Design of an automated prototype for the classification of blueberries using IoT technologies for the San Alejo crop (Guasca, Colombia)

Abstract

This study presents the design, construction, and validation of an automated prototype for blueberry size classification, implemented at the San Alejo farm (Guasca, Colombia). The objective was to replace the manual process with a more precise, efficient, and standardized classification system. The prototype integrates Internet of Things (IoT) technologies using an ESP32 microcontroller and remote control through the Blynk IoT application. Together with a perforated cylindrical stainless-steel structure, it enables the classification of berries into three categories: Jumbo, First, and Second. Methodologically, tools such as statistical analysis, the PDCA cycle, Ishikawa and Pareto diagrams, morphological material analysis, SolidWorks modeling, and validation using a Quality Function Deployment (QFD) matrix were applied. Results demonstrated a 91.5% improvement in operational capacity (from 47 to 90 lb/day), a 15.4% increase in classification accuracy (from 78% to 90%), and a 55.8% reduction in errors. Operator variability was eliminated, and processing time per batch was reduced by 55%. Final user validation yielded a score of 4.6 out of 5.0. The developed system represents a replicable, low-cost, and low-maintenance solution suitable for small-scale producers. It is concluded that the prototype is an effective tool for advancing toward more innovative, more sustainable, and technology-driven agriculture.