Scholarlink

This paper presents the development, implementation, and validation of a dedicated software tool for integrating BIG DATA into macroscopic and multimodal transport models. The tool processes location data from mobile applications (Yanosik, Google Timeline) to generate origin-destination matrices classified by transport mode. Designed with modular architecture, it includes data preprocessing (Python), matrix generation, and direct integration with the PTV Visum environment (C#). Laboratory testing confirmed accurate mode detection and trip segmentation, while real-world validation using a transport model for the city of Bydgoszcz demonstrated a significant increase in model accuracy (R2 improved from 0.54 to 0.82). The solution proved efficient, scalable, and operable with high-volume data. This approach offers a cost-effective alternative to traditional surveys, supporting sustainable urban mobility planning and data-driven decision-making. The tool enables frequent model updates and represents a transferable framework for transport authorities seeking to integrate real-world mobility patterns into analytical workflows.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

A Tool for Integrating Big Data with Traffic Models: Design Process, Validation, and Implementation

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Marcin Jacek Kłos,
  • Łukasz Jaroszek,
  • Jakub Oziomek

摘要

This paper presents the development, implementation, and validation of a dedicated software tool for integrating BIG DATA into macroscopic and multimodal transport models. The tool processes location data from mobile applications (Yanosik, Google Timeline) to generate origin-destination matrices classified by transport mode. Designed with modular architecture, it includes data preprocessing (Python), matrix generation, and direct integration with the PTV Visum environment (C#). Laboratory testing confirmed accurate mode detection and trip segmentation, while real-world validation using a transport model for the city of Bydgoszcz demonstrated a significant increase in model accuracy (R2 improved from 0.54 to 0.82). The solution proved efficient, scalable, and operable with high-volume data. This approach offers a cost-effective alternative to traditional surveys, supporting sustainable urban mobility planning and data-driven decision-making. The tool enables frequent model updates and represents a transferable framework for transport authorities seeking to integrate real-world mobility patterns into analytical workflows.