Neural inertial navigation method for wheeled robots based on self-supervised learning

Authors

  • Fengrong Huang School of Mechanical Engineering, Hebei University of Technology, Tianjin 300400, China
  • Mengqi Gao School of Mechanical Engineering, Hebei University of Technology, Tianjin 300400, China
  • Qinglin Liu National Key Laboratory of Electromagnetic Space Security, Tianjin 300308, China
  • Min Gao School of Mechanical Engineering, Hebei University of Technology, Tianjin 300400, China

DOI:

https://doi.org/10.24425/mms.2025.154345

Keywords

MEMS-IMU, inertial navigation, self-supervised learning, wheeled robot navigation, motion state recognition

Abstract

Low-cost Micro-Electro-Mechanical System Inertial Measurement Units (MEMS-IMUs) are plagued by large, complex, and variable errors. Traditional strap-down inertial navigation systems that utilize MEMS-IMUs are unable to meet the positioning requirements of wheeled robots. Although inertial navigation based on deep learning has been explored, it necessitates a substantial amount of carefully selected and labelled data, resulting in high costs. Consequently, this paper proposes a self-supervised neural inertial navigation method for wheeled robots that solely depends on MEMS-IMU data. Firstly, a representation learning model is established to extract general IMU features for self-supervised denoising. Subsequently, an intelligent framework employing contrastive learning is adopted to explore the latent information of the IMU and acquire the motion state of the robot. Specific motion state information is regarded as observations, and an invariant extended Kalman filter (IEKF) is applied for information fusion to enhance positioning accuracy. Experiments conducted on public datasets demonstrate that, in the absence of additional ground truth values, the Absolute Trajectory Error (ATE) and Temporal Relative Trajectory Error (T-RTE) of the proposed method are 20.23% and 30.71% lower than those of supervised learning-based methods, respectively. The proposed method offers a more cost-effective and practical solution for the development of inertial navigation technology for wheeled robots.

References

[1] Chen, Y., Xu, H., Yang, W., Yang, C., & Xu, K. (2021). Rat robot motion state identification based on a wearable inertial sensor. Metrology and Measurement Systems, Vol. 28 (2021), No. 2, 255-268. https://doi.org/10.24425/mms.2021.136605

[2] Chen, C., & Pan, X. (2024). Deep Learning for Inertial Positioning: A Survey. IEEE Transactions on Intelligent Transportation Systems, 25(9), 10506-10523. https://doi.org/10.1109/tits.2024.3381161

[3] Rong, H., Wu, X., Wang, H., Jin, T., & Zou, L. (2024). Attitude estimation based on multi-scale grouped spatio-temporal attention neural networks. Metrology and Measurement Systems, Vol. 31 (2024), No. 1, 195-211. https://doi.org/10.24425/mms.2024.148542

[4] Lan, J., Wang, K., Song, S., Li, K., Liu, C., He, X., Hou, Y., & Tang, S. (2024). Method for measuring non-stationary motion attitude based on MEMS-IMU array data fusion and adaptive filtering. Measurement Science and Technology, 35(8), 086304. https://doi.org/10.1088/1361-6501/ad44c8

[5] Wang, Z., & Cheng, X. (2021). Adaptive optimization online IMU self-calibration method for visual-inertial navigation systems. Measurement, 180, 109478. https://doi.org/10.1016/j.measurement.2021.109478

[6] Gao, Y., Shi, D., Li, R., Liu, Z., & Sun, W. (2022). Gyro-NeT: IMU Gyroscopes Random Errors Compensation Method Based on Deep Learning. IEEE Robotics and Automation Letters, 8(3), 1471-1478. https://doi.org/10.1109/lra.2022.3230594

[7] Brossard, M., Bonnabel, S., & Barrau, A. (2020b). Denoising IMU Gyroscopes with Deep Learning for Open-Loop Attitude Estimation. IEEE Robotics and Automation Letters, 1. https://doi.org/10.1109/lra.2020.3003256

[8] Liu, W., Caruso, D., Ilg, E., Dong, J., Mourikis, A. I., Daniilidis, K., Kumar, V., & Engel, J. (2020). TLIO: Tight Learned Inertial Odometry. IEEE Robotics and Automation Letters, 5(4), 5653-5660. https://doi.org/10.1109/lra.2020.3007421

[9] Wang, Y., Kuang, J., Niu, X., & Liu, J. (2022). LLIO: Lightweight Learned Inertial Odometer. IEEE Internet of Things Journal, 10(3), 2508-2518. https://doi.org/10.1109/jiot.2022.3214087

[10] Brossard, M., Barrau, A., & Bonnabel, S. (2019). RINS-W: Robust Inertial Navigation System on wheels. 2021 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS). https://doi.org/10.1109/iros40897.2019.8968593

[11] Yang, M., Zhu, R., Xiao, Z., & Yan, B. (2021). Symmetrical-Net: Adaptive zero velocity detection for ZUPT-Aided pedestrian navigation system. IEEE Sensors Journal, 22(6), 5075-5085. https://doi.org/10.1109/jsen.2021.3094301

[12] Jeong, J., Cho, Y., Shin, Y., Roh, H., & Kim, A. (2018). Complex Urban LiDAR Data set. In 2018 IEEE International Conference on Robotics and Automation (ICRA), 6344-6351. https://doi.org/10.1109/icra.2018.8460834

[13] Feng, D., Qi, Y., Zhong, S., Chen, Z., Chen, Q., Chen, H., Wu, J., & Ma, J. (2024). S3E: a Multi-Robot Multimodal dataset for collaborative SLAM. IEEE Robotics and Automation Letters, 9(12), 11401-11408. https://doi.org/10.1109/lra.2024.3490402

[14] Wei, H., Jiao, J., Hu, X., Yu, J., Xie, X., Wu, J., Zhu, Y., Liu, Y., Wang, L., & Liu, M. (2024). FusionPortableV2: A unified multi-sensor dataset for generalized SLAM across diverse platforms and scalable environments. The International Journal of Robotics Research. https://doi.org/10.1177/02783649241303525

[15] Zhou, H., Zhang, S., Peng, J., Zhang, S., Li, J., Xiong, H., & Zhang, W. (2021). Informer: Beyond Efficient Transformer for Long Sequence time-series Forecasting. In Proceedings of the AAAI Conference on Artificial Intelligence, 35(12), 11106-11115. https://doi.org/10.1609/aaai.v35i12.17325

[16] Wang, Y., Sun, X., Cui, D., Wang, X., Jia, Z., & Zhang, Z. (2024). An adaptive estimation of ground vehicle state with unknown measurement noise. Metrology and Measurement Systems, Vol. 31 (2024) No. 2, 389-399. https://doi.org/10.24425/mms.2024.149705

[17] Barrau, A., & Bonnabel, S. (2016). The invariant extended Kalman filter as a stable observer. IEEE Transactions on Automatic Control, 62(4), 1797-1812. https://doi.org/10.1109/tac.2016.2594085

[18] Brossard, M., Barrau, A., & Bonnabel, S. (2020). AI-IMU Dead-Reckoning. IEEE Transactions on Intelligent Vehicles, 5(4), 585-595. https://doi.org/10.1109/tiv.2020.2980758

[19] Jeong, J., Cho, Y., Shin, Y., Roh, H., & Kim, A. (2018). Complex Urban LiDAR Data set. In In 2018 IEEE International Conference on Robotics and Automation (ICRA), 6344-6351. https://doi.org/10.1109/icra.2018.8460834

[20] Guo, F., Yang, H., Wu, X., Dong, H., Wu, Q., & Li, Z. (2023). Model-Based deep Learning for Low-Cost IMU dead reckoning of wheeled mobile robot. IEEE Transactions on Industrial Electronics, 71(7), 7531-7541. https://doi.org/10.1109/tie.2023.3301531

[21] El-Sheimy, N., Hou, H., & Niu, X. (2007). Analysis and modeling of inertial sensors using Allan Variance. IEEE Transactions on Instrumentation and Measurement, 57(1), 140-149. https://doi.org/10.1109/tim.2007.908635

Downloads

Published

2025-06-30

How to Cite

Huang, Fengrong, et al. “Neural Inertial Navigation Method for Wheeled Robots Based on Self-Supervised Learning”. Metrology and Measurement Systems, vol. 32, no. 2, June 2025, pp. art. no. 13, 1-19, doi:10.24425/mms.2025.154345.

Issue

Vol. 32 No. 2 (2025)

Section

Articles