Building a Banana Sorter AI Machine (www.bvaibhav.info)

A new project exploring the development of an AI-based banana sorting machine has emerged, highlighting the role of artificial intelligence in automating routine agricultural tasks. Instead of relying on human workers to identify and separate ripe bananas from overripe ones, the proposed system employs various vision models, including both open-source and cloud-based options, to classify bananas using image recognition. The machine uses a conveyor belt integrated with cameras to make binary decisions: whether a banana should be placed in the "TRASH" or "BASKET" based on its ripeness, evaluated through a series of test images. This initiative is significant for the AI/ML community as it delves into the practical applications of machine learning models in real-world contexts, such as agriculture. Although offline models like Qwen2.5-VL and Gemma 3 demonstrated high accuracy in rejecting overripe bananas, the results revealed that advanced models like GPT-6.1 struggled with consistency, occasionally misclassifying acceptable bananas. The project raises critical questions about the reliability of AI in decision-making roles—especially in nuanced scenarios—pointing to the necessity for ongoing refinement in the design of AI systems that aim to replace human judgment in agricultural practices.