Convergent evolution in silico reveals shape and dynamic principles of directed locomotion

Abstract

Active, directed locomotion on the ground is present in many phylogenetically distant species. Bilateral symmetry and modularity of the body are common traits often associated with improved directed locomotion. Nevertheless, both features result from natural selection, which is contingent (history-dependent) and multifactorial (several factors interact simultaneously). Based solely on the unique natural history on Earth, it is difficult to conclude that bilateral symmetry and modularity of the body are required traits for an improved locomotion ability as they can result from chance or be related to other body functions. As a way to avoid these caveats, we propose using physics-based simulations of 3D voxel-based soft robots evolved under different evolutionary scenarios to test the necessity of both traits for efficient directed locomotion on the ground. We found that an intermediate number of body modules (appendages) and high body symmetry are evolutionarily selected regardless of gravitational environments, robot sizes, and genotype encoding. Therefore, we conclude that both traits are strong candidates for universal principles related to efficient directed locomotion.