Dorian Scholz

CONCEPT AND DESIGN OF THE BIOBIPED1 ROBOT FOR HUMAN-LIKE WALKING AND RUNNING

Biomechanics research shows that the ability of the human locomotor system depends on the functionality of a highly compliant motor system that enables a variety of different motions (such as walking and running) and control paradigms (such as flexible combination of feedforward and feedback controls strategies) and reliance on stabilizing properties of compliant gaits. As a new approach of transferring this knowledge into a humanoid robot, the design and implementation of the first of a planned series of biologically inspired, compliant, and musculoskeletal robots is presented in this paper. Its three-segmented legs are actuated by compliant mono- and biarticular structures, which mimic the main nine human leg muscle groups, by applying series elastic actuation consisting of cables and springs in combination with electrical actuators. By means of this platform, we aim to transfer versatile human locomotion abilities, namely running and later on walking, into one humanoid robot design. First experimental results for passive rebound, as well as push-off with active knee and ankle joints, and synchronous and alternate hopping are described and discussed. BioBiped1 will serve for further evaluation of the validity of biomechanical concepts for humanoid locomotion.

A new biarticular actuator design facilitates control of leg function in BioBiped3

Bioinspired legged locomotion comprises different aspects, such as (i) benefiting from reduced complexity control approaches as observed in humans/animals, (ii) combining embodiment with the controllers and (iii) reflecting neural control mechanisms. One of the most important lessons learned from nature is the significant role of compliance in simplifying control, enhancing energy efficiency and robustness against perturbations for legged locomotion. In this research, we investigate how body morphology in combination with actuator design may facilitate motor control of leg function. Inspired by the human leg muscular system, we show that biarticular muscles have a key role in balancing the upper body, joint coordination and swing leg control. Appropriate adjustment of biarticular spring rest length and stiffness can simplify the control and also reduce energy consumption. In order to test these findings, the BioBiped3 robot was developed as a new version of BioBiped series of biologically inspired, compliant musculoskeletal robots. In this robot, three-segmented legs actuated by mono- and biarticular series elastic actuators mimic the nine major human leg muscle groups. With the new biarticular actuators in BioBiped3, novel simplified control concepts for postural balance and for joint coordination in rebounding movements (drop jumps) were demonstrated and approved.

Simulation and experimental evaluation of the contribution of biarticular gastrocnemius structure to joint synchronization in human-inspired three-segmented elastic legs

The humanoid robot BioBiped2 is powered by series elastic actuators (SEA) at the leg joints. As motivated by the human muscle architecture comprising monoarticular and biarticular muscles, the SEA at joint level are supported by elastic elements spanning two joints. In this study we demonstrate in simulation and in robot experiments, to what extend synchronous joint operation can be enhanced by introducing elastic biarticular structures in the leg, reducing the risk of over-extending individual joints.

Bio-inspired motion control of the musculoskeletal BioBiped1 robot based on a learned inverse dynamics model

Based on the central hypothesis that a humanoid robot with human-like walking and running performance requires a bio-inspired embodiment of the musculoskeletal functions of the human leg as well as of its control structure, a bio-inspired approach for joint position control of the BioBiped1 robot is presented in this paper. This approach combines feed-forward and feedback control running at 1 kHz and 40 Hz, respectively. The feed-forward control is based on an inverse dynamics model which is learned using Gaussian process regression to account for the robots body dynamics and external influences. For evaluation the learned model is used to control the robot purely feed-forward as well as in combination with a slow feedback controller. Both approaches are compared to a basic feedback PD-controller with respect to their tracking ability in experiments. It is shown, that the combined approach yields good results and outperforms the basic feedback controller when applied to the same set-point trajectories for the leg joints.

Efficient design parameter optimization for musculoskeletal bipedal robots combining simulated and hardware-in-the-loop experiments

The design and tuning of bio-inspired musculoskeletal bipedal robots with tendon driven series elastic actuation (TD-SEA) including biarticular structures is more complex than for conventional rigid bipedal robots. To achieve a desired dynamic motion goal additional hardware parameters (spring coefficients, rest lengths, lever arms) of both, the TD-SEAs and the biarticular structures, need to be adjusted. Furthermore, the biarticular structures add correlations over multiple joints which increase the complexity of tuning of these parameters. Parameter adaption and tuning is needed to fit active and passive dynamics of the actuators and the control system. For the considered class of musculoskeletal bipedal robots no fully satisfying systematic approach to efficiently tune all of these parameters has been demonstrated yet. Conventional approaches for tuning of hardware parameters in rigid robots are either simulation based or use a hardware-in-the-loop optimization. This paper presents a new approach to efficiently optimize these parameters, by combining the advantages of simulation-in-the-loop and hardware-in-the-loop optimizations. Grahical interpretation of suitable metrics, like resulting quality values, are used to interpret the simulation results in order to efficiently guide the hardware experiments. By carefully considering the simulation results and adjusting the sequence of robot experiments based on biomechanical insights, the required number of hardware experiments can be significantly reduced. This approach is applied to the musculoskeletal BioBiped2 robot where the hardware parameters of the elastic actuation of the Gastrocnemius and Soleus structures are optimized. A comparison with a state-of-the-art hardware-in-the-loop optimization method demonstrates the efficiency of the presented approach.

SIMPLE YET EFFECTIVE TECHNIQUE FOR ROBUST REAL-TIME INSTABILITY DETECTION FOR HUMANOID ROBOTS USING MINIMAL SENSOR INPUT

Legged locomotion of autonomous humanoid robots is advantageous but also challenging since it inherently suffers from high posture instability. External disturbances such as collisions with other objects or robots in the environment can cause a robot to fall. Many of the existing approaches for instability detection and falling prevention include a large number of sensors resulting in complex multi-sensor data fusion and are not decoupled from the walking motion planning. Such methods can not simply be integrated into an existing low-level controller for real-time motion generation and stabilization of a humanoid robot. A procedure that is both easily implementable using a minimal number of affordable sensors and capable of reliable detection of posture instabilities is missing to date. We propose a simple, yet reliable balance control technique consisting of a filtering module for the used data from two-axes-gyroscopes and -accelerometers located at the trunk, an instability classification algorithm, and a lunge step module. The modules are implemented on our humanoid robots which participate at the yearly RoboCup competitions in the humanoid kid-size league of soccer playing robots. Experimental results show that the approach is suited for real-time operation during walking.