International Journal of Advanced Technology and Engineering Exploration ISSN (Print): 2394-5443    ISSN (Online): 2394-7454 Volume-13 Issue-138 May-2026
  1. 4774
    Citations
  2. 2.8
    CiteScore
Comparative performance analysis of YOLOv8 and YOLO-NAS for asphalt road damage detection

Leonardus Sandy Ade Putra1 and Eka Kusumawardhani1

Department of Electrical Engineering,University of Tanjungpura, Pontianak,West Kalimantan 78124,Indonesia1
Corresponding Author : Leonardus Sandy Ade Putra

Recieved : 26-October-2025; Revised : 20-May-2026; Accepted : 22-May-2026

Abstract

Potholes are a common form of asphalt road damage that not only reduces driving comfort and damages vehicles but also significantly increases the risk of traffic accidents. Conventional road inspection methods for detecting such damage are often costly, time-consuming, and inefficient. This paper aims to develop an asphalt road damage detection (RDD) system based on computer vision using the You Only Look Once version 8 (YOLOv8) and You Only Look Once–neural architecture search (YOLO-NAS) models. The dataset was constructed using publicly available Kaggle datasets along with original dashcam recordings collected from Pontianak City. It was divided into 6,091 training images, 2,094 validation images, and 1,055 testing images. Model performance was enhanced through hyperparameter optimization and data augmentation during training. The evaluation results reveal a significant trade-off between detection accuracy and inference speed. The YOLO-NAS model achieved the highest performance, with a precision of 0.932, recall of 0.919, mAP@0.5 of 0.924, and F1-score of 0.921, but operated at 25 frames per second (FPS). In contrast, the YOLOv8m model achieved a considerably faster inference speed of 60 FPS, with slightly lower performance metrics, including a precision of 0.895, recall of 0.861, mAP@0.5 of 0.886, and F1-score of 0.873. Although YOLO-NAS demonstrated superior detection accuracy, YOLOv8m provided a more balanced trade-off between speed and performance, making it more suitable for real-time applications. This study contributes to the development of an efficient road condition monitoring system that supports improved transportation safety and sustainable infrastructure maintenance.

Keywords

Pothole detection, Asphalt Road damage, Computer vision, YOLOv8, YOLO-NAS, Intelligent transportation systems.

Cite this article

Putra LSA, Kusumawardhani E. Comparative performance analysis of YOLOv8 and YOLO-NAS for asphalt road damage detection. International Journal of Advanced Technology and Engineering Exploration. 2026;13(138):665-682. DOI : 10.19101/IJATEE.2025.121221413

References
[1]
Bruwer MM, Andersen SJ. Using commercial floating car data to remotely infer the presence of potholes along rural road segments. African Transport Studies. 2025; 3:1-12.
[Crossref] [Google Scholar]
[2]
Bouhsissin S, Assemlali H, Sael N. Enhancing road safety: a convolutional neural network based approach for road damage detection. Machine Learning with Applications. 2025; 20:1-14.
[Crossref] [Google Scholar]
[3]
Xu Z, Lei X, Guan H. Multi-scale local feature enhanced transformer network for pavement crack detection. Journal of Image and Graphics. 2023; 28(4):1019-28.
[Crossref] [Google Scholar]
[4]
Hu X, Assaad RH. Real-time robotic teleoperation for pavement pothole segmentation, quantification, and localization using multimodal sensing and efficient multi-scale attention-enhanced edge deep learning. Automation in Construction. 2026; 183:1-22.
[Crossref] [Google Scholar]
[5]
Tello-cifuentes L, Thomson P, Marulanda J, Sandoval E, Bassier M. Deep learning-based detection and evaluation of pavement surface damage. Results in Engineering. 2025:1-23.
[Crossref] [Google Scholar]
[6]
Nimma D, Al-omari O, Pradhan R, Ulmas Z, Krishna RV, El-ebiary TY, et al. Object detection in real-time video surveillance using attention based transformer-YOLOv8 model. Alexandria Engineering Journal. 2025; 118:482-95.
[Crossref] [Google Scholar]
[7]
Roman-garay M, Rodriguez-rangel H, Hernandez-beltran CB, Lepej P, Arreygue-Rocha JE, Morales-rosales LA. Architecture for pavement pothole evaluation using deep learning, machine vision, and fuzzy logic. Case Studies in Construction Materials. 2025; 22:1-26.
[Crossref] [Google Scholar]
[8]
Xu X, Wu T, Ge W, Dai L, Gong J, Chen D, et al. Enhanced pothole detection in low-light conditions using DCE-Pothole and YOLOv10 fusion. Results in Engineering. 2025; 28:1-16.
[Crossref] [Google Scholar]
[9]
Saisree C, Kumaran U. Pothole detection using deep learning classification method. Procedia Computer Science. 2023; 218:2143-52.
[Crossref] [Google Scholar]
[10]
Zhou C, Sun Y, Bi H, Zhou S, Deng X, Wang P, et al. Real-time road defect detection method based on improved YOLOv8-S model. Engineering Research Express. 2026.
[Crossref] [Google Scholar]
[11]
Li J, Qu Z, Wang S, Xia S. A method of road damage detection for complex background images based on region guidance network. Pattern Recognition. 2025; 168:111780.
[Crossref] [Google Scholar]
[12]
Omodaratan B, Jamali A, Wiley T, Al-saadi Z, Mallipeddi R, Asadi E, et al. Advances in you only look once (YOLO) algorithms for lane and object detection in autonomous vehicles. Engineering Applications of Artificial Intelligence. 2026; 168:1-19.
[Crossref] [Google Scholar]
[13]
Yang X, Del RCE, Zou Y, Wotherspoon L, Yang J, Li H. Automated concrete bridge damage detection using an efficient vision transformer-enhanced anchor-free yolo. Engineering. 2025; 51:311-26.
[Crossref] [Google Scholar]
[14]
Kang J, Guan H, Zhang D, Ma L, Wang L, Yu Y, et al. RPDNet: street-level road pavement damage detection with a real-time anchor-free network. International Journal of Applied Earth Observation and Geoinformation. 2026; 146:1-22.
[Crossref] [Google Scholar]
[15]
Omarov B, Kulambayev B. Deep learning-based image processing for real-time detection of road surface damage. Procedia Computer Science. 2024; 251:609-14.
[Crossref] [Google Scholar]
[16]
Li Y, Yin C, Lei Y, Zhang J, Yan Y. RDD-YOLO: road damage detection algorithm based on improved you only look once version 8. Applied Sciences. 2024; 14(8):1-17.
[Crossref] [Google Scholar]
[17]
Ma L, Chen M. Road damage detection based on improved YOLO algorithm. Scientific Reports. 2025; 15(1):1-12.
[Crossref] [Google Scholar]
[18]
Yu H, Liu J, Lin M. A comprehensive literature review on YOLO-based small object detection: methods, challenges, and future trends. Computers, Materials, & Continua. 2026; 87(1):1-53.
[Crossref] [Google Scholar]
[19]
Casas E, Ramos L, Romero C, Rivas-echeverria F. A comparative study of YOLOv5 and YOLOv8 for corrosion segmentation tasks in metal surfaces. Array. 2024; 22:1-16.
[Crossref] [Google Scholar]
[20]
Jrondi Z, Moussaid A, Hadi MY. Exploring end-to-end object detection with transformers versus YOLOv8 for enhanced citrus fruit detection within trees. Systems and Soft Computing. 2024; 6:1-9.
[Crossref] [Google Scholar]
[21]
Zeng J, Zhong H. YOLOv8-PD: an improved road damage detection algorithm based on YOLOv8n model. Scientific Reports. 2024; 14(1):1-14.
[Crossref] [Google Scholar]
[22]
Zhang S, He X, Xue B, Wu T, Ren K, Zhao T. Segment-anything embedding for pixel-level road damage extraction using high-resolution satellite images. International Journal of Applied Earth Observation and Geoinformation. 2024; 131:103985.
[Crossref] [Google Scholar]
[23]
Paramarthalingam A, Sivaraman J, Theerthagiri P, Vijayakumar B, Baskaran V. A deep learning model to assist visually impaired in pothole detection using computer vision. Decision Analytics Journal. 2024; 12:100507.
[Crossref] [Google Scholar]
[24]
Xue Y, Yao C, Wahib M, Gabbouj M. YOLO-DKR: differentiable architecture search based on kernel reusing for object detection. Information Sciences. 2025; 713:122180.
[Crossref] [Google Scholar]
[25]
Gong L, Wang Y, Huang X, Liang J, Fan Y. An improved YOLO algorithm with multisensing for pedestrian detection. Signal, Image and Video Processing. 2024; 18(8):5893-906.
[Crossref] [Google Scholar]
[26]
https://www.kaggle.com/code/farzadnekouei/pothole-segmentation-for-road-damage-assessment. Accessed 27 April 2026.
[27]
Nissimagoudar PC, Miskin SR, Sali VN, SK R, SK D, HM G, et al. Detection of potholes and speed breaker for autonomous vehicles. Procedia Computer Science. 2024; 237:675-82.
[Crossref] [Google Scholar]