Structure and data communication problems in vision-based Real time application

[turtle]

Hi, all,
I am working on a project that is vision-based real-time application. I have two questiones about the structure and data communication. The system acquires images continuously and process them and send out control signals for motion control. Since it is a real time application, so my basic thought is to develop the code on a host computer and download the executable program to another pc (which is a transferred real-time target with RT engine installed). And i also want to use another pc (with a touchscreen) to run as the front panel. Through indicators and controls, the front panel should be able to communicate with the real-time target pc continuously.

So I wonder in this case, if it is feasible to use one computer as the host pc to develop all the codes and download part of them to the real-time target pc and download the executable program for the front panel to the other pc with touchscreen. After disconnecting the host development pc, these two PCs should still be able to run in real-time and exchange data through shared variable engine?

Regarding the software, i plan to install all the development modules (labview professional, vision development module, real-time development module) on the host development PC. And install RT engine, vision runtime engine on the real-time target PC and labview runtime engine on the other PC with touchscreen.

To drive down the cost, I use PC with touchscreen rather than NI touchscreen panel PC.

I don't know if my thoughts about the structure of the system is right or not and if the data communication between real-time target PC and PC with touchscreen can be done without connection to the host development PC? Please let me know all your thoughts. And regarding the software, the point here is to deploy the real-time application with minimum cost by using as many as possible runtime engines rather than a full package of development module for the two PCs that performs the real-time functions.

Thank you all in advance for any of your thoughts.

Regards,

Wei

[viScience]

Yes, you should be able to proceed as planned.  You can develop and test on your host computer and then build and deploy the RT application to any RT target.  The TP PC should only need the LV Runtime and possibly some other drivers.  As you mentioned you can use the SVE to facilitate binding between the TP PC HMI controls and NSV's residing on your RT target.  This binding can be done with little or no programming.  In the end you will have an autonomous RT PC running your vision and motion app and communicating to your TP PC without any connection to the original development machine.

[turtle]

Thank you so much for your answer, Sachsm.