ITU and FAO’s new report investigates AI’s positive impact on agriculture
By Erin Kalejs
According to the United Nations the world’s population will increase by 2 billion people by 2050, requiring a 60% increase in food productivity to feed them. In order to meet this global challenge, the agricultural sector is turning to artificial intelligence (AI). This is exactly what is explored in the recent “Digital Agriculture in Action” report published by the Food and Agriculture Organization of the United Nations (FAO) and International Telecommunication Union (ITU).
Why agriculture matters
The exponential advancement of AI technology offers a new opportunity for optimizing farming practices while simultaneously making them more equitable at a time when progress is critical to achieve an increase in food productivity. AI has the power to improve current agricultural applications by providing support services previously considered too resource-intensive, expensive, or unavailable and can reduce current operational costs by saving time and labor performed by farmers.
Fixing the challenges farmers are currently facing such as soil health and herbicide resistance is imperative. Without doing so, the chances of achieving other SDGs are slim because at least seven of the seventeen UN SDGs are, directly linked to agriculture: SDG 1 (end poverty); SDG 2 (zero hunger); SDG 8 (decent work and economic growth); SDG 12 (responsible consumption and production); SDG 13 (climate action); SDG 14 (aquatic life); and SDG 15 (life on the land). Additionally, agriculture has an indirect impact on global health (food quality being a disease risk factor), and gender equality (through its ability to increase access to resources and employment for women and children).
Therefore, AI technologies can help farmers increase crop yields, address the challenges of soil health and herbicide resistance and use resources more sustainably and efficiently to decrease the agricultural sector’s overall greenhouse gas emissions.
In an effort to collaboratively tackle global geospatial issues related to food security, AI For Good has also launched the new GeoAI Challenge to encourage new AI applications to help advance the United Nations Sustainable Development Goals (SDGs).
The challenge has been created in collaboration with the Food and Agriculture Organization of the United Nations (FAO) and the UN Open GIS Initiative to map cropland with satellite images and Google Earth with machine learning solutions.
Click here to register and learn more about this global challenge.
Zhongxin Chen, Senior IT Officer, and Head of Agro-informatics Unit, Digitalization and Informatics Division, at FAO explains what type of solutions will be chosen from the GeoAI challenge and how they will help drive sustainable agriculture.
“Cropland distribution information is fundamental to sustainable agriculture. There are obvious disagreements between various existing global remote-sensed land cover data products. With this Geo-AI challenge, novel and robust applications and algorithms using machine learning and AI will be selected from the participants all around the world, in the framework of UN OpenGIS initiative.”
Chen urges people to get involved with the Geo-AI challenge because participants “will have the opportunity to compete with the best researchers and programmers who are engaging in using AI to improve cropland mapping and agriculture monitoring with satellite remote sensing images. The finalists will be recognized by the leading agencies of agriculture geospatial IT applications in the world, which will benefit their future career.”
Andrea Manara, Senior System Analyst, at ITU Focal Point for Geospatial Information Management, elaborated on the benefits participants can gain from this challenge. “Participants will benefit from having the opportunity to define, provide inputs and shape the technologies related to GeoAI, network with experts and peers, enjoy a platform to gain hands-on experience related to GeoAI and possibility to gain global recognition in the form of prizes, appreciation, and publications of the results in the ITU News Magazine and ITU Journal on Future and Evolving Technologies.”
Key findings from the report
The report identified four broad areas of AI application in agriculture: crop, soil, and livestock monitoring, detection of pests and diseases, weather and temperature forecasting, and autonomous agricultural robots and farm equipment.
Key findings around AI included how an agricultural community could be taught to train machine learning algorithms. This would allow for machine learning models to be constantly updated in order to deal with changes in the environment such as the identification of new pests or new crops planted.
Another main finding from the report’s pilot project was that the cost of the digital farmhand should be decreased for smallholders to access it through a service. The essential challenge being to ensure that AI practices in agriculture do not cause any “significantly negative externalities and that their benefits are widely distributed and fairly shared among target communities.”
The report also predicts that within the next few years, the application of machine learning in various agricultural applications is expected to increase significantly as long as several challenges to its widespread deployment are fixed, especially among developing country farmers. These challenges include the expensive costs of the technologies, lack of standardization, lack of AI awareness among farmers and limited availability of historical data.
While the agricultural sector is likely to see further usage of AI, it is important that farmers are provided with up-to-date training to ensure technologies are used and continue to improve. Extensive testing and validation of emerging AI applications will be crucial since environmental factors that impact agriculture are not possible to regulate, unlike other industries where risk factors are easier to evaluate and predict.
Join the work of ITU and FAO in the Focus Group on “Artificial Intelligence (AI) and Internet of Things (IoT) for Digital Agriculture” (FG-AI4A). Register for the upcoming meeting here.
