Crossing the reality gap in evolutionary robotics by promoting transferable controllers

The reality gap, that often makes controllers evolved in sim-ulation inefficient once transferred onto the real system, re-mains a critical issue in Evolutionary Robotics (ER); it pre-vents ER application to real-world problems. We hypothe-size that this gap mainly stems from a conflict between the efficiency of the solutions in simulation and their transfer-ability from simulation to reality: best solutions in simula-tion often rely on bad simulated phenomena (e.g. the most dynamic ones). This hypothesis leads to a multi-objective formulation of ER in which two main objectives are opti-mized via a Pareto-based Multi-Objective Evolutionary Al-gorithm: (1) the fitness and (2) the transferability. To eval-uate this second objective, a simulation-to-reality disparity value is approximated for each controller. The proposed method is applied to the evolution of walking controllers for a real 8-DOF quadrupedal robot. It successfully finds effi-cient and well-transferable controllers with only a few ex-periments in reality