Search

The 11th International Symposium on Adaptive Motion of Animals and Machines. Kobe University, Japan. 2023-06-06/09. Adaptive Motion of Animals and Machines Organizing Committee., Poster Session P71

These datasets were used in the paper "Exploring the Contribution of Proprioceptive Reflexes to Balance Control in Perturbed Standing". the different database files contain information about the progress of the optimizations that were performed. The zip files contain txt files with joint angles, joint velocities, joint moments, ground reaction forces, muscle activations, lengths, and forces. Base refers to the base model, TD refers to the base model with different time delay, length refers to the controller with length feedback only, force refers to the controller with force feedback only, and com refers to the model with time delay on the centre of mass.

In this paper, we present a new walking controller based on 3LP model. Taking advantage of linear equations and closed-form solutions of 3LP, the proposed controller can project the state of the robot at any time during the phase back to a certain event for which, a discrete LQR controller is designed. After the projection, a proper control policy is generated by the expert discrete controller and used online. This projecting architecture reacts to disturbances with minimal delay and compared to discrete controllers, it provides superior performance in recovering intermittent external pushes. Further analysis of closed-loop eigenvalues and disturbance rejection shows that the proposed time-projecting controller has strong stabilization properties. Controllable regions also show that the projecting architecture covers most of the maximal controllable set of states. It is computationally much faster than model predictive controllers, but still optimal., Comment: Journal paper under review

Reconstructing the locomotion of extinct vertebrates offers insights into their palaeobiology and helps to conceptualize major transitions in vertebrate evolution. However, estimating the locomotor behaviour of a fossil species remains a challenge because of the limited information preserved and the lack of a direct correspondence between form and function. The evolution of advanced locomotion on land—that is, locomotion that is more erect, balanced and mechanically power-saving than is assumed of anamniote early tetrapods—has previously been linked to the terrestrialization and diversification of amniote lineages. To our knowledge, no reconstructions of the locomotor characteristics of stem amniotes based on multiple quantitative methods have previously been attempted: previous methods have relied on anatomical features alone, ambiguous locomotor information preserved in ichnofossils or unspecific modelling of locomotor dynamics. Here we quantitatively examine plausible gaits of the stem amniote Orobates pabsti, a species that is known from a complete body fossil preserved in association with trackways. We reconstruct likely gaits that match the footprints, and investigate whether Orobates exhibited locomotor characteristics that have previously been linked to the diversification of crown amniotes. Our integrative methodology uses constraints derived from biomechanically relevant metrics, which also apply to extant tetrapods. The framework uses in vivo assessment of locomotor mechanics in four extant species to guide an anatomically informed kinematic simulation of Orobates, as well as dynamic simulations and robotics to filter the parameter space for plausible gaits. The approach was validated using two extant species that have different morphologies, gaits and footprints. Our metrics indicate that Orobates exhibited more advanced locomotion than has previously been assumed for earlier tetrapods, which suggests that advanced terrestrial locomotion preceded the diversification of crown amniotes. We provide an accompanying website for the exploration of the filters that constrain our simulations, which will allow revision of our approach using new data, assumptions or methods.

We present a new walking foot-placement controller based on 3LP, a 3D model of bipedal walking that is composed of three pendulums to simulate falling, swing and torso dynamics. Taking advantage of linear equations and closed-form solutions of the 3LP model, our proposed controller projects intermediate states of the biped back to the beginning of the phase for which a discrete LQR controller is designed. After the projection, a proper control policy is generated by this LQR controller and used at the intermediate time. This control paradigm reacts to disturbances immediately and includes rules to account for swing dynamics and leg-retraction. We apply it to a simulated Atlas robot in position-control, always commanded to perform in-place walking. The stance hip joint in our robot keeps the torso upright to let the robot naturally fall, and the swing hip joint tracks the desired footstep location. Combined with simple Center of Pressure (CoP) damping rules in the low-level controller, our foot-placement enables the robot to recover from strong pushes and produce periodic walking gaits when subject to persistent sources of disturbance, externally or internally. These gaits are imprecise, i.e., emergent from asymmetry sources rather than precisely imposing a desired velocity to the robot. Also in extreme conditions, restricting linearity assumptions of the 3LP model are often violated, but the system remains robust in our simulations. An extensive analysis of closed-loop eigenvalues, viable regions and sensitivity to push timings further demonstrate the strengths of our simple controller.

We investigated the capabilities of a reflex-based neuromuscular controller with a knee and hip gait trainer worn by a subject with a complete spinal cord injury. With controller assistance, this subject was able to reach a walking speed of 1.0m/s. Measured joint torques agreed reasonably well with those of healthy subjects. The controller was also robust, recovering from manual swing foot perturbations. These preliminary results are promising for future implementation of neuromuscular controllers on wearable prototypes for real-world walking conditions.

Propulsion of swimming robots at the surface and underwater is largely dominated by rotary propellers due to high thrust, but at the cost of low efficiency. Due to their inherently high speed turning motion, sharp propeller blades and generated noise, they also present a disturbance to maritime ecosystems. Our work presents a bio-inspired approach to efficient and eco-friendly swimming with moderate to high thrust. This paper describes the concept, development and experimental validation of the novel anguilliform robot MAR. With 15 elements making up the 0.5 m long propulsive section and driven by a single, speed-controlled brushless DC motor (BLDC), the robot creates a smooth continuous traveling wave for propulsion. Steering and autonomy are realized by an actuated head with integrated batteries that serves as a front-rudder. Almost neutral buoyancy paired with individually actuated pectoral fins furthermore enable submerged swimming and diving maneuvers. MAR accomplished high thrusts at a moderate power consumption in first performance tests. The achieved maximum velocity and the speed related efficiency (defined as the achieved speed over the power consumption m Ws(-1)) did not fulfill the expectations in the first tests (in comparison to commercial rotary thrusters), which can be largely attributed to the spatial limitations and an imperfect test setup. Nevertheless, the potential towards highly efficient and high thrust propulsion is visible and will be further investigated in future efforts.

Bio-inspired robotic designs introducing and benefiting from morphological aspects present in animals allowed the generation of fast, robust, and energy-efficient locomotion. We used engineering tools and interdisciplinary knowledge transferred from biology to build low-cost robots, able to achieve a certain level of versatility. Serval, a compliant quadruped robot with actuated spine and high range of motion in all joints, was developed to address the question of what mechatronic complexity is needed to achieve rich motion skills. In our experiments, the robot presented a high level of versatility (number of skills) at medium speed, with a minimal control effort and, in this article, no usage of its spine. Implementing a basic kinematics-duplication from dogs, we found strengths to emphasize, weaknesses to correct, and made Serval ready for future attempts to achieve more agile locomotion. In particular, we investigated the following skills: walk, trot, gallop, bound (crouched), sidestep, turn with a radius, ascend slopes including flat ground transition, perform single and double step-downs, fall, trot over bumpy terrain, lie/sit down, and stand up.

The paper addresses the problem of learning internal task-specific dynamic models for a reaching task. Using task-specific dynamic models is crucial for achieving both high tracking accuracy and compliant behaviour, which improves safety concerns while working in unstructured environment or with humans. The proposed approach uses programming by demonstration to learn new task-related movements encoded as Compliant Movement Primitives (CMPs). CMPs are a combination of position trajectories encoded in a form of Dynamic Movement Primitives (DMPs) and corresponding task-specific Torque Primitives (TPs) encoded as a linear combination of kernel functions. Unlike the DMPs, TPs cannot be directly acquired from user demonstrations. Inspired by the human sensorimotor learning ability we propose a novel method which autonomously learns task-specific TPs, based on a given kinematic trajectory in DMPs.

In this paper, we present a simple control framework for on-line push recovery with dynamic stepping properties. Due to relatively heavy legs in our robot, we need to take swing dynamics into account and thus use a linear model called 3LP which is composed of three pendulums to simulate swing and torso dynamics. Based on 3LP equations, we formulate discrete LQR controllers and use a particular time-projection method to adjust the next footstep location on-line during the motion continuously. This adjustment, which is found based on both pelvis and swing foot tracking errors, naturally takes the swing dynamics into account. Suggested adjustments are added to the Cartesian 3LP gaits and converted to joint-space trajectories through inverse kinematics. Fixed and adaptive foot lift strategies also ensure enough ground clearance in perturbed walking conditions. The proposed structure is robust, yet uses very simple state estimation and basic position tracking. We rely on the physical series elastic actuators to absorb impacts while introducing simple laws to compensate their tracking bias. Extensive experiments demonstrate the functionality of different control blocks and prove the effectiveness of time-projection in extreme push recovery scenarios. We also show self-produced and emergent walking gaits when the robot is subject to continuous dragging forces. These gaits feature dynamic walking robustness due to relatively soft springs in the ankles and avoiding any Zero Moment Point (ZMP) control in our proposed architecture., 20 pages journal paper