Scholarlink

<p>The rapid advancement of new technologies and the widespread availability of data have enabled new business paradigms in urban logistics and transportation. Although optimization&#xa0;theories and algorithms are well-established, a persistent gap remains between these theoretical foundations and the practical skills required to formulate models and implement effective, data-driven solutions. This positioning paper aims to bridge this gap by introducing optimization modeling frameworks and solution methodologies tailored to urban logistics and transportation&#xa0;applications.&#xa0;The paper targets researchers, students, and practitioners in non-business disciplines such as civil engineering, industrial engineering, economics, and computer science. We present several widely used data-driven optimization paradigms, including deterministic optimization with sensitivity analysis, two-stage stochastic programming, integrated simulation-optimization, and Markov decision process, and&#xa0;discuss how mathematical programming, statistical methods, simulation, and machine learning can support decision making. We then provide a systematic classification of urban transportation applications based on key decision&#xa0;characteristics,&#xa0;such as whether decisions are&#xa0;static or dynamic, deterministic or stochastic, and whether they involve interactions among multiple decision makers. Building on this classification, we propose a practical roadmap to guide the selection and implementation of&#xa0;appropriate optimization approaches. Illustrative applications, demonstrate the relevance and versatility of the proposed frameworks.</p>

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

Data-driven optimization for modern urban logistics and transportation: modeling framework and applications

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年3月6日

查看全文
  • Haitao Li

摘要

The rapid advancement of new technologies and the widespread availability of data have enabled new business paradigms in urban logistics and transportation. Although optimization theories and algorithms are well-established, a persistent gap remains between these theoretical foundations and the practical skills required to formulate models and implement effective, data-driven solutions. This positioning paper aims to bridge this gap by introducing optimization modeling frameworks and solution methodologies tailored to urban logistics and transportation applications. The paper targets researchers, students, and practitioners in non-business disciplines such as civil engineering, industrial engineering, economics, and computer science. We present several widely used data-driven optimization paradigms, including deterministic optimization with sensitivity analysis, two-stage stochastic programming, integrated simulation-optimization, and Markov decision process, and discuss how mathematical programming, statistical methods, simulation, and machine learning can support decision making. We then provide a systematic classification of urban transportation applications based on key decision characteristics, such as whether decisions are static or dynamic, deterministic or stochastic, and whether they involve interactions among multiple decision makers. Building on this classification, we propose a practical roadmap to guide the selection and implementation of appropriate optimization approaches. Illustrative applications, demonstrate the relevance and versatility of the proposed frameworks.