A New Agricultural Revolution, Courtesy Of The Internet Of Things And Machine Learning

Anja Strothkämper

How important is agriculture to society? Important enough that the first human civilizations rose as a result of the agricultural revolution. Growing food from the soil allowed us to avoid famine while granting us leisure time to pursue art and science.

In other words, agriculture is one of the most fundamental human activities. As long as we’ve pursued it, we’ve tried to master it. Better techniques meant greater yields. This, in turn, kept humans happier and healthier – and helped birth modern society as we know it.

There’s only one hitch in this success story, however. As our farming capacity has expanded, usage of resources such as land, fertilizer, and water have grown exponentially. Environmental pressures from modern farming techniques have stressed our natural landscapes. Still, by some estimates, worldwide food production will need to increase 70% by 2050 to keep up with global demand.

With global populations rising, it falls to technology to make farming processes more efficient and keep up with the growing demand. Fortunately, the combination of more data from the Internet of agricultural things and new machine learning capabilities can contribute a crucial part.

What is the Internet of Things?

For the uninitiated, the Internet of Things (IoT) simply refers to a world of connected devices. These devices are networked and equipped with sensors and software that allow the collection, analysis, and exchange of data.

The IoT extends the benefits of such connectivity outside the realm of computers and smartphones. Thermostats, vehicles, kitchen appliances, and a vast assortment of other objects can all be connected and networked via embedded technology. The number of connected devices is staggering: It’s estimated that there are 12 billion such devices in the world today. IDC has estimated that, by 2020, there will be 26 connected devices for every person on Earth. Adoption in the agriculture space is also rising quickly, with the total number of connected devices expected to grow from 30 million in 2015 to 75 million in 2020.

How machine learning adds to this equation

Data’s value comes when it is used to optimize processes, interactions, or create new business models. In agribusiness, all interactions with farmers and farming processes become more and more data driven. In a lot of cases, agronomists and industry experts can get better insights through this data, helping them to make better, more precise decisions or give better, more individualized recommendations to farmers. The critical element is analytical tools that are simple to use yet powerful to provide the right information at the right time.

Now machine learning becomes helpful for scalability and automation. It becomes helpful to learn patterns and extract information from large amounts of data, structured and unstructured. The potential becomes clear when we look at specific use cases.

How oil palm plantations become digital

Palm oil is one of the most efficient crops for the production of oils and fats. One hectare of palm is able to produce up to 10 times more oil than other leading oilseed crops. Indonesia and Malaysia produce about 85% of the world’s palm oil, which is used in a wide variety of products such as food, bioenergy, and cosmetics.

Despite its efficiency, palm oil suffers from reputation challenges around deforestation, responsible use of input products, and social standards. Industry leaders in this sector are working hard to promote sustainable practices in this segment, striving for full transparency and traceability along the palm oil value chain.

In addition, there are strong efforts to increase efficiency. The potentials are enormous: Average productivity worldwide has stagnated around 3 tons of oil per hectare per year. However, efficient producers can already achieve yields of 8 tons of oil per hectare per year, peak oil yields of 12 tons per hectare per year have been achieved in small plantations, and maximum theoretical yields are calculated with simulation models at 18.5 tons of oil per hectare per year. Higher yields can reduce the required amount of arable land and decrease the threat of expansion into valuable rainforests.

To achieve this mission, it is crucial to capture as much valuable and granular data as possible at the source. This requires an integrated end-to-end process, from data capture to driving optimization and automation of operations at scale.

Where this becomes clear is when we want to identify each individual tree on a plantation that can span thousands of hectares with millions of trees. By using drone imagery and applying image analysis and machine learning, we can recognize each tree and create a digital twin in its exact position.

Individual features, such as leaf cover area or color distribution, can be extracted from the drone images and related for each digital twin. Also, additional data points from satellite imagery, weather stations, or soil sampling can be related to each tree. Based on digital models for each tree, industry leaders have the means to build models for how each tree should be fertilized, irrigated, and harvested. Similarly, disease risks, nutrient deficiencies, and yield can be modeled.

At the same time, progress is being made on autonomous farm machinery and robots that can automate plantation tasks. The opportunity to optimize the planned tasks per tree to ensure that each individual palm tree thrives lays in front of us.

The takeaway

As technology matures, agricultural processes will become more efficient and precise. Higher efficiency helps reduce required resources. Keeping in mind that agricultural production uses around 55% of non-forest land, 80% of total freshwater, and 30% of fossil fuels, this is a huge lever for sustainability.

Digital capabilities, algorithms, and machine learning models are a key differentiator and a competitive advantage in the era of digital farming. With the right platform, solutions, and capabilities, farmers can digitize their existing expertise in farming processes, products, and practices, be it to enable data-driven farmer collaboration, better product recommendations, precision farming services, or new business models.

Learn how to bring new technologies and services together to power digital transformation by downloading The IoT Imperative for Consumer Industries. Explore how to bring Industry 4.0 insights into your business today by reading Industry 4.0: What’s Next?

Anja Strothkämper

About Anja Strothkämper

Anja Strothkämper is the Vice President for Agribusiness and Commodity Management within the Industry Cloud organization at SAP. Anja’s focus is to drive SAP’s digital strategy as well as solution portfolio and roadmap for SAP Agribusiness and Commodity Management solutions. She is passionate about customer co-innovation and thought leadership in the digital era.