Green AI and energy efficient deep learning for sustainable and scalable precision agriculture

Precision agriculture represents a major advancement in farming by leveraging advanced technologies to improve efficiency, productivity, and sustainability. The use of deep learning an advanced form of artificial intelligence (AI) has transformed this field by providing accurate, data-driven insights for optimizing agricultural practices, such as monitoring crop health and predicting yields. However, challenges remain in implementing deep learning in precision agriculture. Data limitations such as incomplete, inconsistent, or sparsely sampled datasets, restricts robust model development and generalisation. Additionally, the complexity of deep learning architectures often leads to ‘black-box’ systems where decisions are produced without clear, understandable reasoning, making the model’s outputs difficult to interpret. This lack of transparency diminishes trust and limits the practical usability of these models in field-level decision-making. Additionally, these models often require significant computational power, raising concerns about energy consumption and sustainability. Green AI is an emerging approach to developing artificial intelligence systems with reduced environmental impact. This review synthesises recent advancements in leveraging AI for environmental benefits and in creating energy and resource-efficient AI, with particular emphasis on the application of deep learning in contemporary agriculture. This review also examines current deep learning applications in precision agriculture, identifies challenges related to data quality and model complexity, and emphasizes the need for lightweight models and energy-efficient hardware. It also highlights the importance of collaboration among technologists, agronomists, and farmers in overcoming these obstacles and fostering sustainable agricultural solutions. By addressing these issues, advancements in this area can enhance agricultural productivity while promoting environmental sustainability.