From bean to bar 5.0: Artificial intelligence, sensing technologies and robotics across the cocoa–chocolate value chain

Background Cocoa and chocolate production face persistent challenges related to climate variability, pests and diseases, postharvest losses, quality inconsistency, traceability requirements and sustainability concerns. Advances in artificial intelligence (AI), machine learning (ML), sensing technologies, robotics and automation offer opportunities to improve productivity, quality management and operational efficiency across the cocoa–chocolate value chain. Scope and approach This review proposes an integrated bean-to-bar 5.0 framework that synthesizes recent applications of AI, ML, sensing systems, robotics and automation in cocoa farming, postharvest processing, quality and safety evaluation, chocolate product innovation, manufacturing, packaging and logistics. Studies involving computer vision, deep learning, electronic noses, spectroscopy, Internet of Things (IoT) platforms and robotic handling systems are analyzed with respect to datasets, algorithms, technology readiness and deployment constraints to identify current capabilities and future research needs. Key findings and conclusions AI-enabled vision, spectroscopic and sensor-based approaches show strong potential for cocoa pod disease detection, bean grading, and rapid, non-destructive cacao quality assessment, while AI-assisted inspection, 3D printing, robotic handling and automated packaging can enhance sorting accuracy, process consistency and operational efficiency in chocolate manufacturing. Yet most technologies remain at laboratory, pilot or prototype stages, constrained by narrow datasets, controlled testing conditions, fragmented interoperability and limited techno-economic or socio-technical evaluation. Future progress will require representative multimodal datasets, rigorous field and factory validation, interoperable digital ecosystems, responsible AI governance and human-centred automation strategies tailored to smallholder-dominated cocoa regions. Overall, AI, sensing and robotics offer considerable potential to support more efficient, traceable, resilient and sustainable cocoa–chocolate systems.