IRO Journals
Journal of Soft Computing Paradigm

Machine Learning-Enabled Rover For Autonomous Navigation In Rough Terrain

Open Access
Published: 04 November, 2026
Pages: 432-454
  • Parameswari S. , Parameswari S.
    Associate Professor, Department of Electronics and Communication Engineering, Sri Sairam Institute of Technology, Tambaram, Chennai - 44, India.
    Associate Professor, Department of Electronics and Communication Engineering, Sri Sairam Institute of Technology, Tambaram, Chennai - 44, India.
  • Mohammed Daiyaan R. , Mohammed Daiyaan R.
    Students, Department of Electronics and Communication Engineering, Sri Sairam Institute of Technology, Tambaram, Chennai - 44, India.
    Students, Department of Electronics and Communication Engineering, Sri Sairam Institute of Technology, Tambaram, Chennai - 44, India.
  • Chella Jayeshwar K. , Chella Jayeshwar K.
    Students, Department of Electronics and Communication Engineering, Sri Sairam Institute of Technology, Tambaram, Chennai - 44, India.
    Students, Department of Electronics and Communication Engineering, Sri Sairam Institute of Technology, Tambaram, Chennai - 44, India.
  • Kannan K. Kannan K.
    Students, Department of Electronics and Communication Engineering, Sri Sairam Institute of Technology, Tambaram, Chennai - 44, India.
    Students, Department of Electronics and Communication Engineering, Sri Sairam Institute of Technology, Tambaram, Chennai - 44, India.
view PDF
view PDF

How to Cite

S., Parameswari, Mohammed Daiyaan R., Chella Jayeshwar K., and Kannan K. 2026. “Machine Learning-Enabled Rover For Autonomous Navigation In Rough Terrain”. Journal of Soft Computing Paradigm 8 (4): 432-54. https://doi.org/10.36548/jscp.2026.4.008.
ACM ACS APA ABNT Chicago Harvard IEEE MLA Turabian Vancouver AMA

Download Citation

Endnote/Zotero/Mendeley (RIS) BibTeX

Keywords

Autonomous Rover
Terrain Classification
YOLOv8n
ESP32
Hybrid Perception–Control
Obstacle Avoidance
Fail-Safe Navigation.

Abstract

A hybrid perception-control framework for navigating rough terrains by remote YOLOv8n-based terrain classification combined with on-board safety-based control of rover is proposed. The system uses ESP32-WROOM-32 module for controlling motors, ultrasonic sensors for detecting obstacles, infrared sensors for detecting cliff edges, and watchdog communication, whereas ESP32-CAM is responsible for sending terrain images to a laptop where the inference algorithm works. Four-class terrain data consisting of 800 images of grass, sand, gravel, and soil is employed for fine-tuning YOLOv8n using stratified dataset. Terrain predictions control movements depending on traversability coefficients, while safety controls take precedence over remote controls when an obstacle or communication link is detected. The evaluation over 60 trials of outdoor navigation has resulted in terrain classification with an accuracy of 78.3% and successful navigation with an 80.0% success rate. The perception-to-action latency was 170±23 ms, fail-safe time was 1.02±0.04 s, and emergency braking was 310±45 ms. The power consumption of the rover was 4.22±0.35 W, and the estimated operational time was around 4.5 h. Such values characterize the performance of the limited resources rover that implements the separation of complex computation-based inference from motion control in embedded systems.

References

  1. Shilpa, D., S. Pavani, M. Triveni, P. Nandini, G. Venkateswarao, and M. Navya. "An ESP32-Based Autonomous Mobile Robot for Rough Terrain Navigation Using Multi-Sensor Obstacle Avoidance and Hazard Detection." International Journal of AI Electronics and Nexus Energy, 2026, vol. 2, no. 2: 77-85.
  2. Zhao, Jianwei, Jianhua Fang, Shouzhong Wang, Kun Wang, Chengxiang Liu, and Tao Han. "Obstacle avoidance of multi-sensor intelligent robot based on road sign detection." Sensors, 2021, vol. 21, no. 20: 6777.
  3. Zhou, Lanfeng, Ziwei Liu, and Wenfeng Wang. "Terrain Classification Algorithm for Lunar Rover Using a Deep Ensemble Network with High-Resolution Features and Interdependencies between Channels." Wireless Communications and Mobile Computing, 2020, vol. 2020, no. 1: 8842227.
  4. Egan, Rebecca, and Ali Haydar Göktogan. "Deep learning-based terrain classification for traversability analysis, path planning and control of a Mars rover." In Australasian Conference on Robotics and Automation, 2021, 1-10.
  5. Zhang, Bo, Guobin Li, Qixin Zheng, Xiaoshan Bai, Yu Ding, and Awais Khan. "Path planning for wheeled mobile robot in partially known uneven terrain." Sensors, 2022, vol. 22, no. 14: 5217.
  6. Zhang, Qingfeng, Wenpeng Ma, Qingchun Zheng, Xiaofan Zhai, Wenqian Zhang, Tianchang Zhang, and Shuo Wang. "Path planning of mobile robot in dynamic obstacle avoidance environment based on deep reinforcement learning." IEEE Access, 2024, vol. 12: 189136-189152.
  7. Zim, Md Ziaul Haque. "TinyML: analysis of Xtensa LX6 microprocessor for neural network applications by ESP32 SoC." arXiv preprint arXiv:2106.10652, 2021.
  8. Gupta, Shubbham, and Shiv Naresh Shivhare. "Embedded tinyml for predictive maintenance: Vibration analysis on esp32 with real-time fault detection in industrial equipment." International Journal on Computational Modelling Applications, 2025, vol. 2, no. 2: 1-17.
  9. Ng, Yung Chuen, Qi Wen Shervina Lim, Chun Ye Tan, Zhen Hao Gan, and Meng Yee Michael Chuah. "Tactile Aware Dynamic Obstacle Avoidance in Crowded Environment with Deep Reinforcement Learning." arXiv preprint arXiv:2406.13434, 2024.
  10. Jocher, Glenn, Ayush Chaurasia, and Jing Qiu. "Ultralytics YOLOv8." Version 8.0.0, Ultralytics, 2023. Available: https://github.com/ultralytics/ultralytics.
  11. Off-Road Robotics Research Group. "Unstructured Rough Terrain Surface Classification Benchmark Dataset." GitHub Repository, 2023. Available: https://github.com/offroad-robotics