After hard working for half of a year, we have successfully replaced all expensive components on MIT RaceCar architecture with low cost choices and still keep the high speed capability on our platform. Instead of using encoder and stereo VO to estimate the odometry, our platform adopt the combination of laser odom generated by RPLidar-A1 and yaw measuement from gy85 IMU with fine tuning EKF fusion.
The total cost of our platform is only around 600 USD (inc. lidar, car, battery, etc)
The Youtube link for 3 m/s demo video: [https://www.youtube.com/watch?v=jgS6gpmWPoY](https://www.youtube.com/watch?v=jgS6gpmWPoY)
In addition, a step by step tutorial for beginner now is available on our github:
it includes from hardware assembling to software installation and operation.
Currently, we are integrating teb_local_planner on our platform in order to achieve auto-parking capability and trying to add low-cost VIO to obtain better odometry info.
Please feel free to leave any suggestion below or open an issue on our github.
HyphaROS Workshop is one of the most famous group in Taiwan for ROS related activities. We have organized several ROS based workshops and are planning to hold a small competition for auto-racecar in public by the end of this year.