The goal of the work presented here is to influence the overall behaviour of specific animal societies by integrating computational mechatronic devices (robots) into those societies. To do so, these devices should be accepted by the animals as part of the society and/or as part of the collectively formed environment. For that, we have developed two sets of robotic hardware for integrating into societies of two different animals: zebrafish and young honeybees. We also developed mechanisms to provide feedback from the behaviours of societies for the controllers of the robotic system. Two different computational methods are then used as the controllers of the robots in simulation and successfully adapted by evolutionary algorithms to influence the simulated animals for desired behaviours. Together, these advances in mechatronic hardware, feedback mechanisms, and controller methodology are laying essential foundations to facilitate experiments on modulating self-organised behaviour in mixed animal–robot societies.