Abstract
Artificial evolution of physical systems is a stochastic optimization method in which physical machines are iteratively adapted to a target function. The key for a meaningful design optimization is the capability to build variations of physical machines through the course of the evolutionary process. The optimization in turn no longer relies on complex physics models that are prone to the reality gap, a mismatch between simulated and real-world behavior. We report model-free development and evaluation of phenotypes in the artificial evolution of physical systems, in which a mother robot autonomously designs and assembles locomotion agents. The locomotion agents are automatically placed in the testing environment and their locomotion behavior is analyzed in the real world. This feedback is used for the design of the next iteration. Through experiments with a total of 500 autonomously built locomotion agents, this article shows diversification of morphology and behavior of physical robots for the improvement of functionality with limited resources.
Highlights
The adaptation of physical shapes and structures is a fundamental mechanism which allows biological systems to survive in a large variety of environments
Through evolutionary adaptation some animals changed their morphologies to live on land instead of under water, and phenotypic plasticity allows plants to adapt their structures for the survival on an ontogenetic time-scale [1]
Variations of designs are built during evolution and an increase of performance can be observed after a few generations
Summary
The adaptation of physical shapes and structures is a fundamental mechanism which allows biological systems to survive in a large variety of environments. Through evolutionary adaptation some animals changed their morphologies to live on land instead of under water, and phenotypic plasticity allows plants to adapt their structures for the survival on an ontogenetic time-scale [1]. An engineered counterpart of evolutionary adaptation was investigated in the past [5, 6, 7]. The co-optimization of body and mind was demonstrated with the simulated evolution of virtual animal-like creatures [8, 9, 10, 11], which exploited evolutionary dynamics to generate variations of mechanical bodies as well as motor control circuitry for meaningful dynamic behaviors such as walking, running and swimming.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
