It is widely thought that sensorimotor synchronization, underpinning cultural domains such as music and dance, played a critical role in the evolution of human sociality. Here, we present virtual legged robots controlled by central pattern generators (CPGs) that evolve to synchronize motion to rhythmic sensory input in real time. Multi-stage, multi-objective evolutionary algorithms were used to maximize flexibility of the CPGs with respect to control parameters, and then to optimize a neural input layer for wide-ranging susceptibility to rhythmic inputs. The evolved CPGs self-organize to accommodate the input sequence over a range of frequencies and patterns while keeping the agents upright. We show how this behaviour can be scaled up to multiple interacting agents, including with differing morphologies, to produce novel behaviours. We then outline how spike timing dependent plasticity can be used for the acquisition of new motor patterns. Finally, taking inspiration from biocultural evolution and cognitive neuroscience, we suggest ways in which real-time social adaptation can play a key role in the evolution of complex social behaviours in robots.

There are many challenges pertaining to how one should jointly evolve the morphology and controllers of robots and virtual creatures. Innervation between decentralized control approaches can allow for coordinated rhythmic movement in organisms, and this can therefore be beneficial when evolving the bodies and brains of virtual creatures. To test how decentralized control could be beneficial when evolving the morphology and control of 2D virtual creatures, three open-loop decentralized control schemes were compared for their effectiveness: (1) a simple sinusoidal wave generator, (2) a phase-coupled oscillator and (3) a neural network. The latter two controllers could innervate to descending controllers enabling the expression of coordinated movement. In addition, the performance of the controllers were compared when the creatures were made through either a direct or indirect encoding. The results show that a phase-coupled oscillator gives significantly better performance than a simple wave when using either of the two encodings. The neural network approach performed somewhere in-between both controller approaches, although seeding an evolving population with manually designed neural networks improved the performance especially for the direct encoding. Controller modulation through descending innervation can lead to coordinated movements that can benefit decentralized control strategies when evolving the morphology and control of virtual creatures.