Advancing Canning Quality in Common Beans: An Integrated Farm‐to‐Can Framework Combining Breeding, Processing, and Artificial Intelligence

ABSTRACT Common beans ( Phaseolus vulgaris L.) are essential raw material for the canning industry. This article reviews recent advances in assessing canning quality and the integration of artificial intelligence (AI) into breeding methodologies aimed at developing genotypes with superior yield and canning‐quality traits. Cultivars destined for canning must consistently meet strict quality standards in addition to high agronomic performance. Conventional phenotypic quality parameters, such as washed drained weight, processing quality index, sensory properties, and texture, are central to predicting canning performance. However, their assessment is labor‐intensive, costly, and often limited to advanced filial generations, making early selection challenging. Recent progress in artificial intelligence, imaging, and data analysis provides new opportunities to evaluate canning traits at early stages of breeding, complementing conventional sensory and laboratory evaluations. These innovations enable breeders to optimize selection pipelines, reduce dependency on external facilities, and accelerate the release of superior cultivars. The review highlights the potential of AI coupled with nondestructive imaging to transform canning‐quality assessment by offering high‐throughput, cost‐effective, and scalable tools that improve prediction accuracy. Future directions include harmonizing evaluation protocols, developing cultivars that combine nutritional enrichment with drought tolerance and canning quality, and expanding genotype testing across multiple environments. The integration of AI with traditional breeding strategies offers a promising pathway to enhance both the efficiency and sustainability of dry bean improvement programs, ensuring cultivars that align with market requirements and consumer expectations.