Resumen de Behavioral robustness and the distributed mechanisms hypothesis: lessons from bio-inspired and theoretical biology
José A. Fernández León
Theoretical discussions and computational models of bio-inspired embodied and situated agents are presented in this article capturing in simpliied form the dynamical essence of robust and adaptive behavior. The general problem of how dynamical coupling between internal control (brain), body, and environment are exploited in the generation of behavior is particularly analyzed. Using evolutionary algorithms based on Evolutionary Robotics methodology to generate the appropriate neural control, four experiments are introduced to support discussions. The irst model evolves dynamically robust engagements for goal seeking in the presence of neural noise perturbations. The second model develops cognitive-behavioral dependencies for minimal-cognitive behavior in dynamically limited agents. The third one evolves experience-dependent robust behavior in one-legged agent walking. Finally, the last model shows functional dependencies in a mobile-object tracking task. These experiments include a series of structural, sensorimotor, or mutational perturbations, or in the absence of them. Experimental results indicate that neural controls are not suficient to generate robust behavior in each case, suggesting the absence of internal control ‘ensuring’ robustness. The general observation is that coupling dynamics ‘forces’ evolution to behavioral robustness in whatever dynamical form evolution cares to come up with, but relying on behavioral mechanisms that distributes on brain, body, and environment dynamics. Experimental observations provide testable hypothesis that are likely to address in simple organisms in the biological realm, which has some implications for theoretical biology and artiicial systems design.
