Artificial intelligence applications in agriculture continue to grow, driven by the increased demands of precision farming.
This growth is due to increasing demand for agriculture produce, real-time livestock monitoring, and the need for enhanced decision making to optimize farm management.
Other factors contributing to the rising interest in precision farming solutions are growing food demand and government assistance to farmers.
India, for instance, is witnessing significant growth in applications of AI in agriculture due to its government’s effort to promote the use of farm analytics tools among farmers.
Precision farming uses AI technologies to increase crop and livestock yield and production, monitor crop growing conditions, monitor health of individual animals, and improve a wide range of agricultural practices throughout the supply chain.
These AI technologies operate by combining large volumes of data with intelligent and iterative algorithms. Such technologies can recognize patterns, predict future outcomes, and recommend or make decisions using historical data. They can process data in various forms such as text, images, videos, and sounds. But, their performance depends on the availability of large and high quality data.
Farms collect a large volume of data these days and this data is expected to grow exponentially in the near future. This growth is mainly driven by the increasing use of sensing and monitoring devices, control systems, satellite and global positioning systems, and other smart technologies such as smartphones. Enhanced broadband connectivity in rural areas also contributes to farm data growth.
AI technologies embedded in precision farming solutions help farmers improve the accuracy and productivity of a variety of farming practices.
Farmers can create and use models to forecast weather patterns and seasonal changes in the environment to boost the growth of high-yield crops. AI-based sensors can be used to identify weeds and disease and apply chemicals only in the areas that are needed to control them. Furthermore, images and videos collected by satellites and unmanned drones are analyzed to understand soil conditions over time and enhance decisions on the outlook on crop yield and production.
In addition, these technologies assist with early detection of disease outbreaks in animals and even recommend prevention strategies. Motion sensors combined with AI algorithms are capable of monitoring animals’ real-world behavior, such as eating, chewing, walking, detecting abnormal behavior of the individual animals and then providing farmers with information.
These technologies can predict the severity of disease in an animal in advance and recommend treatments to improve their welfare.
AI-based precision farming platforms can also tackle labour and skill challenges. Robots and automation can reduce the need for seasonal workers and augment human labour by harvesting, planting crops and eliminating weeds.
Another application of these technologies includes the use of chatbots which has recently gained popularity among farmers. Farm chatbots, which are capable of processing a large amount of data collected from different sources and through communication with farmers to intelligently aggregate and analyze the information in real-time and assist farmers in their decision making.
Despite the growing need and potential advantages, adoption of AI technologies in farming has been slow.
There is still a long road ahead until we can see full-scale automation of agricultural practices. This is due to a number of factors including the inherent complexity of agricultural production systems, cost of launching and maintaining technologies, limited availability of suitable technology, and lack of legal and farming best practices to guide the establishment of new technologies.
To overcome these hurdles, access to infrastructure for collecting and processing large amounts of heterogeneous data is essential. Embedded in the infrastructure, effective data management systems are required to ensure data quality and standardized data formats that are ready for processing.
In addition, various aspects of data and AI system governance should be addressed. Legal issues such as data ownership require careful attention to gain trust of the farmers and other actors in the supply chain.
Novel business models and concepts are needed for value co-creation and to encourage farmers to share data. New technology solutions and policies are needed to protect farm and sensitive data throughout the AI system development lifecycle. Standardized algorithms, systems, data and protocols are needed for comprehensive and end-to-end automation. A high level of standardization enables reuse of resources to their fullest capacity and enhances usability.
Limited adoption of AI-based technology in farms is also linked to the lack of availability of simple solutions that are easy to use and fit in the farmers’ day to day practices without extensive effort.
It is essential to test prototype solutions in a large-scale on-farm trial to evaluate their constraints and improve their usability and performance. Lastly, it is essential to co-create farm technology solutions with farmers. This may motivate a large number of farmers to participate in the implementation and testing of these technologies, and in turn enhance the trust and adoption of AI-based solutions afterward.
There is no doubt that AI technologies will enable farms to work more efficiently. Future farms will operate with fewer workers and will be more sustainable and responsible. We just need to make sure that farmers, scientists, technologists and governments cooperate and strategically invest towards this important goal.
Rozita Dara is an associate professor in computer science at the University of Guelph.