Optimization-based excavator pose estimation using real-time location systems

Faridaddin Vahdatikhaki, Amin Hammad*, Hassaan Siddiqui

*Corresponding author for this work

Research output: Contribution to journalArticleAcademicpeer-review

44 Citations (Scopus)

Abstract

Of various types of equipment engaged in earthwork projects, excavators account for the largest proportion of fatalities on site, with 36 reported incidents in the U.S. only in 2012. Furthermore, the operations of excavators play a defining role in the productivity of earthwork projects. Thus, it is of a crucial importance to constantly monitor the operations of excavators to ensure the smooth progress of the work, early detection of anomalies and prompt corrective measures. The first step in the monitoring of an excavator is estimating its pose. Real-time location systems (RTLSs) provide a robust and reliable technology to track and monitor excavators in near real-time. Location data captured by RTLSs can contribute to the identification of machine-induced safety hazards and the analysis of operation productivity. Nevertheless, while high-accuracy RTLSs require a considerable financial commitment, the more affordable variations of the technology, e.g. Ultra-Wideband (UWB), does not generate accurate enough data that can be readily used for the excavator pose estimation. As a result, the location data generated from such RTLSs require some processing before they can be effectively deployed for the pose estimation. The present research proposes a robust optimization-based method that uses geometric and operational characteristics of an excavator to improve the quality of the pose estimation through maximizing the compliance with the machine-imposed constraints and minimizing the amount of required corrections. The feasibility of the proposed method is validated through two laboratory case studies. The method generates very promising results in terms of maintaining the geometric integrity of the equipment data and estimating the pose of the equipment.

Original languageEnglish
Pages (from-to)76-92
Number of pages17
JournalAutomation in construction
Volume56
DOIs
Publication statusPublished - Aug 2015

Keywords

  • Excavator
  • Optimization
  • Pose estimation
  • RTLS

Fingerprint Dive into the research topics of 'Optimization-based excavator pose estimation using real-time location systems'. Together they form a unique fingerprint.

Cite this