Ioannis Havoutis

Path planning with force-based foothold adaptation and virtual model control for torque controlled quadruped robots

We present a framework for quadrupedal locomotion over highly challenging terrain where the choice of appropriate footholds is crucial for the success of the behaviour. We use a path planning approach which shares many similarities with the results of the DARPA Learning Locomotion challenge and extend it to allow more flexibility and increased robustness. During execution we incorporate an on-line force-based foothold adaptation mechanism that updates the planned motion according to the perceived state of the environment. This way we exploit the active compliance of our system to smoothly interact with the environment, even when this is inaccurately perceived or dynamically changing, and update the planned path on-the-fly. In tandem we use a virtual model controller that provides the feed-forward torques that allow increased accuracy together with highly compliant behaviour on an otherwise naturally very stiff robotic system. We leverage the full set of benefits that a high performance torque controlled quadruped robot can provide and demonstrate the flexibility and robustness of our approach on a set of experimental trials of increasing difficulty.

Dynamic trot-walking with the hydraulic quadruped robot — HyQ: Analytical trajectory generation and active compliance control

This paper presents a trajectory generator and an active compliance control scheme, unified in a framework to synthesize dynamic, feasible and compliant trot-walking locomotion cycles for a stiff-by-nature hydraulically actuated quadruped robot. At the outset, a CoP-based trajectory generator that is constructed using an analytical solution is implemented to obtain feasible and dynamically balanced motion references in a systematic manner. Initial conditions are uniquely determined for symmetrical motion patterns, enforcing that trajectories are seamlessly connected both in position, velocity and acceleration levels, regardless of the given support phase. The active compliance controller, used simultaneously, is responsible for sufficient joint position/force regulation. An admittance block is utilized to compute joint displacements that correspond to joint force errors. In addition to position feedback, these joint displacements are inserted to the position control loop as a secondary feedback term. In doing so, active compliance control is achieved, while the position/force trade-off is modulated via the virtual admittance parameters. Various trot-walking experiments are conducted with the proposed framework using HyQ, a ~ 75kg hydraulically actuated quadruped robot. We present results of repetitive, continuous, and dynamically equilibrated trot-walking locomotion cycles, both on level surface and uneven surface walking experiments.

Planning and execution of dynamic whole-body locomotion for a hydraulic quadruped on challenging terrain

We present a framework for dynamic quadrupedal locomotion over challenging terrain, where the choice of appropriate footholds is crucial for the success of the behaviour. We build a model of the environment on-line and on-board using an efficient occupancy grid representation. We use Any-time-Repairing A* (ARA*) to search over a tree of possible actions, choose a rough body path and select the locally-best footholds accordingly. We run a n-step lookahead optimization of the body trajectory using a dynamic stability metric, the Zero Moment Point (ZMP), that generates natural dynamic whole-body motions. A combination of floating-base inverse dynamics and virtual model control accurately executes the desired motions on an actively compliant system. Experimental trials show that this framework allows us to traverse terrains at nearly 6 times the speed of our previous work, evaluated over the same set of trials.

Quadrupedal trotting with active compliance

We present a trotting controller for a torque controlled quadruped robot. Our approach uses active compliance to overcome difficulties that are crucial for the realisation of symmetric gaits, i.e. force equalization, disturbance rejection and impact absorption. We present a scheme for the compliant control of each leg that is based on a virtual spring abstraction. This active compliance scheme allows us to greatly vary the dynamical behaviour of the system on-the-fly, without altering the physical characteristics of the robot, by changing the parameters of the virtual springs. This way we are able to evaluate a wide range of trotting gaits with varying parametrizations. We report results of robust trotting in various speeds and push recovery in simulation, and continue with results of actively compliant trotting on the real quadruped robot. We further discuss difficulties and limitations with the implementation of such dynamic gait controllers on the real system.

LOCAL REFLEX GENERATION FOR OBSTACLE NEGOTIATION IN QUADRUPEDAL LOCOMOTION

Legged robots that dynamically locomote through rough terrain need to constantly handle unpredicted collisions (e.g. foot stumbling due to an obstacle) due to the unstructured nature of the environment. If these disturbances are strong enough they can cause errors in the robot’s trunk that are difficult to control with a common feedback-based controller, imposing a serious risk to the overall system stability. The impulsive nature of such disturbances demands a very short reaction time, especially in case of dynamic gaits (trot, gallop, etc.). A quick reaction becomes increasingly crucial when the robot is deprived of reliable visual feedback (e.g. smoky areas or thick vegetation) or when an accurate map of the environment is not available. In this paper we propose a local elevator reflex which enables the robot to reactively overcome high obstacles. The reflex is implemented and experimentally evaluated on the hydraulic quadruped HyQ. We demonstrate the feasibility and effectiveness of our approach showing that the robot is able to step over a platform of 11cm height (14% of the leg length) without prior knowledge of the terrain.

Constrained geodesic trajectory generation on learnt skill manifolds

This paper addresses the problem of compactly encoding a continuous family of trajectories corresponding to a robotic skill, and using this representation for the purpose of constrained trajectory generation in an environment with many (possibly dynamic) obstacles. With a skill manifold that is learnt from data, we show that constraints can be naturally handled within an iterative process of minimizing the total geodesic path length and curvature over the manifold. We demonstrate the utility of this process with two examples. Firstly, a three-link arm whose joint space and corresponding skill manifold can be explicitly visualized. Then, we demonstrate how this procedure can be used to generate constrained walking motions in a humanoid robot.

Synthesising Novel Movements through Latent Space Modulation of Scalable Control Policies

We propose a novel methodology for learning and synthesising whole classes of high dimensional movements from a limited set of demonstrated examples that satisfy some underlying latent low dimensional task constraints. We employ non-linear dimensionality reduction to extract a canonical latent space that captures some of the essential topology of the unobserved task space. In this latent space, we identify suitable parametrisation of movements with control policies such that they are easily modulated to generate novel movements from the same class and are robust to perturbations. We evaluate our method on controlled simulation experiments with simple robots (reaching and periodic movement tasks) as well as on a data set of very high-dimensional human (punching) movements. We verify that we can generate a continuum of new movements from the demonstrated class from only a few examples in both robotic and human data.

Quadruped robot trotting over irregular terrain assisted by stereo-vision

Legged robots have the potential to navigate in challenging terrain, and thus to exceed the mobility of wheeled vehicles. However, their control is more difficult as legged robots need to deal with foothold computation, leg trajectories and posture control in order to achieve successful navigation. In this paper, we present a new framework for the hydraulic quadruped robot HyQ, which performs goal-oriented navigation on unknown rough terrain using inertial measurement data and stereo-vision. This work uses our previously presented reactive controller framework with balancing control and extends it with visual feedback to enable closed-loop gait adjustment. On one hand, the camera images are used to keep the robot walking towards a visual target by correcting its heading angle if the robot deviates from it. On the other hand, the stereo camera is used to estimate the size of the obstacles on the ground plane and thus the terrain roughness. The locomotion controller then adjusts the step height and the velocity according to the size of the obstacles. This results in a robust and autonomous goal-oriented navigation over difficult terrain while subject to disturbances from the ground irregularities or external forces. Indoor and outdoor experiments with our quadruped robot show the effectiveness of this framework.

Onboard perception-based trotting and crawling with the Hydraulic Quadruped Robot (HyQ)

This paper presents a framework developed to increase the autonomy and versatility of a large (~75kg) hydraulically actuated quadrupedal robot. It combines onboard perception with two locomotion strategies, a dynamic trot and a static crawl gait. This way the robot can perceive its environment and arbitrate between the two behaviours according to the situation at hand. All computations are performed on-board and are carried out in two separate computers, one handles the high-level processes while the other is concerned with the low-level hard real-time control. The perception and subsequently the appropriate gait modifications are performed autonomously. We present outdoor experimental trials of the robot trotting over unknown terrain, perceiving a large obstacle, altering its behaviour to the cautious crawl gait and stepping onto the obstacle. This allows the robot to locomote quickly on relatively flat terrain and gives the robot the ability to overcome large irregular obstacles when required.

Motion planning and reactive control on learnt skill manifolds

We address the problem of encoding and executing skills, i.e. motion tasks involving a combination of specifications regarding constraints and variability. We take an approach that is model-free in the sense that we do not assume an explicit and complete analytical specification of the task – which can be hard to obtain for many realistic robot systems. Instead, we learn an encoding of the skill from observations of an initial set of sample trajectories. This is achieved by encoding trajectories in a skill manifold which is learnt from data and generalizes in the sense that all trajectories on the manifold satisfy the constraints and allowable variability in the demonstrated samples. In new instances of the trajectory-generation problem, we restrict attention to geodesic trajectories on the learnt skill manifold, making computation more tractable. This procedure is also extended to accommodate dynamic obstacles and constraints, and to dynamically react against unexpected perturbations, enabling a form of model-free feedback control with respect to an incompletely modelled skill. We present experiments to validate this framework using various robotic systems – ranging from a three-link arm to a small humanoid robot – demonstrating significant computational improvements without loss of accuracy. Finally, we present a comparative evaluation of our framework against a state-of-the-art imitation-learning method.