Scholarlink

The increase in port terminals demand over the last decades highlighted a challenging scenario, especially for port terminals located in urban areas, of truck congestion, long waiting times, and, consequently, a significant amount of CO2 emission by trucks during this waiting time. Smart technologies and data-driven approaches are possible solutions to minimize the impact of demand increase on port efficiency and reduce the waiting times at port hinterland. Thus, in the context of smart technologies, the main objective of this research is to develop a flexible Truck Appointment System using a Machine Learning algorithm to predict and coordinate truck arrivals, reducing the waiting time that causes congestion and pollution emissions at the port area. Using a Discrete-event simulation in R Language, we compare three different scenarios with data from a real use case of a port terminal located in South America. From the simulated scenarios, it was possible to achieve a reduction of about 90% in CO2 emission using a Machine Learning algorithm to autonomously identify disruptions regarding truck appointments and propose a real-time reschedule, avoiding waiting time and congestion at port area.

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

Port Logistics Emissions Control Using Machine Learning to Coordinate the Truck Flow

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Maurício Randolfo Flores,
  • Julia Cristina Bremen,
  • Enzo Morosini Frazzon,
  • Gisele de Lorena Diniz Chaves

摘要

The increase in port terminals demand over the last decades highlighted a challenging scenario, especially for port terminals located in urban areas, of truck congestion, long waiting times, and, consequently, a significant amount of CO2 emission by trucks during this waiting time. Smart technologies and data-driven approaches are possible solutions to minimize the impact of demand increase on port efficiency and reduce the waiting times at port hinterland. Thus, in the context of smart technologies, the main objective of this research is to develop a flexible Truck Appointment System using a Machine Learning algorithm to predict and coordinate truck arrivals, reducing the waiting time that causes congestion and pollution emissions at the port area. Using a Discrete-event simulation in R Language, we compare three different scenarios with data from a real use case of a port terminal located in South America. From the simulated scenarios, it was possible to achieve a reduction of about 90% in CO2 emission using a Machine Learning algorithm to autonomously identify disruptions regarding truck appointments and propose a real-time reschedule, avoiding waiting time and congestion at port area.