This two-part paper discusses the analysis and control of legged locomotion in terms of N-step capturability: the ability of a legged system to come to a stop without falling by taking N or fewer steps. We consider this ability to be crucial to legged locomotion and a useful, yet not overly restrictive criterion for stability. Part 1 introduced the N-step capturability framework and showed how to obtain capture regions and control sequences for simplified gait models. In Part 2, we describe an algorithm that uses these results as approximations to control a humanoid robot. The main contributions of this part are (1) step location adjustment using the 1-step capture region, (2) novel instantaneous capture point control strategies, and 3) an experimental evaluation of the 1-step capturability margin. The presented algorithm was tested using M2V2, a 3D force-controlled bipedal robot with 12 actuated degrees of freedom in the legs, both in simulation and in physical experiments. The physical robot was able to recover from forward and sideways pushes of up to 21 Ns while balancing on one leg and stepping to regain balance. The simulated robot was able to recover from sideways pushes of up to 15 Ns while walking, and walked across randomly placed stepping stones.

Capturability-based analysis and control of legged locomotion, Part 2: Application to M2V2, a lower-body humanoid

Walking is the most common terrestrial form of locomotion in animals. Its great versatility and flexibility has led to many attempts at building walking machines with similar capabilities. The control of walking is an active research area both in neurobiology and robotics, with a large and growing body of work. This paper gives an overview of the current knowledge on the control of legged locomotion in animals and machines and attempts to give walking control researchers from biology and robotics an overview of the current knowledge in both fields. We try to summarize the knowledge on the neurobiological basis of walking control in animals, emphasizing common principles seen in different species. In a section on walking robots, we review common approaches to walking controller design with a slight emphasis on biped walking control. We show where parallels between robotic and neurobiological walking controllers exist and how robotics and biology may benefit from each other. Finally, we discuss where research in the two fields diverges and suggest ways to bridge these gaps.

http://iopscience.iop.org/article/10.1088/1748-3190/10/4/041001/pdf

Controlling legs for locomotion—insights from robotics and neurobiology

This PhD thesis covers simulation and control of biped walking robots. Simulations and experiments were performed with the robots Johnnie and Lola developed at the Institute of Applied Mechanics, Technische Universitat Munchen, Germany.

The chapter on modeling and simulation describes the approach to simulating the rigid multibody dynamics. It includes specialized models for simulating drive friction, nonlinear drive kinematics and the unilateral ground contact. Component models, contact and differential equation solvers are combined to a family of robot simulations with varying modeling depth and computational cost.

The thesis also presents a hierarchical system for real-time walking control. Novel aspects include a trajectory generator based on spline collocation and a stabilizing controller based on hybrid position/force control. Lola’s redundant toe and pelvis joints are used to reduce joint speeds and avoid joint limits. 

All methods were verified on both robots in walking experiments and simulations. In experiments Lola reached a maximum walking speed of 3.34 km/h. using a computer vision system developed at the Institute of Autonomous Systems Technology at the Universitat der Bundeswehr in Munich, Germany, Lola is also capable of safely exploring previously unknown environments.

https://mediatum.ub.tum.de/doc/997204/file.pdf

Simulation and Control of Biped Walking Robots

Despite several decades of research, locomotion of bipedal robots is still far from achieving the graceful motions and the dexterity observed in human walking. Most of today’s bipeds are controlled by analytical approaches based on multibody dynamics, pre-calculated joint trajectories, and Zero-Moment Point considerations to ensure stability. However, beside their considerable achievements these methods show several drawbacks like strong model dependency, high energetic and computational costs, and vulnerability to unknown disturbances. In contrast to this, human locomotion is elegant, highly robust, fast, and energy efficient. These facts gave rise to the main hypothesis of this thesis, namely that a control system based on insights into human motion control can yield human-like walking capabilities in two-legged robots. This thesis thus presents a control methodology for bipeds relying heavily on the transfer of concepts found in the locomotion control of humans. Based on a thorough review on biomechanics and neuroscience literature, a control approach is derived that can achieve dynamic, efficient, and robust walking of three-dimensional and fully articulated bipeds. Being located above the neural level, the control system is structured as a hierarchical network of local feed-forward and feedback units, without using a complete dynamic model or pre-calculated joint trajectories. Sensor event-based spinal pattern generators coordinate the stimulation and synchronization of control units and the compliance of passive joints. By applying local torque commands instead of joint angle control, passive dynamics and self-stabilizing effects of elasticities can be exploited. Postural control is achieved by the phase-dependent activity of several reflexes of various complexity. The suggested approach is tested in a full-featured dynamics simulation framework on an anthropomorphic biped with 21 degrees of freedom and human-like morphology, weight, and actuation. The control system can achieve three-dimensional dynamic walking of variable velocity as well as balanced standing. It is able to cope with the high complexity and the mechanical elasticities of the modeled biped. The emerging, naturally looking walking gait shows remarkable similarities to human walking. Simultaneous control of only a subset of joints is sufficient to direct the passive dynamics of the robot towards a walking gait. The achieved walking velocity of up to 5 km/h can compete with even the most advanced of today’s bipeds. At the same time, energy efficiency is much better than in joint angle controlled robots. The control system shows considerable robustness against unknown and unexpected disturbance like steps, slopes, or external forces.

Human-like Control of Dynamically Walking Bipedal Robots

This work reports on the modelling of a compliant humanoid robot. The introduction of intrinsic compliance in some of the joints of the COmpliant huMANoid robot (COMAN) affects its dynamics making no trivial the generation of walking gaits. To permit the development of effective gait generators which take into account these additional dynamic effects due to intrinsic compliance, an appropriate model which can predict the robot motion dynamics is required. To predict the motion of the centre of mass (COM) of the compliant robot we propose a new model for compliant humanoid systems which combines the inverted pendulum model approach with a three dimensional compliant model (Cartesian) at the level of the COM. The derivation of the model is introduced followed by experimental validation which demonstrates the adequate performance achieved by the proposed reduced model. In particular the efficacy of the model was experimentally validated on the COMAN humanoid platform using a series of ZMP based walking gaits.

A compact model for the compliant humanoid robot COMAN

This video highlights work to date on the Yobotics-IHMC Lower Body Humanoid Robot. The robot is a twelve degree-of-freedom robot with force controllable Series Elastic Actuators at each degree of freedom. Control algorithms utilize Virtual Model Control, and foot placement is determined using Capture Regions. The robot can recover from moderate disturbances and walk on flat ground. Ongoing work is focused on improving robustness to disturbances, walking more quickly and efficiently, and walking over rough terrain.

/pdf/the-yobotics-ihmc-lower-body-humanoid-robot-4tvt7u2ay9.pdf

Peter Neuhaus

Papers

The Yobotics-IHMC Lower Body Humanoid Robot