Deep learning based identification and tracking of railway bogie parts

The Train Rolling-Stock Examination (TRSE) is a safety examination process that physically examines the bogie parts of a moving train, typically at speeds over 30 km/h. Currently, this inspection process is done manually by railway personnel in many countries to ensure safety and prevent interruptio...

Full description

Saved in:
Bibliographic Details
Main Authors: Muhammad Zakir Shaikh, Zeeshan Ahmed, Enrique Nava Baro, Samreen Hussain, Mariofanna Milanova
Format: Article
Language:English
Published: Elsevier 2024-11-01
Series:Alexandria Engineering Journal
Subjects:
Online Access:http://www.sciencedirect.com/science/article/pii/S111001682400797X
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1846167087423160320
author Muhammad Zakir Shaikh
Zeeshan Ahmed
Enrique Nava Baro
Samreen Hussain
Mariofanna Milanova
author_facet Muhammad Zakir Shaikh
Zeeshan Ahmed
Enrique Nava Baro
Samreen Hussain
Mariofanna Milanova
author_sort Muhammad Zakir Shaikh
collection DOAJ
description The Train Rolling-Stock Examination (TRSE) is a safety examination process that physically examines the bogie parts of a moving train, typically at speeds over 30 km/h. Currently, this inspection process is done manually by railway personnel in many countries to ensure safety and prevent interruptions to rail services. Although many earlier attempts have been made to semi-automate this process through computer-vision models, these models are iterative and still require manual intervention. Consequently, these attempts were unsuitable for real-time implementations. In this work, we propose a detection model by utilizing a deep-learning based classifier that can precisely identify bogie parts in real-time without manual intervention, allowing an increase in the deployability of these inspection systems. We implemented the Anchor-Free Yolov8 (AFYv8) model, which has a decoupled-head module for recognizing bogie parts. Additionally, we incorporated bogie parts tracking with the AFYv8 model to gather information about any missing parts. To test the effectiveness of the AFYv8-model, the bogie videos were captured at three different timestamps and the result shows the increase in the recognition accuracy of TRSE by 10 % compared to the previously developed classifiers. This research has the potential to enhance railway safety and minimize operational interruptions.
format Article
id doaj-art-0129aa0264a5422a99160894b6098506
institution Kabale University
issn 1110-0168
language English
publishDate 2024-11-01
publisher Elsevier
record_format Article
series Alexandria Engineering Journal
spelling doaj-art-0129aa0264a5422a99160894b60985062024-11-15T06:11:14ZengElsevierAlexandria Engineering Journal1110-01682024-11-01107533546Deep learning based identification and tracking of railway bogie partsMuhammad Zakir Shaikh0Zeeshan Ahmed1Enrique Nava Baro2Samreen Hussain3Mariofanna Milanova4NCRAAI MUET & NCRA-CMS Lab, Mehran University of Engineering and Technology (MUET), Jamshoro 76062, Pakistan; Universidad de Malaga, Málaga, Spain; Correspondence to: NCRAAI MUET & NCRA-CMS Lab, Mehran University of Engineering and Technology (MUET), Jamshoro 76062, Pakistan and School of Industrial Engineering, University of Malaga, Spain.NCRAAI MUET & NCRA-CMS Lab, Mehran University of Engineering and Technology (MUET), Jamshoro 76062, PakistanUniversidad de Malaga, Málaga, SpainDawood University of Engineering & Technology, Karachi 74800, PakistanDepartment of Computer Science, University of Arkansas, Little Rock, USAThe Train Rolling-Stock Examination (TRSE) is a safety examination process that physically examines the bogie parts of a moving train, typically at speeds over 30 km/h. Currently, this inspection process is done manually by railway personnel in many countries to ensure safety and prevent interruptions to rail services. Although many earlier attempts have been made to semi-automate this process through computer-vision models, these models are iterative and still require manual intervention. Consequently, these attempts were unsuitable for real-time implementations. In this work, we propose a detection model by utilizing a deep-learning based classifier that can precisely identify bogie parts in real-time without manual intervention, allowing an increase in the deployability of these inspection systems. We implemented the Anchor-Free Yolov8 (AFYv8) model, which has a decoupled-head module for recognizing bogie parts. Additionally, we incorporated bogie parts tracking with the AFYv8 model to gather information about any missing parts. To test the effectiveness of the AFYv8-model, the bogie videos were captured at three different timestamps and the result shows the increase in the recognition accuracy of TRSE by 10 % compared to the previously developed classifiers. This research has the potential to enhance railway safety and minimize operational interruptions.http://www.sciencedirect.com/science/article/pii/S111001682400797XComputer VisionObject DetectionDeep LearningYoloTrain Rolling StockWheelset
spellingShingle Muhammad Zakir Shaikh
Zeeshan Ahmed
Enrique Nava Baro
Samreen Hussain
Mariofanna Milanova
Deep learning based identification and tracking of railway bogie parts
Alexandria Engineering Journal
Computer Vision
Object Detection
Deep Learning
Yolo
Train Rolling Stock
Wheelset
title Deep learning based identification and tracking of railway bogie parts
title_full Deep learning based identification and tracking of railway bogie parts
title_fullStr Deep learning based identification and tracking of railway bogie parts
title_full_unstemmed Deep learning based identification and tracking of railway bogie parts
title_short Deep learning based identification and tracking of railway bogie parts
title_sort deep learning based identification and tracking of railway bogie parts
topic Computer Vision
Object Detection
Deep Learning
Yolo
Train Rolling Stock
Wheelset
url http://www.sciencedirect.com/science/article/pii/S111001682400797X
work_keys_str_mv AT muhammadzakirshaikh deeplearningbasedidentificationandtrackingofrailwaybogieparts
AT zeeshanahmed deeplearningbasedidentificationandtrackingofrailwaybogieparts
AT enriquenavabaro deeplearningbasedidentificationandtrackingofrailwaybogieparts
AT samreenhussain deeplearningbasedidentificationandtrackingofrailwaybogieparts
AT mariofannamilanova deeplearningbasedidentificationandtrackingofrailwaybogieparts