In this lecture, we will take a look at how FlexBE states are implemented ‘under the hood’: how do they interact with the larger ROS system?

In this video, we have learned the design principles behind FlexBE states:

• Separate your application-specific behavior (implemented in your FLexBE behavior), from the general robot capabilities you implement in ROS nodes.
• FlexBE state implementations should interface with the nodes that implement the capabilities that they coordinate.

A state implementation should:

• Internally execute a specific robotic operation by implementing a client to the robot capability provided by some ROS node.
• Provide an external interface that allows to configure the state for different behaviors:
  • Parameters to configure its static properties for the behavior.
  • Input and output keys to connect to runtime data.
  • Define the set of possible outcomes of the state execution.

In the next videos, we will learn how we can program our own state implementations.