Factors Impacting Diversity and Effectiveness of Evolved Modular Robots

Abstract

In many natural environments, different forms of living organisms successfully accomplish the same task while being diverse in shape and behavior. This biodiversity is what made life capable of adapting to disrupting changes. Being able to reproduce biodiversity in artificial agents, while still optimizing them for a particular task, might increase their applicability to scenarios where human response to unexpected changes is not possible. In this work, we focus on Voxel-based Soft Robots (VSRs), a form of robots that grants great freedom in the design of both morphology and controller and is hence promising in terms of biodiversity. We use evolutionary computation for optimizing, at the same time, morphology and controller of VSRs for the task of locomotion. We investigate experimentally whether three key factors—representation, Evolutionary Algorithm (EA), and environment—impact the emergence of biodiversity and if this occurs at the expense of effectiveness. We devise an automatic machine learning pipeline for systematically characterizing the morphology and behavior of robots resulting from the optimization process. We classify the robots into species and then measure biodiversity in populations of robots evolved in a multitude of conditions resulting from the combination of different morphology representations, controller representations, EAs, and environments. The experimental results suggest that, in general, EA and environment matter more than representation. We also propose a novel EA based on a speciation mechanism that operates on morphology and behavior descriptors and we show that it allows to jointly evolve morphology and controller of effective and diverse VSRs.

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

“…This way, with proper choices of the simulation environment, evolutionary algorithm, and encoding structures, it is possible to study the contingency level of a biological outcome given specific selection pressures. There are already evolutionary simulation studies in robots that analyze the effects of development ( Corucci et al, 2017 ), environment ( Miras and Eiben, 2019; Pigozzi et al, 2023 ), material properties and environment transitions ( Corucci et al, 2018 ), representation ( Medvet et al, 2021 ) and control ( Cheney et al, 2014 ) on morphologies and behavior. Environment’s parameters as gravity - a constant variable on Earth that is difficult to test experimentally and affects life’s evolutionary outcomes ( Rayner, 2003 ) - can be tested in simulation studies ( Morey-Holton, 2003 ).…”

MorphVAE: Advancing Morphological Design of Voxel-Based Soft Robots with Variational Autoencoders

“…• Diversity: The diversity metric assesses a population's capacity to adapt to dynamic environments. It has been a longstanding evaluation criterion in previous literature to gauge the effectiveness of robot design algorithms (Medvet et al 2021;Pigozzi et al 2023). See our appendix on Github for more details about how we measure diversity.…”

Human Control of Simulated Modular Soft Robots May Predict the Performance of Optimized AI-Based Controllers

“…As robots, we consider Voxel-based Soft Robots (VSRs) [19,20]. They allow for a diverse set of morphologies [21], which are assemblies of homogeneous modules of soft material. Voxels contract or expand their volume following an actuation signal dictated by the controller.…”