Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Optimizing Modular Robot Composition: A Lexicographic Genetic Algorithm Approach

Jonathan Külz, Matthias Althoff

TL;DR

This work proposes combining a genetic algorithm with a lexicographic evaluation of solution candidates to overcome the problem of identifying an optimal module composition for a specific task and navigate search spaces exceeding those in prior work by magnitudes in the number of possible compositions.

Abstract

Industrial robots are designed as general-purpose hardware with limited ability to adapt to changing task requirements or environments. Modular robots, on the other hand, offer flexibility and can be easily customized to suit diverse needs. The morphology, i.e., the form and structure of a robot, significantly impacts the primary performance metrics acquisition cost, cycle time, and energy efficiency. However, identifying an optimal module composition for a specific task remains an open problem, presenting a substantial hurdle in developing task-tailored modular robots. Previous approaches either lack adequate exploration of the design space or the possibility to adapt to complex tasks. We propose combining a genetic algorithm with a lexicographic evaluation of solution candidates to overcome this problem and navigate search spaces exceeding those in prior work by magnitudes in the number of possible compositions. We demonstrate that our approach outperforms a state-of-the-art baseline and is able to synthesize modular robots for industrial tasks in cluttered environments.

Optimizing Modular Robot Composition: A Lexicographic Genetic Algorithm Approach

TL;DR

This work proposes combining a genetic algorithm with a lexicographic evaluation of solution candidates to overcome the problem of identifying an optimal module composition for a specific task and navigate search spaces exceeding those in prior work by magnitudes in the number of possible compositions.

Abstract

Industrial robots are designed as general-purpose hardware with limited ability to adapt to changing task requirements or environments. Modular robots, on the other hand, offer flexibility and can be easily customized to suit diverse needs. The morphology, i.e., the form and structure of a robot, significantly impacts the primary performance metrics acquisition cost, cycle time, and energy efficiency. However, identifying an optimal module composition for a specific task remains an open problem, presenting a substantial hurdle in developing task-tailored modular robots. Previous approaches either lack adequate exploration of the design space or the possibility to adapt to complex tasks. We propose combining a genetic algorithm with a lexicographic evaluation of solution candidates to overcome this problem and navigate search spaces exceeding those in prior work by magnitudes in the number of possible compositions. We demonstrate that our approach outperforms a state-of-the-art baseline and is able to synthesize modular robots for industrial tasks in cluttered environments.
Paper Structure (14 sections, 10 equations, 3 figures, 1 table)

This paper contains 14 sections, 10 equations, 3 figures, 1 table.

Table of Contents

  1. Introduction
  2. Contribution
  3. Related Work
  4. Problem Statement
  5. Method
  6. Genetic Operators
  7. Fitness Function
  8. Experiments
  9. Robot Modules
  10. Task Definition
  11. Baseline
  12. Setup
  13. Results
  14. Conclusion

Figures (3)

  • Figure 1: Adapting task-tailored modrob compositions leads to optimized setup and production costs: In the example shown above, a small change in the task requirements results in two robots of significantly different complexity. While a six-degree-of-freedom robot is necessary to reach all desired goal positions exactly (a), a relaxation of orientation tolerances allows us to design an modrob with four degrees of freedom only to solve the task (b).
  • Figure 2: We evaluated our algorithm on two synthetic and two manufacturing settings: Synthetic obstacles are shown in red, milling machines and conveyor belts are shown in grey. Every goal consists of a desired position and orientation, indicated by an end effector. The robot base placement is indicated by a red base module.
  • Figure 3: Our approach outperformed the baseline regardless of the weighting factor $w_J$. Especially for our approach, the tradeoff between $n_J$ and $t$ becomes visible. Shaded areas indicate 95% confidence intervals computed using bootstrapping.