Enhancing Logistics Demand Prediction Accuracy Through Client–Vendor Provider Hyperconnected Data Ensembles
Author(s)
Advisor(s)
Editor(s)
Collections
Supplementary to:
Abstract
In this paper, we present a hyperconnected data ensemble framework under the Physical Internet (PI) paradigm. As the current world is more volatile, uncertain, complex and ambiguous than before, it is well known that today’s logistics and supply chain management (LSCM) is facing greater difficulties and the idea of PI is introduced as a solution to the logistics sustainability grand challenge. PI aims to achieve significant logistics system efficiency and sustainability improvement through universal interconnectivity and smart open coordination. Meanwhile, PI will facilitate data sharing among supply chain parties. Instead of traditional data sharing by integrating datasets in common cases, we suppose that the hyperconnected data ensembles require as minimal data as possible. With the utilization of aggregated information, such as the overall activity forecast, the hyperconnected data ensembles can enhance the accuracy of the logistic demand prediction while preserving data privacy. A framework of logistics demand prediction with hyperconnected data ensembles is established and results of some experiments conducted based on the framework support our hypothesis that the demand prediction accuracy can be increased by integrating the forecast data that clients are willing to provide.
Sponsor
Date
2021-06
Extent
Resource Type
Text
Resource Subtype
Paper