Motion Primitives Research Articles

Muscle synergies inherent in simulated hypogravity running reveal flexible but not unconstrained locomotor control

With human space exploration back in the spotlight, recent studies have investigated the neuromuscular adjustments to simulated hypogravity running. They have examined the activity of individual muscles, whereas the central nervous system may rather activate groups of functionally related muscles, known as muscle synergies. To understand how locomotor control adjusts to simulated hypogravity, we examined the temporal (motor primitives) and spatial (motor modules) components of muscle synergies in participants running sequentially at 100%, 60%, and 100% body weight on a treadmill. Our results highlighted the paradoxical nature of simulated hypogravity running: The reduced mechanical constraints allowed for a more flexible locomotor control, which correlated with the degree of spatiotemporal adjustments. Yet, the increased temporal (shortened stance phase) and sensory (deteriorated proprioceptive feedback) constraints required wider motor primitives and a higher contribution of the hamstring muscles during the stance phase. These results are a first step towards improving astronaut training protocols.

Dynamic Motion Primitives-Based Trajectory Learning for Physical Human–Robot Interaction Force Control

One promising function of interactive robots is to provide a specific interaction force to human users. For example, rehabilitation robots are expected to promote patients' recovery by interacting with them with a prescribed force. However, motion uncertainties of different individuals, which are hard to predict due to the varying motion speed and noises during motion, degrade the performance of existing control methods. This paper proposes a method to learn a desired reference trajectory for a robot based on dynamic motion primitives (DMPs) and iterative learning (IL). By controlling the robot to follow the generated desired reference trajectory, the interaction force can achieve a desired value. In our proposed approach, DMPs are first employed to parameterize the demonstration trajectories of the human user. Then a recursive least square (RLS)-based estimator is developed and combined with the Adam optimization method to update the trajectory parameters so that the desired reference trajectory of the robot is iteratively obtained by resolving the DMPs. Since the proposed method parameterizes the trajectories depending on the phrase variable, it removes the essential assumption of traditional IL methods where the iteration period should be invariant, and thus has improved robustness compared with the existing methods. Experiments are performed using an interactive robot to validate the effectiveness of our proposed scheme.

Learning Skills From Demonstrations: A Trend From Motion Primitives to Experience Abstraction

Learning Skills From Demonstrations: A Trend From Motion Primitives to Experience Abstraction

Controlling Musculoskeletal System with Spinal Cord-Inspired Constraint Force Field Primitives

Organisms naturally have extraordinary motor ability. They can autonomously control the highly redundant and nonlinear musculoskeletal system to complete fast and flexible movements. Motor primitive theory proposed that a complex movement can be realized by modular organization of simple movement patterns. This theory provides a convincing theoretical basis for explaining the extraordinary motor ability of organisms, and has been supported by plenty of experimental evidences in neurosciences. As a kind of typical motor primitive related to the modular regulation of the spinal motor system, convergent force field was thought to be significant for reducing the complexity of controlling numerous muscles and joints simultaneously. Inspired by the primitive property of convergent force field, in this paper, we proposed a new algorithm to efficiently construct constraint force field on musculoskeletal system with highly redundant actuators by taking optimized parameters of constraint force field as motor primitives. Compared with previous methods of constructing constraint force field, the proposed algorithm is able to reduce the dimension of the solution space so as to effectively improve the computational efficiency of constructing constraint force field in musculoskeletal system with redundant muscles. Validation experiments were carried out in a musculoskeletal system with 10 redundant muscles. The number of optimization iterations and computational time required for constructing constraint force field at a new position were significantly reduced. The system can accurately reach the target position using constant activations. This work may bring in new inspiration for realizing efficient and accurate motion control of musculoskeletal robots with limited feedback accuracy.

Dynamic Via-points and Improved Spatial Generalization for Online Trajectory Generation with Dynamic Movement Primitives

Dynamic Movement Primitives (DMP) have found remarkable applicability and success in various robotic tasks, which can be mainly attributed to their generalization, modulation and robustness properties. However, the spatial generalization of DMP can be problematic in some cases, leading to excessive overscaling and in turn large velocities and accelerations. While other DMP variants have been proposed in the literature to tackle this issue, they can also exhibit excessive overscaling as we show in this work. Moreover, incorporating intermediate points (via-points) for adjusting the DMP trajectory to account for the geometry of objects related to the task, or to avoid or push aside objects that obstruct a specific task, is not addressed by the current DMP literature. In this work we tackle these unresolved so far issues by proposing an improved online spatial generalization, that remedies the shortcomings of the classical DMP generalization, and moreover allows the incorporation of dynamic via-points. This is achieved by designing an online adaptation scheme for the DMP weights which is proved to minimize the distance from the demonstrated acceleration profile to retain the shape of the demonstration, subject to dynamic via-point and initial/final state constraints. Extensive comparative simulations with the classical and other DMP variants are conducted, while experimental results validate the practical usefulness and efficiency of the proposed method.

A KMP-based interactive learning approach for robot trajectory adaptation with obstacle avoidance

Purpose Imitation learning is a powerful tool for planning the trajectory of robotic end-effectors in Cartesian space. Present methods can adapt the trajectory to the obstacle; however, the solutions may not always satisfy users, whereas it is hard for a nonexpert user to teach the robot to avoid obstacles in time as he/she wishes through demonstrations. This paper aims to address the above problem by proposing an approach that combines human supervision with the kernelized movement primitives (KMP) model. Design/methodology/approach This approach first extracts the reference database used to train KMP from demonstrations by using Gaussian mixture model and Gaussian mixture regression. Subsequently, KMP is used to modulate the trajectory of robotic end-effectors in real time based on feedback from its interaction with humans to avoid obstacles, which benefits from a novel reference database update strategy. The user can test different obstacle avoidance trajectories in the current task until a satisfactory solution is found. Findings Experiments performed with the KUKA cobot for obstacle avoidance show that this approach can adapt the trajectories of the robotic end-effector to the user’s wishes in real time, including trajectories that the robot has already passed and has not yet passed. Simulation comparisons also show that it exhibits better performance than KMP with the original reference database update strategy. Originality/value An interactive learning approach based on KMP is proposed and verified, which not only enables users to plan the trajectory of robotic end-effectors for obstacle avoidance more conveniently and efficiently but also provides an effective idea for accomplishing interactive learning tasks under constraints.

Robot skill learning system of multi-space fusion based on dynamic movement primitives and adaptive neural network control

This article develops a robot skill learning system with multi-space fusion, simultaneously considering motion/stiffness generation and trajectory tracking. To begin with, surface electromyography (sEMG) signals from the human arm is captured based on the MYO armband to estimate endpoint stiffness. Gaussian Process Regression (GPR) is combined with dynamic movement primitive (DMP) to extract more skills features from multi-demonstrations. Then, the traditional DMP formulation is improved based on the Riemannian metric to encode the robot’s quaternions with non-Euclidean properties. Furthermore, an adaptive neural network (NN)-based finite-time admittance controller is designed to track the trajectory generated by the motion model and to reflect the learned stiffness characteristics. In this controller, a radial basis function neural network (RBFNN) is employed to compensate for the uncertainty of the robot dynamics. Finally, experimental validation is conducted using the ROKAE collaborative robot, confirming the effectiveness of the proposed approach. In summary, the presented framework is suitable for human-robot skill transfer method that require simultaneous consideration of position and stiffness in Euclidean space, as well as orientation on Riemannian manifolds.

Continuous adaptive gaits manipulation for three-fingered robotic hands via bioinspired fingertip contact events

The remarkable skill of changing its grasp status and relocating its fingers to perform continuous in-hand manipulation is essential for a multifingered anthropomorphic hand. A commonly utilized method of manipulation involves a series of basic movements executed by a high-level controller. However, it remains unclear how these primitives evolve into sophisticated finger gaits during manipulation. Here, we propose an adaptive finger gait-based manipulation method that offers real-time regulation by dynamically changing the primitive interval to ensure the force/moment balance of the object. Successful manipulation relies on contact events that act as triggers for real-time online replanning of multifinger manipulation. We identify four basic motion primitives of finger gaits and create a heuristic finger gait that enables the continuous object rotation of a round cup. Our experimental results verify the effectiveness of the proposed method. Despite the constant breaking and reengaging of contact between the fingers and the object during manipulation, the robotic hand can reliably manipulate the object without failure. Even when the object is subjected to interfering forces, the proposed method demonstrates robustness in managing interference. This work has great potential for application to the dexterous operation of anthropomorphic multifingered hands.

MPTree: A Sampling-based Vehicle Motion Planner for Real-time Obstacle Avoidance

This paper presents a novel and modular framework, named MPTree, for real-time vehicle motion planning with dynamic obstacle avoidance. MPTree adopts a sampling-based algorithm, to explore a tree of Motion Primitives (MPs) and return near-optimal trajectories. Specifically, MPTree builds motion primitives to connect the tree nodes (waypoints), sampled in a mesh of waylines on the local planning horizon. The tree exploration is based on a semi-structured RRTa, with an application-specific cost function (e.g., minimum-jerk or minimum-time) and high-level behavioral policy. We show examples of MPTREE’s specialization for urban environments and autonomous racing, using fast-to-evaluate motion primitives to accelerate the tree exploration phase. A prototype implementation is tested in a closed-loop simulation environment. Our preliminary results show that MPTree provides feasible collision-free trajectories while ensuring low computational times.

Learning Motion Primitives for the Quantification and Diagnosis of Mobility Deficits.

The severity of mobility deficits is one of the most critical parameters in the diagnosis and rehabilitation of Parkinson's disease (PD). The current approach for severity evaluation is clinical scaling that relies on a clinician's subjective observations and experience, and the observation in laboratories or clinics may not suffice to reflect the severity of motion deficits as compared to daily living activities. The paper presents an approach to modeling and quantifying the severity of mobility deficits from motion data by using nonintrusive wearable physio-biological sensors. The approach provides a user-specific metric that measures mobility deficits in terms of the quantities of motion primitives that are learned from motion tracking data. The proposed method achieved 99.84% prediction accuracy on laboratory data and 93.95% prediction accuracy on clinical data. This approach presents the potential to supplant traditional observation-based clinical scaling, providing an avenue for real-time feedback to fortify positive progression throughout the course of rehabilitation.

Biomimetic learning of hand gestures in a humanoid robot.

Hand gestures are a natural and intuitive form of communication, and integrating this communication method into robotic systems presents significant potential to improve human-robot collaboration. Recent advances in motor neuroscience have focused on replicating human hand movements from synergies also known as movement primitives. Synergies, fundamental building blocks of movement, serve as a potential strategy adapted by the central nervous system to generate and control movements. Identifying how synergies contribute to movement can help in dexterous control of robotics, exoskeletons, prosthetics and extend its applications to rehabilitation. In this paper, 33 static hand gestures were recorded through a single RGB camera and identified in real-time through the MediaPipe framework as participants made various postures with their dominant hand. Assuming an open palm as initial posture, uniform joint angular velocities were obtained from all these gestures. By applying a dimensionality reduction method, kinematic synergies were obtained from these joint angular velocities. Kinematic synergies that explain 98% of variance of movements were utilized to reconstruct new hand gestures using convex optimization. Reconstructed hand gestures and selected kinematic synergies were translated onto a humanoid robot, Mitra, in real-time, as the participants demonstrated various hand gestures. The results showed that by using only few kinematic synergies it is possible to generate various hand gestures, with 95.7% accuracy. Furthermore, utilizing low-dimensional synergies in control of high dimensional end effectors holds promise to enable near-natural human-robot collaboration.

Contextualized Trajectory Parsing with Spatio-Temporal Graph

This work investigates how to automatically parse object trajectories in surveillance videos, which aims at jointly solving three subproblems: 1) spatial segmentation, 2) temporal tracking, and 3) object categorization. We present a novel representation spatiotemporal graph (ST-Graph) in which: 1) Graph nodes express the motion primitives, each representing a short sequence of small-size patches over consecutive images, and 2) every two neighbor nodes are linked with either a positive edge or a negative edge to describe their collaborative or exclusive relationship of belonging to the same object trajectory. Phrasing the trajectory parsing as a graph multicoloring problem, we propose a unified probabilistic formulation to integrate various types of context knowledge as informative priors. An efficient composite cluster sampling algorithm is employed in search of the optimal solution by exploiting both the collaborative and the exclusive relationships between nodes. The proposed framework is evaluated over challenging videos from public datasets, and results show that it can achieve state-of-the-art tracking accuracy.

PRIMP: PRobabilistically-Informed Motion Primitives for Efficient Affordance Learning From Demonstration

PRIMP: PRobabilistically-Informed Motion Primitives for Efficient Affordance Learning From Demonstration

Computationally Efficient Minimum-Time Motion Primitives for Vehicle Trajectory Planning

Computationally Efficient Minimum-Time Motion Primitives for Vehicle Trajectory Planning

Kinodynamic Motion Planning for a System with Squid Dynamics

This paper introduces a path planning algorithm for a system with squid dynamics in a cluttered environment. We capture the complex interactions of fin, arms, and body patterning by analyzing experimental data collected from observing squid motion. We extract nine motion primitives to build the control sequence for a time-optimal trajectory. This task is formulated as a mixed-integer program, and we generate the minimum-time trajectory using a sample-based approach. Numerical simulations illustrate the efficacy of this strategy and motivate ongoing and future efforts to exploration of squid motion features, improvement of the modeling, and experimental demonstrations of the motion planning strategy.

A Comparison of Approaches for Segmenting the Reaching and Targeting Motion Primitives in Functional Upper Extremity Reaching Tasks.

There is growing interest in the kinematic analysis of human functional upper extremity movement (FUEM) for applications such as health monitoring and rehabilitation. Deconstructing functional movements into activities, actions, and primitives is a necessary procedure for many of these kinematic analyses. Advances in machine learning have led to progress in human activity and action recognition. However, their utility for analyzing the FUEM primitives of reaching and targeting during reach-to-grasp and reach-to-point tasks remains limited. Domain experts use a variety of methods for segmenting the reaching and targeting motion primitives, such as kinematic thresholds, with no consensus on what methods are best to use. Additionally, current studies are small enough that segmentation results can be manually inspected for correctness. As interest in FUEM kinematic analysis expands, such as in the clinic, the amount of data needing segmentation will likely exceed the capacity of existing segmentation workflows used in research laboratories, requiring new methods and workflows for making segmentation less cumbersome. This paper investigates five reaching and targeting motion primitive segmentation methods in two different domains (haptics simulation and real world) and how to evaluate these methods. This work finds that most of the segmentation methods evaluated perform reasonably well given current limitations in our ability to evaluate segmentation results. Furthermore, we propose a method to automatically identify potentially incorrect segmentation results for further review by the human evaluator. Clinical impact: This work supports efforts to automate aspects of processing upper extremity kinematic data used to evaluate reaching and grasping, which will be necessary for more widespread usage in clinical settings.

Trajectory Deformation-Based Multi-Modal Adaptive Compliance Control for a Wearable Lower Limb Rehabilitation Robot.

Adaptive compliance control is critical for rehabilitation robots to cope with the varying rehabilitation needs and enhance training safety. This article presents a trajectory deformation-based multi-modal adaptive compliance control strategy (TD-MACCS) for a wearable lower limb rehabilitation robot (WLLRR), which includes a high-level trajectory planner and a low-level position controller. Dynamic motion primitives (DMPs) and a trajectory deformation algorithm (TDA) are integrated into the high-level trajectory planner, generating multi-joint synchronized desired trajectories through physical human-robot interaction (pHRI). In particular, the amplitude modulation factor of DMPs and the deformation factor of TDA are adapted by a multi-modal adaptive regulator, achieving smooth switching of human-dominant mode, robot-dominant mode, and soft-stop mode. Besides, a linear active disturbance rejection controller is designed as the low-level position controller. Four healthy participants and two stroke survivors are recruited to conduct robot-assisted walking experiments using the TD-MACCS. The results show that the TD-MACCS can smoothly switch three control modes while guaranteeing trajectory tracking accuracy. Moreover, we find that appropriately increasing the upper bound of the deformation factor can enhance the average walking speed (AWS) and root mean square of trajectory deviation (RMSTD).

Combination of Dynamic Movement Primitives and Recurrent Neural Network to Solve Kinematics and Motion Planning Problems of an Exoskeleton‐Based Rehabilitation Robot

In the field of robotics, especially in exoskeleton‐based rehabilitation robots, motion planning is an important problem. An effective motion trajectory must ensure naturalness and similarity to daily human movements. In this study, the researchers proposed using the dynamic movement primitives (DMPs) method to create trajectories similar to the actual trajectories of daily human activities. Subsequently, the study proposed and constructed a recurrent neural network (RNN) model to solve the inverse kinematics (IK) problem. This model was trained using a large dataset generated from the DMP algorithm mentioned earlier. The IK results applied to simulations on the upper limb exoskeleton robot of Vietnam (UExosVN) under development showed that the obtained trajectories have high accuracy and stability, suitable for control and similar in form to the desired trajectories. The study conducted experiments with two activities of daily living (ADL), namely the playing ball task and drinking water task. The errors in joint variables for the IK problem are all under 1°, except for joint five, which does not significantly impact the robot’s end‐point accuracy. The shape similarity level in the predicted trajectory compared to the test data is mostly above 0.75. The combined approach of DMP and RNN provides a potential solution for designing motion trajectories for rehabilitation robots.

Cooperative Control for Dual-Arm Robots Based on Improved Dynamic Movement Primitives

Cooperative Control for Dual-Arm Robots Based on Improved Dynamic Movement Primitives

Probabilized Dynamic Movement Primitives and Model Predictive Planning for Enhanced Trajectory Imitation Learning

Probabilized Dynamic Movement Primitives and Model Predictive Planning for Enhanced Trajectory Imitation Learning