Closed Alchemistqqqq closed 10 months ago
Hi, thanks for your interest in our work. In principle, it should be able to be deployed on a Turtlebot 3 robot. The main concern is that I am not sure if the Raspberry Pi could run the DRL-VO policy in real time. But it's worth a good try. To do that, there are several key steps: 1) Follow our paper (i.e. Sec.V.B) to modify the lidar historical map input based on your lidar sensor configuration (i.e. how many lidar scan points). Code files need to be modified: barn_data_pub.py drl_vo_inference.py 2) Find the control command topic for Turtlebot 3 and modify the topic in drl_vo_inference.py 3) Try to install the singularity container software on the Raspberry Pi, then follow the ReadMe to depoly the DRL-VO policy; If it does not work, follow the ReadMe in drl_vo_nav package to manually install learning-based packages:
pip install torch==1.7.1+cu110 -f https://download.pytorch.org/whl/torch_stable.html
pip install gym==0.18.0 pandas==1.2.1
pip install stable-baselines3==1.1.0
pip install tensorboard psutil cloudpickle
After configuring the required environment you can depoly the DRL-VO policy to see how it works in the Turtlebot 3.
Thanks for your help, I will try the following real environment experiment according to your help.
Hello, I am very honored to read your paper and get your help in the simulation experiment a while ago. In your introduction, I saw that you used the turtlebot2 robot, but you used the Jackal robot in the competition. I would like to know if I can migrate this code to the turtlebot3 waffle pi for experiments in real environment. Because I am a beginner, if I can transfer to Waffle school for experiment, there may be some problems in the future, I hope you can give me some advice, thank you very much.