By Jan Keim

Artificial Intelligence (AI) and its subsets, such as Machine Learning (ML), are among the most heavily discussed technologies, both in academia and in business. AI is predicted to fundamentally disrupt many areas of business, from cybersecurity to supply chain management. In fact, most people interact with some sort of AI on a regular basis, for example when using a chatbot or filtering emails. A 2018 study conducted by PricewaterhouseCoopers (PwC) estimates a worldwide Gross Domestic Product (GDP) growth by up to 14% by 2030 as a result of accelerated use of AI, with China growing its GDP by up to 26%, compared to Northern Europe with an expected growth of 9.9% and North America with 14.5%. The Organisation for Economic Cooperation and Development (OECD) forecasts a strong impact of AI, especially on manufacturing, with the potential creation of entirely new industries. The digitalisation of industry, commonly referred to as “Industry 4.0”, is a prime example of rapid change driven by AI. Technologies such as the Internet of Things (IoT), big data analytics, cloud computing, augmented reality (AR) and 3D printing underpin this process and may lead to the transformation of manufacturing into “a single cyber-physical system in which digital technology, internet and production are merged in one”, as the European Parliament Research Service puts it.

While AI certainly has tremendous potential to transform manufacturing, one industry that is less talked about in this context is agriculture, even though the agriculture industry is among the most relevant to populations’ daily lives. There are various applications of AI in agriculture already. However, most of these applications are limited to bigger farms, currently neglecting smallholder farmers. Three areas of application of AI in agriculture are outlined below:

Precision Agriculture

Precision agriculture refers to the observation, measurement and responses to variability in crops, fields and animals. By using AI to increase crop yields and animal performance, precision agriculture can reduce costs and optimise processes. For example, Blue River Technology, a U.S. based start-up that has been acquired by tractor giant John Deere in 2017, uses computer vision and AI to precisely apply herbicides, instead of spraying entire fields. This approach not only saves money, it also decreases the environmental impact of plant protection products by eliminating up to 90% of the herbicide volumes.

Field Monitoring & Harvest Forecasting

Analysing the current condition of fields has long been a labour-intensive challenge for farmers. By analysing drone and satellite pictures using AI, farmers are now able to receive accurate data on their fields’ condition, vegetation issues and problem areas. For instance, IBM’s Watson Decision Platform for Agriculture provides farmers with tools that alert them should there be threats from weather forecasts, soil conditions, evapotranspiration rates, or crop stress. This helps farmers improve crop protection and optimise crop yields, for example. Ultimately, field monitoring helps farmers estimate their agricultural yield and plan security measures accordingly.

Process Automation

The United Nations (UN) predicts that by 2050, 68% of the world’s population will live in urban areas. This will lead to a decrease in labour force in rural areas. By automating processes, easier risk identification, faster decision making and remote operations, AI can significantly reduce the need for labour in the agriculture industry and decrease labour costs.

While there are many benefits to using AI in agriculture, there are a few challenges that have to be taken into consideration while moving towards a more automated and AI-enhanced future. Firstly, many applications of AI, or digitalisation more generally, can be cost intensive, require technological knowledge and demand special infrastructure. While big farms can largely benefit from AI applications, smallholder farmers may be left behind. Hence, ensuring that smallholder farmers equally benefit from the technological progress is a crucial task for politics and science alike. Secondly, the AI-supported automation of agricultural processes tends to benefit countries with large farmlands, such as the United States, Germany or France. Yet, many smaller countries are dependent on agriculture, such as Togo, Sierra Leone or Guinea-Bissau. So far, trade barriers have helped some smaller countries to protect their agricultural sector. However, the advancing globalisation and increasing international trade may exacerbate such policy, which could endanger smaller countries’ agriculture. Thirdly, the technological development in agriculture tends to benefit developed countries. High wages in developed countries create a strong incentive to automate processes and thereby save labour costs. In developing countries with lower wages, this incentive is weaker. According to a discussion paper by McKinsey & Company, the automation could bring back production from poorer countries to developed countries, which would likely increase the lead of developed countries over developing countries.

AI can help farmers tackle some of the most pressing problems they face today. Therefore, the steady adoption of AI will most likely continue and ultimately become mainstream. However, to ensure a level playing field, policymakers, scientists and innovators need to make sure that neither smallholder farmers nor entire developing countries are left behind.