An Agentic artificial intelligence framework for Agricultural Technologies

Human civilisation is dependent on agriculture. With population growth and extreme climate conditions, various countries are suffering from constrained resources while facing the challenge of optimising production efficiency to meet the increasing demand. Hence, researchers, governments, and organisations are exploring different avenues of machine learning to solve some of the issues associated with agricultural sustainability. Despite a strong potential, scarcity of data and lack of integration of agricultural sensors and systems, the adoption of AI/ML paradigms is currently limited, making the future smart agricultural infrastructure counterproductive. ;An agentic AI system is an autonomous artifical system which can take intelligent decision without no or minimal human supervison. Integration of such systems in agriculture can benefit the agricultural production. The agentic AI overcomes the traditional machine learning limitations by making it more adaptive, goal driven and enhances the decision making capabilities. Hence, the use of such systems in agriculture will consider a holistic dynamic environment to make decisions autonomously improving the production. For instance, using Agricultural IoT sensors and drones to gather real-time environment status and use them to take action autonomously. The sensors and drone images will allow the intelligent system to detect the disease early and nutrient deficiencies way ahead of time. In addition, agricultural sensor data can help us monitor the soil health, optimise irrigation and weather patterns. Such advanced predictions will help make the agriculture sustainable. ;This PhD research aims to explore agentic AI-based systems collectively working to address the aforementioned data and system bottlenecks in resource-constrained challenging environments. The project will explore research gaps in traditional machine learnings in agriculture. In addition, we will focus on exploring the current limitations in agentic AI in agriculture. Finally, the project will focus on developing state-of-the-art methodologies, techniques, and proof-of-concept (PoC) to help ensure the resilience and sustainability of future agriculture.