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

2026-05-15

AI Summary

This study addresses the challenges posed by environmental dynamics to the co-evolution of robot morphology and control, where conflicting and unpredictable environmental changes limit the efficacy of Lamarckian inheritance. The authors integrate evolutionary optimization for morphology with lifelong learning for controller adaptation, employing a simulated soft robot platform to compare Lamarckian and Darwinian inheritance strategies under Bayesian optimization and reinforcement learning. Their findings reveal that the performance of Lamarckian inheritance critically depends on both the conflictuality and predictability of environmental changes: it underperforms relative to Darwinian mechanisms only when these two factors coexist. Notably, equipping robots with environmental sensors enables Lamarckian inheritance to significantly enhance adaptability in conflicting environments, thereby restoring its evolutionary advantage.

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.

Problem

• Lamarckian inheritance
• dynamic environments
• evolutionary robotics
• environmental predictability
• morphology-control co-optimization

Innovation

• Lamarckian inheritance
• dynamic environments
• co-optimization
• predictability
• evolutionary robotics