Lamarckian Inheritance in Dynamic Environments: How Key Variables Affect Evolutionary Dynamics

TL;DR

This paper investigates how Lamarckian inheritance influences the evolution of robots in dynamic environments, showing that environmental predictability and conflict levels determine its effectiveness.

Contribution

It demonstrates that the benefits of Lamarckian inheritance depend on environmental variables and introduces a sensor-based approach to enhance its advantages.

Findings

Lamarckian inheritance underperforms when environmental changes are conflicting and unpredictable.

Adding environmental sensors improves Lamarckian inheritance benefits in conflicting environments.

The effectiveness of Lamarckian inheritance varies with environmental predictability and conflict levels.

Abstract

The co-optimization of a robot's body and brain presents a coupled challenge: the morphology constrains which control strategies are effective, while the control determines how well the morphology performs. To address this, we combine morphology optimization as evolution with controller optimization as lifetime learning, utilizing Lamarckian inheritance to transfer learned controller parameters from parent to offspring. In dynamic environments, existing literature presents conflicting evidence: while traditional evolutionary theory often suggests Lamarckian inheritance lacks benefit, recent studies in evolutionary robotics indicate it can improve performance. We hypothesize that this is because previous works have not included all relevant variables with dynamic environments. In this work, we show that the benefit of Lamarckian inheritance depends on two variables: how conflicting the environmental changes are to robot control, and the predictability of those changes for the robotic agent. Using virtual soft robots and two different learning approaches, Bayesian optimization and reinforcement learning, we show that Lamarckian inheritance only underperforms Darwinian inheritance when the changes are both conflicting and unpredictable. We find that adding a sensor to detect environmental changes restores the benefits for Lamarckian inheritance in conflicting environments, by allowing robotic agents to predict the need for a different behavior, thereby generalizing their control.