Hi everyone,
I’m Joey, a year one student intern from HKU working on the Mini Pupper team.
I’ve been exploring the use of ROS on Raspberry Pi and have found that a single Pi can have limitations. With advancements in generative AI, I believe there is an opportunity to enhance ROS2 projects and reduce hardware demands.
In my latest work, I’ve integrated a camera module and generative AI with ROS2 to implement a line following function on the Mini Pupper 2, a quadruped robot that is based on the CM4 module. The video below showcases this integration.
By leveraging generative AI models, I’ve been able to offload computationally intensive tasks and create a more versatile line following system than what a Raspberry Pi could achieve. For the next steps, I will try to work on recognizing marks and implementing the function of moving around obstacles, and further explore 3D barrier and mapping capabilities.
To help others get started, I’ve also prepared book and website documentation covering the basic functionality. The GitHub repository with the ROS2 code for the Mini Pupper project is available, and I’ll be uploading the line following demo code soon.
I’d love to hear your thoughts on how we can use AI and ROS2 together even more. Leave a comment and let me know what you think!
