Address management improvement - KLG Venlo

Project description

The goal of this project is to analyze, clean up, and optimize address management at KLG by identifying and merging duplicate addresses. This will allow KLG Venlo to maintain accurate, standardized address records that improve logistics planning and operational workflows.

Therefore, the main challenge of the project lies in answering the question: How can KLG Venlo optimize its address management system to eliminate duplicate entries, improve data consistency, and enhance operational efficiency?

Context

KLG Venlo is a logistics company specializing in international transport and supply chain solutions. With a strong presence in the Netherlands, KLG Venlo handles around 250,000 shipments and manages 500,000 loading and unloading locations. The company is committed to optimizing logistics operations by leveraging technology, data management, and process efficiency to deliver high-quality services to its clients.

KLG Venlo is experiencing inefficiencies in address management due to duplicate and inconsistent address entries. Many addresses are recorded with slight variations in spelling or formatting, leading to data pollution and incorrect address codes. This inconsistency creates challenges in logistics planning, particularly when choosing the right address for specific orders. These issues not only affect operational efficiency but also cause delays and errors in the delivery process.

Results

The Python-based fuzzy matching script effectively identifies and merges duplicate entries by comparing company names and addresses using similarity scoring, while strictly matching house numbers. The testing results show a reduction of ~73% in row number, while maintaining accuracy at 100%, proven to efficiently deal with duplicates. This allows for a cleaner and more reliable address database with measurable improvements.

In addition, restructuring the TMS address input fields by separating street names, house numbers, and additional details will reduce the chance of future errors and create a more consistent and user-friendly data entry process. When combined with a built-in fuzzy matching mechanism at the point of entry, the system can automatically detect potential duplicates before they are saved, ensuring better data quality from the start.

To fully realize the benefits of this solution, it is recommended that KLG invests in both technical implementation and user adoption. This includes integrating the fuzzy matching logic into the TMS, providing clear guidance to users, and offering training and support during the transition.

By improving data structure and cleaning existing records, KLG can significantly increase the accuracy of delivery information, reduce manual corrections, and streamline operational planning across its logistics network.

About the project group

All members are ICT & Business students. Three team members studied at Fontys Eindhoven (course-based), and one from Fontys Tilburg (demand-based). We worked in 2 week sprints, using Scrum. Physical meetings happened 3 times a week, mostly to discuss progress and next steps. We tried approaching the project by working individually on deliverables and comparing outcomes and results, as well as choosing the most suitable option/ combining each other's work. Some documents (such as Project Plan), were divided among group members from the start.