Closed ghost closed 1 year ago
An example of the output from rqt_graph. This was generated on my laptop (not the rover) so the actual input/output nodes are missing. That is why the Drive_Board_Hardware
namespace only has the 6 channels used by the Drivetrain_Ouptut
node.
Here's the version generated from the rover when running. I set it to only show active topics, so the PWM drivers don't show all 16 channels.
The only nodes that are not relocatable are the *_Interface
nodes (relocating them breaks the base station interface) and the nodes in the *_Board_HW
namespaces (relocating them requires re-wiring the rover. Everything else can be run on any board (functionally) As we get closet to a complete system, we will want to run CPU load analysis and see if any boards would benefit from re-balancing the system. We could go to a hardware-based division: Each board is responsible for a set of hardware (GPS, PWM, etc.) instead of each board handling the control of a single subsystem.
Another Update:
I would like to put this into visio, powerpoint, or draw.io - but it has been changing somewhat rapidly as I add new nodes. Once we get more into testing/bugfixing, it will be more stable.
@jensenm7
With some changes to the drive system:
We need to generate documentation for the structure of the system:
These may be combined/blurred some. This is not a one time task. We should probably have these for each milestone.