Modelo predictivo del crecimiento microbiano durante la fermentación del cacao fino de aroma

Abstract

CONDITIONING OF PROJECT PUBLICATION. This project aims to develop a computational system to support the control of the fine aroma cocoa fermentation process, which is commonly carried out empirically in most producing farms. The hypothesis states that the integration of temperature sensors, a mobile application, and a predictive model implemented in software enables the estimation of the fermentation process behavior, allowing its monitoring and analysis without relying on laboratory analyses. The proposal is justified by the need to incorporate accessible technological tools that contribute to improving the competitiveness of Ecuadorian cocoa. The project was developed using an iterative methodology that integrated low-cost temperature sensors with Bluetooth Low Energy connectivity, a mobile application, and a web platform. Multiplatform technologies, local data storage for offline operation, and a statistical predictive model were implemented. The results showed that the system successfully recorded real-time temperature readings, generated process estimations, and synchronized data with a centralized platform. In conclusion, the project validates the feasibility of integrating hardware and software to improve the monitoring of the fine aroma cocoa fermentation process. Keywords: Mobile application, Predictive model, BLE sensors, Fermentation monitoring, Computational system