Modelo predictivo para la elección de proveedores de transporte consolidado de carga (LTL) en Estados Unidos
No Thumbnail Available
Date
2025
Authors
Directors
Journal Title
Journal ISSN
Volume Title
Publisher
Institución Universitaria Pascual Bravo
Abstract
Este trabajo presenta el desarrollo de un modelo predictivo basado en técnicas de machine learning para evaluar la eficiencia de los transportistas que prestan el servicio de transporte terrestre LTL (Less-Than-Truckload, o carga consolidada). Utilizando datos históricos generados por una empresa intermediaria del sector, se construyó una herramienta capaz de predecir la probabilidad de que una empresa transportista cumpla con la recogida y entrega de un envío dentro de los plazos ofrecidos. El modelo considera variables como el área de servicio de origen y destino, características físicas de la orden (volumen, peso), servicios adicionales requeridos y otros factores operativos. Durante el proceso, se compararon y evaluaron diversos algoritmos de aprendizaje automático, identificando el enfoque más efectivo para la predicción de incumplimientos. La
herramienta desarrollada funciona como un sistema de soporte a la decisión para la selección y priorización de proveedores, minimizando retrasos logísticos y mejorando la satisfacción del cliente a través de decisiones informadas y fundamentadas en datos, validadas mediante simulaciones y análisis de desempeño real.
ABSTRACT This study presents the development of a predictive model based on machine learning techniques to evaluate the efficiency of carriers providing Less-Than-Truckload (LTL) freight services. Using historical operational data from a non-asset-based freight broker, the research builds a tool capable of estimating the probability that a carrier will complete both pickup and delivery within the promised timeframes. The model incorporates variables such as origin and destination service areas, physical shipment characteristics (volume, weight), additional required services, and other relevant operational factors. Several machine learning algorithms were compared and evaluated, identifying the most effective approach for predicting non-compliance. The resulting tool functions as a decision-support system for selecting and prioritizing carriers, reducing logistical delays and enhancing customer satisfaction through data-driven and validated decision-making processes, including practical simulations and comparisons with real-world performance.
ABSTRACT This study presents the development of a predictive model based on machine learning techniques to evaluate the efficiency of carriers providing Less-Than-Truckload (LTL) freight services. Using historical operational data from a non-asset-based freight broker, the research builds a tool capable of estimating the probability that a carrier will complete both pickup and delivery within the promised timeframes. The model incorporates variables such as origin and destination service areas, physical shipment characteristics (volume, weight), additional required services, and other relevant operational factors. Several machine learning algorithms were compared and evaluated, identifying the most effective approach for predicting non-compliance. The resulting tool functions as a decision-support system for selecting and prioritizing carriers, reducing logistical delays and enhancing customer satisfaction through data-driven and validated decision-making processes, including practical simulations and comparisons with real-world performance.
Description
Keywords
Machine learning, Transporte terrestre, Satisfacción del cliente, Eficiencia, Logística, Predicción de cumplimiento, Optimización logística, Less Than Truckload, Machine learning