Transforming weed management in sustainable agriculture with artificial intelligence: A systematic literature review towards weed identification and deep learning

In the face of increasing agricultural demands and environmental concerns, the effective management of weeds presents a pressing challenge in modern agriculture. Weeds not only compete with crops for resources but also pose threats to food safety and agricultural sustainability through the indiscriminate use of herbicides, which can lead to environmental contamination and herbicide-resistant weed populations. Artificial Intelligence (AI) has ushered in a paradigm shift in agriculture, particularly in the domain of weed management. AI's utilization in this domain extends beyond mere innovation, offering precise and eco-friendly solutions for the identification and control of weeds, thereby addressing critical agricultural challenges. This article aims to examine the application of AI in weed management in the context of weed detection and the increasing impact of deep learning techniques in the agricultural sector. Through an assessment of research articles, this study identifies critical factors influencing the adoption and implementation of AI in weed management. These criteria encompass factors of AI adoption (food safety, increased effectiveness, and eco-friendliness through herbicides reduction), AI implementation factors (capture technology, training datasets, AI models, and outcomes and accuracy), ancillary technologies (IoT, UAV, field robots, and herbicides), and the related impact of AI methods adoption (economic, social, technological, and environmental). Of the 5821 documents found, 99 full-text articles were assessed, and 68 were included in this study. The review highlights AI's role in enhancing food safety by reducing herbicide residues, increasing effectiveness in weed control strategies, and promoting ecofriendliness through judicious herbicide use. It underscores the importance of capture technology, training datasets, AI models, and accuracy metrics in AI implementation, emphasizing their synergy in revolutionizing weed management practices. Ancillary technologies, such as IoT, UAVs, field robots, and AI-enhanced herbicides, complement AI's capabilities, offering holistic and data-driven approaches to weed control. Additionally, the adoption of AI methods influences economic, social, technological, and environmental dimensions of agriculture. Last but not least, digital literacy emerges as a crucial enabler, empowering stakeholders to navigate AI technologies effectively and contribute to the sustainable transformation of weed management practices in agriculture.