ‘cool’ gadgets that make farmers’ lives easier; these technologies are critical to achieving long-term food production security.

In recent years, new precision agriculture technologies have been a double-edged sword for farmers, as broader adoption of yield enhancing techniques combined with favorable weather has resulted in short-term supply outpacing demand (see Exhibit 1). This has placed downward pressure on crop prices. In our opinion, near-term supply imbalances should not be extrapolated into a long-term outlook for overproduction, though, as global food demand continues to swell and recent crop disruptions such as drought and disease remind us of how short-lived a state of oversupply can be. Due to growing populations and rising incomes, the global agriculture industry is expected to need to generate as much output in the next 40 years alone as it did in the last 10,000 years combined (McKinsey & Company). Declining arable land per capita means that production growth needs to come from yield enhancement, which is precision agriculture’s ultimate goal. 

Exhibit 1: Global production of cereal has outpaced demand, contributing to downward pressure on near-term crop prices  

The Need

Agricultural resource scarcity remains one of our key long-term investment themes given the trend increase in global food consumption and growing supply constraints on arable land. By 2030, the global population is projected to increase by 1.1 billion to an estimated 8.3 billion, with the number of people entering the global consumer class (defined as those with a daily income of between $10 and $100) expected to rise by roughly 3 billion over the same period¹. Among many other implications, this projected increase in the number and income level of global middle-class households means more demand for food — particularly for higher-protein diets via consumption of livestock and meat products.

According to projections by the Food and Agriculture Organization of the United Nations (FAO) and the Organisation for Economic Co-operation and Development, global demand for beef, veal, pork, poultry and mutton will rise from 312 million tons in 2015 to 357 million tons in 2025 — an increase of close to 15%. This rise in protein demand presents a substantial challenge to the agriculture industry, as every pound of meat requires two to seven pounds of grain feed. Taken together, the FAO estimates that global agricultural production will need to rise 70% between 2009 and 2050 to meet growing demand. At the same time, the amount of arable land per capita continues to decline (see Exhibit 2), while available acreage is only expected to grow 4% over the next 35 years, necessitating yield improvements in order to satisfy production demands. We believe that pressures from water scarcity, increasingly erratic weather patterns due to rising global temperatures, and land competition from biofuels, will only further constrain agricultural output. The sum of these challenges will require farmers to adopt more efficient techniques and new technologies that can increase yields at home, while helping to meet import demand from fast-growing markets around the world.

Exhibit 2: Declining global arable land per capita

The Promise of Precision

The ultimate goal of precision agriculture is to increase yields over the long-term. To achieve this, precision agriculture has three main subgoals: to do more with fewer inputs (including land, seeds, fertilizer, and water), to decrease output variability year to year, and to respond more quickly to changes in crops. Making progress towards these goals is an iterative process which increasingly requires advanced data collection and analytics to create a feedback loop for continuous improvement — meaning data from one year informs better decision making and product design in the next year. Traditional agriculture companies have played a major role thus far in developing “smart” equipment and inputs, but as computing needs become more intense, non-traditional agriculture players in big data, artificial intelligence, and machine learning are also finding a place on the farm. For example, large traditional IT companies have acquired an immense amount of weather data in order to help farmers better predict and cope with the capriciousness of nature. Further, high tech agriculture startups are sprouting up and receiving substantial backing from established players.

In addition to the overlay of big data and analytics, what are the other technological solutions for meeting the goals of doing more with fewer inputs, reducing variability year to year, and responding faster to crop changes? It is apt to start with the seed. Through the use of biotechnology, seed companies can insert genes to help plants combat insects, control weeds through herbicide resistance, and increase drought resistance, all to reduce sensitivity to uncontrollable variations in nature and maximize long-term yield growth. For example, some new drought resistant seeds, which use less water during drought conditions and still deliver strong yields when water is plentiful, have shown a 10 bushel/acre yield advantage versus non-drought resistant seeds. Seeds can also be optimized for specific regions and soil types. Farmers using these so called “stacked trait” seeds may require lower total chemical applications to raise their crops and maintain better soil quality over time, effectively reducing costs and enhancing sustainability.

In order to fully leverage the precision of seed technology, advancements in equipment must provide equal precision to the measurement of fields, the execution of planting, and the monitoring of crops. The aim is to plant the optimal amount of the optimal seed in the optimal soil with the optimal application of fertilizer, all while monitoring crops in great detail in order to ensure a robust harvest. With so many variables at play, an array of technologies is needed. 

Yield mapping lays the foundation and is a technique first used in the 1990’s, but continues to become increasingly exact. This is the process of combining GPS data with sensors built into harvesters that measure the amount harvested and the moisture level of the crop. Further sensors can be planted into the ground in order to measure nutrient content and moisture level at various depths. This information is used to generate a digital and visual representation of the field’s yield dispersion, which then informs decision making during the following year’s planting season. The yield map creates the opportunity to execute variable-rate seeding, a highly effective way to both reduce seed waste and maximize yields by planting more seeds in the most fertile locations. New precision planters use the yield map, along with real-time sensors mounted on planters, to automatically and optimally plant seeds with impressive exactitude. Precision fertilizer application and precision spraying to combat weeds follow similar tactics in order to ensure sections of the field receive exactly what they need to flourish, while also reducing input waste. Accuracy of execution is increased and strain on farmers is decreased by the integration of self-driving, GPS guided tractors. Self-driving tractors still require the presence of a human in the cab to prevent collisions, but manufacturers expect to achieve a fully autonomous version in the next decade. These technologies translate into direct savings for farmers, with one Auburn University study estimating an average 22% cost savings for farms who deployed these techniques (includes GPS-based guidance, Variable-Rate Application, and Automatic Section Control). Domestic adoption of these technologies has been broad, but there is still further room for penetration (see Exhibit 3), while there remains a substantial yield enhancement opportunity in deploying these technologies to less developed regions.

Monitoring the progress of crops is critical to maximizing yields. Aerial imaging options reduce the need for labor intensive crop inspection by foot and provide important insight into crop health, soil variation, and even pest and weed infestations. Three options are available to farmers to generate aerial images:

• The first is satellite imagery, which provides lower resolution images that are “good enough” but are susceptible to issues such as cloud cover that prevents timely image generation.
  
• The second is airplane imagery, which can provide higher quality images but for a higher price.
  
• The third is drone imagery, which has the promise to be both a low cost and high resolution option.

Exhibit 3: Adoption of precision agriculture technologies on soybean farms as of 2012

Further advances in autonomous flying (reducing the hassle of training and crashes for farmers), reduced flying regulations, cost reductions in sensor and camera technology, and better data analysis are needed in order to deliver on the promise of drones being the next leader of farm imagery.

Monitoring livestock is also providing new opportunities for technology adoption. Startups have developed wearable tech for cows that use movement detectors to track a cow’s activity (let’s call them early moo-vers). This information can be used to determine the health of the animal, giving an early warning for needed medical attention, as well as revealing ideal times for breeding.

With 40% of the world’s population expected to live in conditions of “severe water stress” by 2050 (Water Environment Federation), water scarcity also presents a considerable challenge for agricultural yield growth, creating the opportunity for precision irrigation to not only reduce water costs but also build more sustainable farms. Currently 70% of total freshwater use goes to agricultural applications, with farming livestock approximately ten times more water intensive than farming crops, according to the FAO. Precision irrigation applications have proven to not only reduce water consumption by 20-50%, but can also increase yields. In precision drip irrigation, data collected from moisture sensors in the field can be combined with large databases of weather data to automatically direct vast grids of drip tape to deliver tailored, small doses of water to individual plants. This can be very costly to implement, however. Therefore, for now, these advanced irrigation techniques are reserved for high value, water intensive crops like grapes and nuts. Adoption is likely to accelerate if water scarcity becomes a near-term constraint to farm profitability.

It’s not just large, industrial farms using these precision technologies. Small scale and organic farms are finding significant benefits in adopting new techniques, assisted by declining entry level prices for technology. For example, weed management is a significant challenge for all farmers, with 40% of crops lost to weeds annually according to the FAO, but organic farmers who have to eschew the use of herbicides for weed management have an even larger challenge. High resolution aerial imaging that uses multispectral analysis to quickly detect the presence of weeds is particularly useful when combined with GPS guided tractors that can accurately complete nonchemical weed management tasks like cultivation and flame weeding. Further goals of many organic farmers such as sustainability of soil (reducing erosion) and reducing consumption of inputs like water will be accomplished through the application of precision techniques.

Conclusion

The need for precision agriculture is clear: a 70% increase in food production to meet demand in 2050, up against arable land constraints and potential input shortages, creates the need for more efficient and productive agriculture solutions. This means leveraging advances in both traditional agriculture technology and innovations in high performance supercomputing and computational sciences. With arable land per capita continuing to decline, these technology advancements have the promise to make existing land more productive and production more sustainable. We continue to see farmland as an increasingly valuable long-term asset, as it continues to be a significant input constraint in the challenge to feed the world’s population in the long run.