Swarms of Cyber-Physical Systems (CPSs) can be used to tackle many challenges that traditional multi-robot systems fail to address. In particular, the self-organizing nature of swarms ensures they are both scalable and adaptable. Such benefits come at the cost of having a complex system that is extremely hard to design manually. Therefore, an automated process is required for designing the local interactions between the agents that lead to the desired swarm behavior. In this work, the authors employ evolutionary design methodologies to generate the local controllers of the agents. This requires many simulation runs and, as a consequence, distributed simulation. The paper first proposes a network-based Application Programming Interface (API) that employs a publish/subscribe broker architecture to distribute simulations among multiple Simulation Servers (SSs). Following this, a file-based API is proposed, which exports the agent controller to the simulator enabling deployment of the evolved solution on CPSs. Both approaches are compared in terms of time needed for the evolutionary optimization process with the support of simulations. A proof of concept demonstrates the portability to CPSs using TurtleBot robots. The results suggest that for most scenarios it is beneficial to export the agent controller to the simulator to avoid the vast communication overhead. The presented network-based approach currently lacks this feature but is well suited to offload computation-heavy simulations to a cluster of SSs.