Intelligent Hoisting with Car-Like Mobile Robots

Heng Li, Xiaochun Luo, Martin Skitmore

Research output: Journal article publicationJournal articleAcademic researchpeer-review

3 Citations (Scopus)

Abstract

The construction industry's traditional hoisting system always needs workers to complete the tasks involved, with concomitant extra labor costs and attention to the workers' safety. This paper describes the development of an intelligent hoisting system to optimize the hoisting process, involving the application of robotic cars and vision-based recognition. In the design of the hoisting system, electric hooks were used and maneuvered by a robotic car while the vision-based recognition system - based on capturing images by the camera - arranges the robotic motion. The Yolo v2 recognition algorithm was used, which provides fast and efficient vision-based recognition. More than 30 trials in an experimental prefabrication factory indicated that the system had a significant success rate of approximately 92.5% (3.7/4) - the proportion of hooks successfully grappling the hoist points - verifying the feasibility of the system. The primary contribution of this paper is in the development and demonstration of an intelligent hoisting system to optimize the hoisting process, involving the application of robotic cars and vision-based recognition, thus furthering the application of computer vision techniques and robotics to construction work.

Original languageEnglish
Article number04020136
JournalJournal of Construction Engineering and Management
Volume146
Issue number12
DOIs
Publication statusPublished - 1 Dec 2020

Keywords

  • Car-like mobile robots
  • Deep learning
  • Intelligent hoisting
  • Visual object detection

ASJC Scopus subject areas

  • Civil and Structural Engineering
  • Building and Construction
  • Industrial relations
  • Strategy and Management

Cite this