Vision Guided Robotics With Supervisory PLC

[Swervomotor]

I'm currently working on a project involving a Unistream 15.6 as a supervisory PLC  for 2 Epson Robots (a 6 axis and a 4 axis) and a keyence vision system for vision guided robotics. They are communicating over Ethernet/IP

I'm wondering if anyone on here has experience with programming practices for how to handle the robot control in this situation. 

Setup:

Unistream-  Master

Robot 1 - Slave

Robot 2- Slave

Vision system- Slave

First idea of data flow and control:   PLC triggers camera----->vision returns position data in a register (pre-calibrated to robot) to PLC-----> PLC writes position data to multiple registers in robot controller which the robot controller then stores as a Point. ----->PLC turns bit on for robot to move to position----->robot returns bit when at the location and writes current location to register (So PLC knows where it is at all the time)-----PLC sends next point data------->repeat----->repeat

This method seems to have very little programming on the robot controller itself and would allow data tables stored on the PLC vs Point data on the robot controller.  Draw backs are when Jog/teaching new points via the robot controller the data would have to be entered into the data table on the PLC manually.

 

Second Idea is to have multiple sub-routines in Robot controllers and "call' them with bits from the PLC, the vision guided position data would still need to be transferred for vision--->PLC--->robot

but all the other point data such as way points and place points would be stored and managed on the robot controller.

PLC triggers camera-->vision position data sent to PLC--->PLC sends position data to Robot---> robot picks part and moves to several points independently then turns on a bit when it finishes. 

 

What are your thoughts on this?