Robotics/Computer Control/Control Architectures/Sense-Plan-Act

From Wikibooks, open books for an open world
Jump to: navigation, search

As obvious from the term, this architecture works through a cycle of Sense, Plan and Act processes.

  • By sensing, it means getting the required information from the available sensors and converting it to some usable form.
  • Planning refers to the use of available information from sensing phase to determine the control parameters and sequences required for various components in order to make the robot proceed towards final goal.
  • Finally, act phase is simply the implementation of the processes and sequences underlined by the planning phase.

This kind of approach is useful if the sensing process is slow and the environment is quite static. For example, on a low speed processor, image processing may require lots of time to process each frame. Hence, it might not be feasible to process every frame and react to the gathered information in real-time. If we use this technique, the information will be processed at a very low frequency, providing more time for other real-time processes to run while still doing the work of planning and acting towards the goal.