Week 4/5 – Controlling the robots
In the fourth and fifth week, I focused on knowing why the robots wheren’t moving.
As I didn’t know why it wasn’t working, I clone Eva Fernandez tfg repository because it contains a Nao able to move. After changing some of the routes in the code, I was able to make it move:
With this working, I look what I was missing in my repositories, and I founded that the unitree models doesn’t have joint controllers, and that waas why they wheren’t responding.
After that, I decided that I needed new models, so we founded this go2 model made by khaledgabr77.
This new version of the go2 had everything it needed to walk, despite I don’t now yet how is done:
The h1 robot is in stand by for the moment, as we hand’t found any alternative, because as we can use the Nao as a humanoid, we also want bo be able to use the arms, something that can barely be done with the Nao.
In another note, I’ve started looking at how to use Mujoco sim, as is were we plan to train the model.
Summary
- Replicating Eva Fernandez tfg
- Obtaining a new go2 model
- Looking at mujoco