Juan Diego Erazo

Prof. Juan Diego Erazo


Research Focus & Metrics

Academic Career Length

4Years (2023-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
7 April 2025 in Computation
Verified
Journal Article
Open access
PDF file
WCWilson Chango
Wilson Chango
PBPamela Buñay-Guisñan
Pamela Buñay-Guisñan
Juan Diego ErazoJuan Diego Erazo
Juan Diego Erazo
Pedro Aguilar-EncarnacionPedro Aguilar-Encarnacion
Pedro Aguilar-Encarnacion
JSJaime Sayago-Heredia
Jaime Sayago-Heredia
Geovanny Euclides Silva PeñafielGeovanny Euclides Silva Peñafiel
Geovanny Euclides Silva Peñafiel
The purpose of this study lies in developing a comparison of neural network-based models for vehicular congestion prediction, with the aim of improving urban mobility and mitigating the negative effects associated with traffic, such as accidents and congestion. This research focuses on evaluating the effectiveness of different neural network architectures, specifically Transformer and LSTM, in order to achieve accurate and reliable predictions of vehicular congestion. To carry out this research, a rigorous methodology was employed that included a systematic literature review based on the PRISMA methodology, which allowed for the identification and synthesis of the most relevant advances in the field. Likewise, the Design Science Research (DSR) methodology was applied to guide the development and validation of the models, and the CRISP-DM (Cross-Industry Standard Process for Data Mining) methodology was used to structure the process, from understanding the problem to implementing the solutions. The dataset used in this study included key variables related to traffic, such as vehicle speed, vehicular flow, and weather conditions. These variables were processed and normalized to train and evaluate various neural network architectures, highlighting LSTM and Transformer networks. The results obtained demonstrated that the LSTM-based model outperformed the Transformer model in the task of congestion prediction. Specifically, the LSTM model achieved an accuracy of 0.9463, with additional metrics such as a loss of 0.21, an accuracy of 0.93, a precision of 0.29, a recall of 0.71, an F1-score of 0.42, an MSE of 0.07, and an RMSE of 0.26. In conclusion, this study demonstrates that the LSTM-based model is highly effective for predicting vehicular congestion, surpassing other architectures such as Transformer. The integration of this model into a simulation environment showed that real-time traffic information can significantly improve urban mobility management. These findings support the utility of neural network architectures in sustainable urban planning and intelligent traffic management, opening new perspectives for future research in this field.

29 January 2025 in Preprints.org
Verified
Preprint
Open access
PDF file
WCWilson Chango
Wilson Chango
PBPamela Buñay-Guisñan
Pamela Buñay-Guisñan
Juan Diego ErazoJuan Diego Erazo
Juan Diego Erazo
Pedro Aguilar-EncarnacionPedro Aguilar-Encarnacion
Pedro Aguilar-Encarnacion
JSJaime Sayago-Heredia
Jaime Sayago-Heredia
Recent research has demonstrated a strong correlation between the exponential growth of the transportation sector and the increase in urban phenomena such as traffic accidents and congestion. In this context, accurate congestion prediction has become a primary objective to improve urban mobility and mitigate the negative effects associated with traffic.The study proposes a machine learning-based classification model to predict vehicular congestion. Using a dataset that includes variables such as speed, traffic flow, and weather conditions, various classification algorithms were trained and evaluated. The results obtained indicate that models based on Random Forest offer superior performance in the task of congestion prediction.To evaluate the impact of these predictions on urban mobility, the model was integrated into a vehicular ad-hoc network (VANET) simulation environment. The simulation results demonstrated that providing real-time traffic information based on model predictions allows drivers to make more informed decisions, consequently reducing travel times and emissions.

21 May 2023 in Journal of Namibian Studies : History Politics Culture
Verified
Journal Article
Open access
PDF file
Juan Diego ErazoJuan Diego Erazo
Juan Diego Erazo
BRBryan Guillermo Guananga Rodríguez
Bryan Guillermo Guananga Rodríguez
EVEUGENIA MERCEDES NARANJO VARGAS
EUGENIA MERCEDES NARANJO VARGAS
RPRaúl Gregorio Martínez Pérez
Raúl Gregorio Martínez Pérez
A documentary review was carried out on the production and publication of research papers related to the study of the variables Legal Standards and Sustainable Manufacturing. The purpose of the bibliometric analysis proposed in this document was to know the main characteristics of the volume of publications registered in the Scopus database during the period 2018-2022, achieving the identification of 107 publications. The information provided by this platform was organized through graphs and figures categorizing the information by Year of Publication, Country of Origin, Area of Knowledge and Type of Publication. Once these characteristics have been described, the position of different authors towards the proposed theme is referenced through a qualitative analysis. Among the main findings made through this research, it is found that the United Kingdom, with 19 publications, was the country with the highest scientific production registered in the name of authors affiliated with institutions in that country. The Area of Knowledge that made the greatest contribution to the construction of bibliographic material referring to Legal Standards and Sustainable Manufacturing was Environmental Science with 63 published documents, and the Type of Publication most used during the period indicated above was the Journal Article with 74 documents of the total scientific production.

5 January 2026 in Esprint Investigación
Verified
Journal Article
Open access
PDF file
Juan Diego ErazoJuan Diego Erazo
Juan Diego Erazo
VGVanessa Lorena Valverde González
Vanessa Lorena Valverde González
RPRaúl Gregorio Martínez Pérez
Raúl Gregorio Martínez Pérez
JJJavier Edmundo Albuja Jácome
Javier Edmundo Albuja Jácome
El presente estudio aborda el modelado y pronóstico de la demanda mensual en un entorno de manufactura, en el marco del Plan de Planificación y Control de la Producción. La unidad de análisis correspondió a una serie univariada de demanda operativa registrada durante 24 meses, representativa de un proceso productivo con comportamiento estable y periodicidad regular. La investigación adoptó una perspectiva cuantitativa y aplicada, con un diseño longitudinal no experimental, centrado en el examen estructural de la secuencia temporal y la validación estadística de un modelo parsimonioso. Se implementaron modelos ARIMA como herramienta metodológica adecuada para entornos fabriles con enfoque de medio plazo. El manejo de la serie incluyó la verificación de la calidad de los datos, la evaluación de la estacionalidad y la ejecución de pruebas de estacionariedad (ADF, KPSS y OCSB), además del análisis de las funciones de autocorrelación (ACF) y autocorrelación parcial (PACF) para identificar los grados de dependencia temporal. Los resultados evidenciaron que no era necesario aplicar diferenciación, lo que condujo a la selección de un modelo ARMA(0,1) estimado mediante máxima verosimilitud. Este modelo mostró un desempeño superior frente al modelo Naïve, con una mejora del 45 % en el MAPE, cumpliendo además con los supuestos de independencia, normalidad y homocedasticidad de los residuos. Finalmente, el modelo fue reestimado con la serie completa para generar una previsión operativa de seis meses con intervalos de confianza del 80 % y 95 %, útil para la definición de puntos de pedido y estrategias de inventario. Los resultados evidencian que un modelo ARMA parsimonioso proporciona precisión, trazabilidad y gobernanza operativa, fortaleciendo la planificación de la producción y la estabilidad de la cadena de suministro.


2 July 2026 in Esprint Investigación
Verified
Journal Article
Open access
PDF file
JPJohana Beatriz Samaniego Palacios
Johana Beatriz Samaniego Palacios
Juan Diego ErazoJuan Diego Erazo
Juan Diego Erazo
BRByron Daniel Erazo Rodríguez
Byron Daniel Erazo Rodríguez
WYWilian Andres Nuñez Yanez
Wilian Andres Nuñez Yanez
Los gemelos digitales (GD) se han consolidado como una de las tecnologías habilitadoras más relevantes de la Industria 4.0 para fortalecer la supervisión, el monitoreo y el control automático de procesos industriales. El presente artículo tuvo como objetivo analizar la evidencia científica disponible sobre la aplicación de gemelos digitales en la supervisión y el control automático de procesos industriales mediante una revisión sistemática basada en la metodología PRISMA 2020. Se realizó una búsqueda bibliográfica en Scopus, Web of Science, IEEE Xplore, ScienceDirect, MDPI y SpringerLink, identificándose inicialmente 120 registros; tras el proceso de cribado y evaluación de elegibilidad, se seleccionaron 25 estudios de acceso abierto publicados en revistas científicas indexadas. Los resultados evidenciaron que los gemelos digitales se aplican principalmente en la supervisión remota mediante sistemas SCADA, la integración con sistemas de control automático, el diagnóstico de fallas y mantenimiento predictivo, la ciberseguridad industrial y la formación de operadores mediante realidad aumentada y realidad virtual. Asimismo, se identificaron aplicaciones en sectores como manufactura discreta, energía hidroeléctrica y fotovoltaica, petróleo y gas, tratamiento de agua y robótica industrial, confirmando su carácter transversal dentro de los ecosistemas industriales inteligentes. Se concluye que los gemelos digitales constituyen una herramienta estratégica para mejorar la eficiencia operacional, la capacidad predictiva y la toma de decisiones industriales; sin embargo, su adopción enfrenta desafíos relacionados con la interoperabilidad, la estandarización de arquitecturas, la ciberseguridad y los costos de implementación.