Multimodal and Explainable AI System Design for Agricultural Sustainability: A Comprehensive Literature Review and Synthesis

Sustainability in agriculture is the key problem due to environmental issues, soil degradation and different crop diseases. Artificial Intelligence (AI) offer the solutions for precision agriculture with the utilization of multimodal systems. However, AI systems, especially utilizing deep learning are black boxes which will question the trust of diagnosis and predictions of the implemented system. This work explores a literature review of multimodal and explainable AI (XAI) systems for sustainable agriculture. It analyses multimodal fusion methodologies, deep learning, and explainable solutions that could be employed to assess sustainability impacts in agriculture. The study analyses the research prospects such as adaptive fusion, edge optimization, explainability evaluation.