<p>Challenges such as a downward trend in cultivation and post-harvest losses lead to increased gap in cocoa bean supply and demand. This review deals with the recent AI models used in farming, processing, and supply chain of cocoa beans. Farming models viz. XAI-CROP, Random Forest, and Gradient Boosting can detect cocoa diseases, recommend appropriate pesticides, enable targeted crop spraying, count the number of pods on cocoa trees, and indicate cocoa pod ripeness. Processing models involving AI viz. Artificial Neural Network, Bootstrap Forest fermentation, and Particle Swarm Optimisation were explored for their efficiency in technological steps viz<i>.</i> drying, roasting, conching, and tempering to obtain high-quality chocolates. The supply chain models used AI such as Decision Tree, Multi-level Perception and Long Short-Term memory for cold storage, traceability, and deforestation prediction. AI can thus be used to standardise the quality of produce by optimal resource utilisation leading to minimal impact on the environment.</p> Graphical abstract <p></p>

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

AI technologies shaping the future of the cocoa industry from farm to fork: a comprehensive review

  • Hemasri Senthil,
  • Madhura Janve

摘要

Challenges such as a downward trend in cultivation and post-harvest losses lead to increased gap in cocoa bean supply and demand. This review deals with the recent AI models used in farming, processing, and supply chain of cocoa beans. Farming models viz. XAI-CROP, Random Forest, and Gradient Boosting can detect cocoa diseases, recommend appropriate pesticides, enable targeted crop spraying, count the number of pods on cocoa trees, and indicate cocoa pod ripeness. Processing models involving AI viz. Artificial Neural Network, Bootstrap Forest fermentation, and Particle Swarm Optimisation were explored for their efficiency in technological steps viz. drying, roasting, conching, and tempering to obtain high-quality chocolates. The supply chain models used AI such as Decision Tree, Multi-level Perception and Long Short-Term memory for cold storage, traceability, and deforestation prediction. AI can thus be used to standardise the quality of produce by optimal resource utilisation leading to minimal impact on the environment.

Graphical abstract