Masters Theses

Keywords and Phrases

Discrete Event Simulation; Hydraulic Shovels; Machine Learning Classification; Operator Behavior Analysis; Stalling; Swing and Bucket Dynamics

Abstract

This study aims to develop a reliable algorithm for stalling detection and investigate how operator behaviors impact the truck-loading process, focusing on swing and bucket rotations. While the study of excavator digging conditions and equipment performance is well understood, describing stalling from telemetry data and the influence of operators’ behavior on overall system efficiency remain underexplored. This study seeks to: (i) develop a reliable algorithm for stall detection by analyzing key operational variables such as velocity and angles; (ii) implement statistical and machine learning techniques, specifically a Support Vector Machine (SVM) classification algorithm, to differentiate between ideal and non-ideal digging behaviors; (iii) develop a discrete event simulation (DES) model in Arena® for evaluating the effect of differences in swing and bucket angles on hydraulic excavator production rates; and (iv) use the developed model to analyze the impact of different operator bucket and swing angle profiles on cycle times and overall production rates. The developed classification algorithm achieves a sensitivity of 100%, indicating it correctly identified all stalling cases, and a specificity of 99%, demonstrating its ability to accurately classify normal classes with minimal false positives. On the other hand, the DES simulation modeled four operators and involved 90 trucks with 85 replications, showing that Operator 1 achieves the highest payload rate with 78.2 tons/min. Thus, this study lays the groundwork for stall detection and modeling operator performance.

Advisor(s)

Awuah-Offei, Kwame, 1975-

Committee Member(s)

Eckert, Andreas
Xu, Guang

Department(s)

Mining Engineering

Degree Name

M.S. in Mining Engineering

Publisher

Missouri University of Science and Technology

Publication Date

Fall 2025

Pagination

xi, 126 pages

Note about bibliography

Includes_bibliographical_references_(pages 104-123)

Rights

© 2026 Mateo Fernando Montenegro Defaz , All Rights Reserved

Document Type

Thesis - Open Access

File Type

text

Language

English

Thesis Number

T 12573

Download

Share

 
COinS