I’ve made a ROS wrapper for the Pimoroni that lets you control your Pimoroni from ROS. I’ve also included a Gazebo simulation of a Pimoroni STS PI so you can record the movements of your Pimoroni with ROSBAG and then replay them in a Gazebo simulation like in the video below:
In the video I have a real Pimoroni who’s movements are recorded and also a simulated Pimoroni in a virtual representation of my flat which I replay the movements to. Gazebo has a good physics engine that simulates the Pimoroni well.
You can get the source at:
The aim of the project is to do reinforcement learning (Q learning0 with a Pimoroni trying to teach it to do tasks like find objects. I will train it within Gazebo and when it works deploy it to a real Pimoroni. I’ll post this project if I am successful.