The objective of this study was to evaluate the performance a magnetorheological-fluid-based variable stiffness actuator leg under high impact forces through optimal tuning and control damping properties. To achieve this, drop testing experiments were conducted with at various heights payload masses. results showed that while lower higher can lead greater energy dissipation, respectively, also ...

Hierarchical learning has been successful at generalizable locomotion skills on walking robots in a sample-efficient manner. However, the low-dimensional “latent” action used to communicate between two layers of hierarchy is typically user-designed. In this letter, we present fully-learned hierarchical framework, that capable jointly low-level controller and high-level latent space. Once space ...

The Theo Jansen mechanism is gaining widespread popularity among the legged robotics community due to its scalable design, energy efficiency, low payload-to-machine-load ratio, bioinspired locomotion, and deterministic foot trajectory. In this paper, we perform for the first time the dynamic modeling and analysis on a four-legged robot driven by a single actuator and composed of Theo Jansen mec...

This paper describes a neural learning architecture for control of legged robots inspired by mammalian neurophysiology. Biological studies indicate that the cerebel-lum is a key part of an adaptive control system which enables mammals to display remarkable limb coordination during loco-motion. We present a distributed control system using reinforcement learning methods and mechanisms inspired b...

In this paper we investigate a biological framework to generate and adapt a motion pattern so that can be energy efficient. In fact, the motion pattern in legged animals and human emerges among interaction between a central pattern generator neural network called CPG and the musculoskeletal system. Here, we model this neuro - musculoskeletal system by means of a leg - like mechanical system cal...

In this work, we present a nonlinear dynamics perspective on generating and connecting gaits for energetically conservative models of legged systems. particular, show that the set constitutes connected space locally defined 1D submanifolds in gait space. These manifolds are coordinate-free parameterized by energy level. We algorithms identifying such families through use numerical continuation ...

Legged robots are popular candidates for missions in challenging terrains due to their versatile locomotion strategies. Terrain classification is a key enabling technology autonomous legged robots, allowing them harness innate flexibility adapt the demands of operating environment. We show how highly capable machine learning techniques, namely gated recurrent neural networks, allow our target r...

With the introduction of commercially available programmable legged robots, a generic software method for detection of abnormalities in the robots’ locomotion is required. Our approach is to gain satisfactory results using a bare minimum amount of hardware feedback; In most cases we are able detect faults using only the joint angle sensors. Methods for recognising several types of collision as ...