Descriptive and Predictive Analytics for Performance Improvement of Inbound Logistics
During the pandemic, many firms realized the importance of the supply chain in their business. Firms realized areas where they were lagging and worked to improve their supply chain operations. The buzzwords like data analytics and machine learning became even more popular at such times because these techniques are effective in making processes more efficient. In this paper, we present practical applications of data analytics for improving inbound logistics. The aim of this paper to improve the On-Time & In-Full (OTIF) metric by using data-driven approaches to distill insights and make appropriate recommendations. To date, this effort has resulted in a 15% improvement in the on-time metric for a particular vendor by introducing a drop-trailer program. These recommendations have been thoroughly discussed with the management team, which led to building a machine learning model to predict the number of incoming pallets. Accurate prediction of incoming pallets enables managers of distribution centers to plan their labor as per the number of pallets. Adopting the machine learning model enabled a reduction in MAPE by 14%, which means we improved the accuracy by 14%. Future work can focus on scaling this to all vendors, which can be expected to contribute significantly to improving productivity and customer service.
Advisor: Dr. Vittaldas Prabhu
Penn State Only
Files are only accessible to users logged-in with a Penn State Access ID.
|Work Title||Descriptive and Predictive Analytics for Performance Improvement of Inbound Logistics|
|License||In Copyright (Rights Reserved)|
|Work Type||Research Paper|
|Deposited||June 20, 2021|
This resource is currently not in any collection.