Efficient and Diverse Generative Robot Designs using Evolution and Intrinsic Motivation

Leni K. Le Goff,Simon C. Smith

2024-11-27

Abstract:Methods for generative design of robot physical configurations can automatically find optimal and innovative solutions for challenging tasks in complex environments. The vast search-space includes the physical design-space and the controller parameter-space, making it a challenging problem in machine learning and optimisation in general. Evolutionary algorithms (EAs) have shown promising results in generating robot designs via gradient-free optimisation. Morpho-evolution with learning (MEL) uses EAs to concurrently generate robot designs and learn the optimal parameters of the controllers. Two main issues prevent MEL from scaling to higher complexity tasks: computational cost and premature convergence to sub-optimal designs. To address these issues, we propose combining morpho-evolution with intrinsic motivations. Intrinsically motivated behaviour arises from embodiment and simple learning rules without external guidance. We use a homeokinetic controller that generates exploratory behaviour in a few seconds with reduced knowledge of the robot's design. Homeokinesis replaces costly learning phases, reducing computational time and favouring diversity, preventing premature convergence. We compare our approach with current MEL methods in several downstream tasks. The generated designs score higher in all the tasks, are more diverse, and are quickly generated compared to morpho-evolution with static parameters.

Robotics

What problem does this paper attempt to address?

### What problems does this paper attempt to solve?

This paper aims to solve the problem of automatically generating efficient and diverse robot designs in complex environments. Specifically, the researchers focus on how to generate innovative and adaptive robot designs through the combination of evolutionary algorithms (EAs) and intrinsic motivation. The following are the main challenges mentioned in the paper:

1. **High computational cost**: Traditional methods such as Morphological Evolution and Learning (MEL) require a learning phase for each generated design, which leads to extremely high computational costs. For example, Gupta et al. [6] required 1,152 CPUs when using deep reinforcement learning (RL) and morphological evolution to generate 4,000 designs.

2. **Premature convergence to sub - optimal solutions**: The MEL framework is prone to premature convergence to local optimal solutions, thus reducing the diversity of designs. Research shows that the learning process exacerbates this premature convergence phenomenon, while a fixed controller can maintain higher design diversity.

To solve these problems, the authors propose a method that combines morphological evolution with intrinsic motivation - **Morphological Evolution and Homeokinesis (MEHK)**. This method utilizes homeokinesis, which can generate exploratory behavior within seconds without in - depth knowledge of the robot design. Homeokinesis quickly assesses the feasibility of a design by maximizing predictability and sensitivity, thereby reducing computational time and increasing design diversity.

### Main contributions

1. **Introduction of homeokinesis**: Apply it to the framework for generating robot designs, including external sensors.
2. **Efficient method for generating diverse designs**: MEHK can quickly generate a large number of robot designs with lower computational resources.
3. **Improvement in downstream task performance**: Compared with morphological evolution combined with a fixed controller, the designs generated by MEHK show higher performance in multiple downstream tasks.

Through these improvements, MEHK can not only generate robot designs more efficiently but also ensure that these designs perform better in actual tasks.