Next Article in Journal
Intelligent Diagnosis of Rolling Element Bearings Under Various Operating Conditions Using an Enhanced Envelope Technique and Transfer Learning
Next Article in Special Issue
Intelligent Decision Framework for Booster Fan Optimization in Underground Coal Mines: Hybrid Spherical Fuzzy-Cloud Model Approach Enhancing Ventilation Safety and Operational Efficiency
Previous Article in Journal
Determining the Range of Applicability of Analytical Methods for Belleville Springs and Novel Approach of Calculating Quasi-Progressive Spring Stacks
Previous Article in Special Issue
Fault Feature Extraction Based on Variational Modal Decomposition and Lifting Wavelet Transform: Application in Gear of Mine Scraper Conveyor Gearbox
 
 
Article
Peer-Review Record

Intelligent Scheduling in Open-Pit Mining: A Multi-Agent System with Reinforcement Learning

Machines 2025, 13(5), 350; https://doi.org/10.3390/machines13050350
by Gabriel Icarte-Ahumada 1,* and Otthein Herzog 2,3
Reviewer 1: Anonymous
Reviewer 2:
Machines 2025, 13(5), 350; https://doi.org/10.3390/machines13050350
Submission received: 20 March 2025 / Revised: 22 April 2025 / Accepted: 22 April 2025 / Published: 23 April 2025
(This article belongs to the Special Issue Key Technologies in Intelligent Mining Equipment)

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Please see the comments in the attached file.

Comments for author File: Comments.pdf

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

This is an interesting work presented in the article. The attempt of this work is timely and valuable. However, I need the following clarifications:

  1. Why dd you keep number of actions limited to binary decision? Was not there any case in the middle?
  2. It sounds that each process starts with a CFP by shovelAgents, then there seems to be no indication about how agents reach agreement. Do you use any kind of equilibrium, such as Nash equilibrium etc?
  3. It seems that some tracks may not be able to function at all subject to the circumstances if they do not have feasible offer in comparison to other competing tracks. How do you measure that all track are functioning to some extent? Or, do you allow truck to remain idle?
  4. why did you use Q-table not any ANN for training? How agents are learning? What is the subject decision variable to learn, and subject to what criteria?
  5. There is no indication or evidence for that agents learn? Can you demonstrate if the agents learn?
  6. Given the experimental results suggest, it sounds to me that the rationale of using RL for the purpose you used does not help improve the performance? I suggest you go for higher complexity to show the approach is useful. 

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

The vast majority of comments have been modified.

Comments on the Quality of English Language

The English could be improved to more clearly express the research.

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

I am happy with the paper as is.

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Back to TopTop