Digital Transformation of Monitoring in The Customs System Development of an App for Tracking Closed Vehicles at Land Border Crossings

By Ernelda NOVAKU, Aurora BINJAKU

The customs system in Albania faces significant challenges in monitoring sealed vehicles at land border points, where manual processes lead to delays, human errors, and a lack of transparency. These issues increase operational costs, reduce service efficiency, and hinder effective risk management due to the absence of advanced analytical tools. The research question addressed is: How can an application leveraging advanced technologies, such as Machine Learning (ML) and a Graphical User Interface (GUI), enhance the monitoring of sealed vehicles at Albanian land border points? This study tackles these gaps by developing a dedicated application designed to boost transparency, minimize operational errors, and provide risk analysis based on historical data. Drawing from international and local literature on successful customs digitization, the research identifies ML and GUI as key tools for efficiency and security. The developed application includes a user-friendly interface for customs authorities, featuring functionalities like vehicle registration, statistical reporting, and risk prediction, supported by technologies such as PHP, MySQL, HTML, CSS, Bootstrap 5, Font Awesome, JavaScript with Chart.js, Python with scikit-learn, and Flask. The methodology combines a qualitative approach, based on professional observations at customs branches, with technical development. The application successfully automates monitoring processes, achieving a risk prediction accuracy of 92.3% through ML, while the GUI simplifies data entry and reporting for customs officers, enhancing operational workflow. This digital transformation offers a scalable solution to improve customs efficiency, reduce manual errors, and align with the Albanian Customs Development Strategy 2021-2025. It sets a foundation for further technological integration, potentially impacting regional customs modernization efforts.

Key words: Customs Monitoring, Risk Prediction, Machine Learning, Random Forest Classifier, Digital Transformation, Web Application Development

 

https://doi.org/10.58944/bbec9633