The Impact of Evolutionary Computation on Robotic Design: A Case Study with an Underactuated Hand Exoskeleton

TL;DR

This paper addresses the challenge of optimizing exoskeleton link lengths to maximize torque transmission under safety constraints, focusing on the underactuated hand exoskeleton U-HEx. It compares brute force optimization with Evolutionary Computation methods, specifically Genetic Algorithms and the Big Bang-Big Crunch algorithm, within both a restricted six variable space and an expanded nine variable space, and uses numerical inverse kinematics to map actuator forces to joint torques. The study demonstrates that EC approaches yield higher or comparable optimality with substantially faster convergence than brute force, enabling additional design variables and revealing differences between GA and BB-BC in convergence speed. The results provide practical guidance for exoskeleton design by showing that EC can meaningfully improve performance while reducing development time, with BB-BC often converging faster and GA offering marginal gains in higher dimensional spaces.

Abstract

Robotic exoskeletons can enhance human strength and aid people with physical disabilities. However, designing them to ensure safety and optimal performance presents significant challenges. Developing exoskeletons should incorporate specific optimization algorithms to find the best design. This study investigates the potential of Evolutionary Computation (EC) methods in robotic design optimization, with an underactuated hand exoskeleton (U-HEx) used as a case study. We propose improving the performance and usability of the U-HEx design, which was initially optimized using a naive brute-force approach, by integrating EC techniques such as Genetic Algorithm and Big Bang-Big Crunch Algorithm. Comparative analysis revealed that EC methods consistently yield more precise and optimal solutions than brute force in a significantly shorter time. This allowed us to improve the optimization by increasing the number of variables in the design, which was impossible with naive methods. The results show significant improvements in terms of the torque magnitude the device transfers to the user, enhancing its efficiency. These findings underline the importance of performing proper optimization while designing exoskeletons, as well as providing a significant improvement to this specific robotic design.

Figures (4)

• Figure 1: U-HEx: (a) A user wearing the first prototype of U-HEx from the state-of-the-art, which is bulky and cumbersome sarac2017design. Note that the picture only shows the device with a single finger. (b) Schema of the comparison between the approaches in terms of run time and optimality of the solution.
• Figure 2: Kinematic model of U-HEx depicting all link lengths, passive joints (rotational and linear), and the active linear actuator. The colored links represent the decision variables of the optimization problem (i.e., the links to be optimized): in yellow, the ones that were used in its original design sarac2017design; in green, the additional three that were included in the current work thanks to evolutionary computation. The rest of the links are set to fixed values to ensure the kinematic chain is closed. Only joint $O$ is actuated.
• Figure 3: Comparison between results of Experiments 1 and 2. Plots include median, interquartile range, and outliers. Details are zoomed in to appreciate variance amongst different conditions.
• Figure 4: CAD models of U-HEx with the lengths retrieved from (a) BF, similar to the original prototype sarac2017design and (b) EC -- specifically BB-BC.