Lizards...

...co-design of robotic lizards at ETH Robotic Systems Lab


The video shows two sessions of a Tree-structured Parzen Estimator-based Bayesian Optimization for single-sized legged lizards to optimize the parameters of a "Hildebrand-like" gait generating controller.


This project is still ongoing. As it just started less than 2 months ago, many details are still evolving while I am writing this description and cannot be described furthermore here.

The project is a collaboration between Prof. Marco Hutter and Prof. Baxi Chong (my former supervisor at Dan Goldman’s lab). My supervisor at ETH is Postdoc Robert Baines. The objective of this project is to study and optimize through successful co-design (optimizing the robot’s morphology and control policy at the same time) lizard locomotion through robotics platforms, both in sim and in real. I am leading the sim/learning aspect using a IsaacLab-based framework. Simultaneously, at Penn, Baxi’s team is testing the hypotheses made on the real robot.


Performance difference of multiple-sized legged lizards using the exact same non-optimized gait.