Gemelo digital para el monitoreo predictivo de puentes en zonas remotas.

Abstract

Abstract: Ecuador faces critical challenges in its road infrastructure due to climatic and geographical factors, evidenced by recent bridge collapses. This thesis presents the development and validation of a digital twin designed for the predictive monitoring of bridges in remote areas. The solution integrates a low-cost IIoT architecture using edge nodes (ESP32 and MPU-9250 sensors) and Microsoft Azure cloud services. The methodology applied a hybrid approach: edge processing to filter in-situ data and artificial intelligence algorithms (Naive Bayes) in the cloud to classify the structural condition in real time. The system was validated through simulations and an experimental deployment on the UTPL pedestrian bridge. The results demonstrated 95% accuracy in anomaly detection and 100% stability in telemetry transmission under the secure MQTT protocol. The proposal confirms that it is possible to implement economically viable and technologically robust early warning systems to mitigate risks in hard-to-access infrastructures.