Are the worker nodes written in a way that would let them be kicked off by something like OpenCue (or other more comprehensive farm management system), maybe with a small wrapper around them? Or would this need to be run in parallel with its own server and machine allocations?