Wheel-INS: A Wheel-mounted MEMS IMU-based Dead Reckoning System

IMU drifts quickly with time! Well...it's true, but what if we mount it on the wheel❓

In Wheel-INS, we proposed a dead reckoning sysytem for the wheeled robots using only one Wheel-mounted IMU. There are two major advantages of mounting IMU to the wheel: 1) The vehicle velocity can be directly obtained by the Wheel-IMU without wheel encoder; 2) The continuous rotation of the wheel can somewhat mitigate the error accumulation of the inertial navigation system because of rotation modulation.

💥News💥

🎉🎉 Oct. 2023 A completely new version of the code supporting both Linux and Windows is released!

🎉🎉 Nov. 2022 Wheel-INS has been extended to a SLAM solution! Check out our paper of Wheel-SLAM accepted to IEEE Robotics and Automation Letters. The source code is also released.

🎉🎉 Nov. 2022 Our paper on multiple IMUs-based wheeled robot localization (Wheel-INS2) has been accepted to IEEE Transactions on Intelligent Transportation Systems. Check it out!

🎉🎉 Aug. 2021 Wheel-INS has been accepted to IEEE Transactions on Vehicular Technology. Check it out!

Run it

1. Linux

We recommend you use the g++ compiler with Ubuntu 20.04. The build-essential libraries should be installed first:

sudo apt-get install cmake
sudo apt-get install build-essential

After preparing the compilation environment, you can clone the repository and run it as follows:

# Clone the repository
git clone git@github.com:i2Nav-WHU/Wheel-INS.git ~/

# Build Wheel-INS
cd ~/Wheel-INS
mkdir build && cd build

cmake ..
make -j10

# Run demo dataset
cd ~/Wheel-INS
./bin/Wheel-INS config/robot.yaml

2. Windows

Here we show how to run the code with Visual Studio Code (VSCode), but you can also use other IDEs, e.g., Visual Studio.

• Install VSCode and the extensions: C/C++, C/C++ Extension Pack, CMake, and CMake Tools.
• Install CMake and Microsoft Visual C/C++ Build Tools.
• Open Wheel-INS with VSCode.
• Set compiler: open the Command Palette (Ctrl+Shift+P) and type "CMake: Select a Kit", select the correct build tool according to your system.
• Configure CMake: type and click "CMake: Configure" in the Command Palette.
• Compile Project: type and click "CMake: Build" in the Command Palette.

Once an executable file Wheel-INS.exe is generated, the compilation is done. Then, you can run it via the terminal in VSCode as follows:

.\bin\Release\Wheel-INS.exe config/robot.yaml

You can then run plot.py in utils to plot the trajectory estimated by Wheel-INS as well as the raw Wheel-IMU data. Here is an example plot for the robot dataset.

Datasets

Two sets of example data with ground truth are provided (see dataset). Please refer to the ReadMe.pdf for details. If git clone is too slow, please try to download the .zip file directly.

Related Papers

Based on the study of Wheel-IMU, we have published three papers. 1) The original Wheel-INS paper where we proposed a wheel-mounted IMU-based dead reckoning system and investigated its characteristics. 2) A thorough and complete comparison on three different measurement models in Wheel-INS with both theoretical analysis and experimental illustration. 3) A multiple IMUs-based localization system for wheeled robots by obtaining different dynamic information of the vehicle and taking advantage of the relative spatial constraints among the inertial sensors.

If you find our studies helpful to your academic research, please consider citing the related papers.

• X. Niu, Y. Wu and J. Kuang, "Wheel-INS: A Wheel-mounted MEMS IMU-based Dead Reckoning System," IEEE Transactions on Vehicular Technology, doi: 10.1109/TVT.2021.3108008, 2021. (pdf) (IEEE Xplore)

@ARTICLE{niu2021tvt,
  author={Niu, Xiaoji and Wu, Yibin and Kuang, Jian},
  journal={IEEE Transactions on Vehicular Technology}, 
  title={{Wheel-INS}: A Wheel-Mounted {MEMS IMU}-Based Dead Reckoning System}, 
  year={2021},
  volume={70},
  number={10},
  pages={9814-9825},
  doi={10.1109/TVT.2021.3108008}
}

• Y. Wu, X. Niu and J. Kuang, "A Comparison of Three Measurement Models for the Wheel-mounted MEMS IMU-based Dead Reckoning System," IEEE Transactions on Vehicular Technology, doi: 10.1109/TVT.2021.3102409, 2021. (pdf) (IEEE Xplore)

@ARTICLE{wu2021tvt,
  author={Wu, Yibin and Niu, Xiaoji and Kuang, Jian},
  journal={IEEE Transactions on Vehicular Technology}, 
  title={A Comparison of Three Measurement Models for the Wheel-Mounted {MEMS IMU}-Based Dead Reckoning System}, 
  year={2021},
  volume={70},
  number={11},
  pages={11193-11203},
  doi={10.1109/TVT.2021.3102409}}

• Y. Wu, J. Kuang and X. Niu, "Wheel-INS2: Multiple MEMS IMU-based Dead Reckoning System for Wheeled Robots with Evaluation of Different IMU Configurations," IEEE Transactions on Intelligent Transportation Systems, 2022. (pdf)(IEEE Xplore)

@ARTICLE{wu2022tits,  
    author={Wu, Yibin and Kuang, Jian and Niu, Xiaoji},  
    journal={IEEE Transactions on Intelligent Transportation Systems},   
    title={{Wheel-INS2}: Multiple {MEMS IMU}-Based Dead Reckoning System With Different Configurations for Wheeled Robots},   
    year={2022}, 
    pages={1-14},  
    doi={10.1109/TITS.2022.3220508}
}

License

The code is released under GPLv3 license.

Acknowledgement

We would like to thank the i2Nav group for sharing KF-GINS.

For any questions, please feel free to contact Mr. Yibin Wu (ybwu at whu.edu.cn) or Dr. Jian Kuang (kuang at whu.edu.cn).