Control hierarchy and communication

In previous chapters, we discussed how to map the DoF of a quadruped animal to a robot, and how to collaborate its limbs for posturing and walking. In this chapter, we will discuss the software implementation on its multi-layer control and communication.

We now start from a static hardware "frame player" that moves limbs according to a list of servo angles. By updating the angle list over time, we can generate continuous movement of the robot.

The simplest motion is converting from one static posture to another one. xxxxxxxx

Motion during walking is more challenging. The frames are highly time-dependent with strict geometry constraints. Though we can generate those angles with previously discussed algorithm, it's still challenging to efficiently implement it with limited computational resources.

Ideally, the workflow should look like: