Thanks! Now I get what you are saying. I didn't even realize that the two sides of the legs looked differently.

Thanks, but I think my Bittle is installed the same way. The video likely looks a bit different because the limbs didn't move properly as they were controlled by RL model.

I wrote a detailed article mentioning some more details and workarounds for some of the issues I faced replicating this. https://www.learnwitharobot.com/p/training-a-reinforcement-learning

Thanks for providing the code and references. I was able to train a model and try it out on Bittle. However my trained model doesn't deliver the same smooth gait compared to the pre-existing one in the repository ( trained_agent_PPO.zip) I trained using the exact same code as in the Jupyter notebook at https://github.com/ger01d/opencat-gym/blob/main/opencat-gym.ipynb. I also trained with 5 million steps instead of 2 million in the repo. The result is much poorer tha what trained_agent_PPO gives. I am wondering if trained_agent_PPO was created using some different settings such as a different reward function? Btw, greta job with open sourcing the code. It provides a great example to learn RL.

This is really great news. I look forward to when this new robot arm will come online.

Thanks! Great documentation.

Would really love to replicate the project. Seems that Bittle starts to move once he sees you? Which version of Raspberry Pi did you use for this project? If you are interested in writing a short writeup for this project, I would welcome that at http://www.learnwitharobot.com

Introducing a delay to limit commands being executed helped. I guess there were too many commands issued becuase of changes in the detected color. Btw, whats the purpose of the following line: ==================== tQueue->addTask('i', ""); ====================

Using NyBoard. Will try printing. I think my code is just adding one set of commands to tQueue, but will check.

Linux box is a Desktop computer. Just connected Bittle via USB adapter to the USB port of desktop. Tried all the 4 USB ports just to make sure that it doesn't think /dev/ttyS0 to be one specific port. Yes, standard firmware is on Bittle. Meaning, it functions properly with help of remote control. I pulled latest main from github. Thanks for the link.. I see that it has pictures of checking Bluetooth connectivity in Windows. But the script doesn't work with Bluetooth, or does it? ON my side, I was able to successfully make a Bluetooth connection and see in Linux that the Bluetooth connection is working. However, neitehr the script, nor the Petoi Desktop App (Linux version) are able to use the Bluetooth connection. They all complain about /dev/ttyS0 not connected to Bittle (which I am unable to resolve although I tried to connect Bittle via USB adapter)

Understand. In that case, thanks for spending the money to get this done. It was trully needed for a better experience with Bittle.

I installed v1.0.1 of this app on my iPhone last week and was amazed by the App capabilities. Wondering if the source code of the App can be open sourced... I think that will really help people learning iOS programming as well as enable others to build more interesting features into this app.

This looks like a very cool project. Do you have instructions somewhere?

Many thanks @Paul Masson and @Rongzhong Li. After downgrading to 2.6.1, the test program works. Here is a sample output on the serial monitor for those who might wish to try this.

Worked after I deleted the I2Cdev library. I was following the instructions written under: // I2Cdev and MPU6050 must be installed as libraries Probably I2Cdev need not be separately installed