Socially interacting robots are widely envisioned to revolutionise learning and care, and as they become more ubiquitous, to perform tasks collaboratively. However, despite astounding progress in engineering adaptive capabilities in fundamental behaviours such as locomotion, social behaviour is still generally considered a separate 'add-on' component. COROBOREES will contribute to efforts to provide a bottom-up approach to social behaviour in automated systems. Combining methods from evolutionary robotics, collective behaviour and sound and music computing, the project will investigate ways to incorporate social information into the motion controllers of virtual agents, and to study how this can enable synchronisation of complex movement patterns. Specifically, the project will build on the central pattern generator (CPG) framework for robot motion. CPGs are widely used in robotic controllers for cyclic motions such as walking, and may include feedback mechanisms so that robots can adapt to new tasks and environments. Using evolutionary methods, COROBOREES will compare mechanisms for entrainment to rhythmic acoustic patterns as well as acoustically transmitted social cues. Methods from sound and music computing will be used to evaluate the complexity and synchronisation of the collective rhythms generated. In addition to the innovation within robotics, this work will result in a unique simulation platform for collective behaviour, complete with built-in sensorimotor mechanisms, for the use of researchers in artificial life, biology and neuroscience. The broadly applicable training-through-research nature of the action, the contact with numerous fields of research, as well as a tailored and extensive focus on transferable skills, will provide the Experienced Researcher with invaluable knowledge and experience for leading applied interdisciplinary research in the future.