Real-time Motion Generation Research Articles

A Unified Real-Time Motion Generation Algorithm for Approximate Position Analysis of Planar N-Bar Mechanisms

Abstract This paper presents a novel real-time kinematic simulation algorithm for planar N-bar linkage mechanisms, both single- and multi-degrees-of-freedom, comprising revolute and/or prismatic joints and actuators. A key feature of this algorithm is a reinterpretation technique that transforms prismatic elements into a combination of revolute joint and links. This gives rise to a unified system of geometric constraints and a general-purpose solver which adapts to the complexity of the mechanism. The solver requires only two types of methods—fast dyadic decomposition and relatively slower optimization-based—to simulate all types of planar mechanisms. From an implementation point of view, this algorithm simplifies programming without requiring handling of different types of mechanisms. This versatile algorithm can handle serial, parallel, and hybrid planar mechanisms with varying degrees-of-freedom and joint types. Additionally, this paper presents an estimation of simulation time and structural complexity, shedding light on computational demands. Demonstrative examples showcase the practicality of this method.

Dynamic Optimization Fabrics for Motion Generation

Optimization fabrics are a geometric approach to real-time local motion generation, where motions are designed by the composition of several differential equations that exhibit a desired motion behavior. We generalize this framework to dynamic scenarios and nonholonomic robots and prove that fundamental properties can be conserved. We show that convergence to desired trajectories and avoidance of moving obstacles can be guaranteed using simple construction rules of the components. In addition, we present the first quantitative comparisons between optimization fabrics and model predictive control and show that optimization fabrics can generate similar trajectories with better scalability, and thus, much higher replanning frequency (up to 500 Hz with a 7 degrees of freedom robotic arm). Finally, we present empirical results on several robots, including a nonholonomic mobile manipulator with 10 degrees of freedom and avoidance of a moving human, supporting the theoretical findings.

Real-time controllable motion transition for characters

Real-time in-between motion generation is universally required in games and highly desirable in existing animation pipelines. Its core challenge lies in the need to satisfy three critical conditions simultaneously: quality, controllability and speed , which renders any methods that need offline computation (or post-processing) or cannot incorporate (often unpredictable) user control undesirable. To this end, we propose a new real-time transition method to address the aforementioned challenges. Our approach consists of two key components: motion manifold and conditional transitioning. The former learns the important low-level motion features and their dynamics; while the latter synthesizes transitions conditioned on a target frame and the desired transition duration. We first learn a motion manifold that explicitly models the intrinsic transition stochasticity in human motions via a multi-modal mapping mechanism. Then, during generation, we design a transition model which is essentially a sampling strategy to sample from the learned manifold, based on the target frame and the aimed transition duration. We validate our method on different datasets in tasks where no post-processing or offline computation is allowed. Through exhaustive evaluation and comparison, we show that our method is able to generate high-quality motions measured under multiple metrics. Our method is also robust under various target frames (with extreme cases).

Regularized Hierarchical Quadratic Program for Real-Time Whole-Body Motion Generation

The goal of this article is to find an optimal and robust solution for on-line hierarchical least-squares optimization subject to both equality and inequality constraints. We focus on the reasoning about the task regularization to ensure the convergence and robustness of a solution in the face a singularity. The mixed problem of a regularization and inequality-constrained hierarchical optimization is not fully discussed due to the mathematical complexity. We address this problem by formulating a regularized hierarchical quadratic programming. The solution is obtained in a unified and computationally efficient manner by leveraging a singular value decomposition and an active set method. At the same time, we concentrate on the realization of the proposed algorithm as a practical means of real-time whole-body motion generation. The effectiveness of the algorithm is validated through extensive numerical simulations and experimental tests of a rescue robot successfully executing manipulation missions in a highly unstructured environment.

RNN for Perturbed Manipulability Optimization of Manipulators Based on a Distributed Scheme: A Game-Theoretic Perspective.

In order to leverage the unique advantages of redundant manipulators, avoiding the singularity during motion planning and control should be considered as a fundamental issue to handle. In this article, a distributed scheme is proposed to improve the manipulability of redundant manipulators in a group. To this end, the manipulability index is incorporated into the cooperative control of multiple manipulators in a distributed network, which is used to guide manipulators to adjust to the optimal spatial position. Moreover, from the perspective of game theory, this article formulates the problem into a Nash equilibrium. Then, a neural network with anti-noise ability is constructed to seek and approximate the optimal strategy profile of the Nash equilibrium problem with time-varying parameters. Theoretical analyses show that the neural network model has the superior global convergence and noise immunity. Finally, simulation results demonstrate that the neural network is effective in real-time cooperative motion generation of multiple redundant manipulators under perturbations in distributed networks.

Versatile and robust bipedal walking in unknown environments: real-time collision avoidance and disturbance rejection

Autonomous navigation in complex environments featuring obstacles, varying ground compositions, and external disturbances requires real-time motion generation and stabilization simultaneously. In this paper, we present and evaluate a strategy for rejection of external disturbances and real-time motion generation in the presence of obstacles and non-flat ground. We propose different solutions for combining the associated algorithms and analyze them in simulations The promising method is validated in experiments with our robot Lola. We found a hierarchical approach to be effective for solving these complex motion generation problems, because it allows us to decompose the problem into sub-problems that can be tackled separately at different levels. This makes the approach suitable for real-time applications and robust against perturbations and errors. Our results show that real-time motion planning and disturbance rejection can be combined to improve the autonomy of legged robots.

Integrated Optimization of Climbing Locomotion for a Humanoid Robot

We investigate real-time motion generation for a humanoid robot climbing a wall. We apply nonlinear model predictive control (NMPC) to optimize the climbing path and motion of the limbs simultaneously. To climb a wall, the robot must determine which hold it will grasp in each step. Kinematic constraints vary depending on the holds grasped by the robot, and these constraints will change as the robot climbs. To determine which hold to grasp while considering kinematics constraints, our proposed method introduces a potential function on the wall and state-dependent weights. We also propose online configuration of the performance index for NMPC to reduce the computational cost. This enables optimization under a complicated wall model. Only information in the vicinity of the humanoid robot is utilized for optimization because the control input over a finite future horizon is optimized at each sampling time of NMPC. We simulate the behavior of a climbing robot to show that the computation time is nearly the same as a sampling period.

Safe and Efficient Human–Robot Collaboration Part II: Optimal Generalized Human-in-the-Loop Real-Time Motion Generation

The coexistence of humans and robots in fenceless robot cells requires robust safety precautions to prevent humans from being injured. Currently, safety is ensured by limiting the robot velocity, force, and power. This results in large cycle times and, hence, very inefficient industrial applications, where no amortization of the robotic system can be expected. In this letter, a novel method for improving the robot performance is presented that still complies with the international safety standards for collaborative robots. The approach of this letter is based on a projection of a human arm motion into the robot's path to estimate a possible collision with the robot. This idea is addressed in an optimization approach by minimizing the time needed by the robot to reach the goal position under human-in-the-loop constraints. The segmented path is optimized by solving a nonlinear programming problem, and the effect of crucial parameters is analyzed. To guarantee a flexible motion of the resulting optimized path, a generalization method using dynamic movement primitives and the compliance of constraints are proposed. Experiments validate this new method that significantly improves the efficiency of human–robot coexistence.

Kinematic optimization for bipedal robots: a framework for real-time collision avoidance

Bipedal locomotion is more than dynamically stable walking. The redundant kinematic design of humanoid robots allows for complex motions in complex scenarios. One challenge of current robotic research is the exploitation of the capacities of redundant robots in real-time applications. In this paper, we present and evaluate methods for real-time motion generation of redundant robots. The proposed methods are based on a model-predictive approach. We propose and compare methods for optimization of robot motions defined by parameterized task-space trajectories and for redundancy resolution. The approaches are successfully combined in a novel algorithm. The methods are introduced with the help of a minimal model. It shows their applicability for a wide range of complex robotic systems. We apply and validate their effectiveness and their real-time character in several experiments with different environments with the humanoid robot Lola.

Real-time whole-body motion generation using torso posture regression and center of mass

For household humanoid robots, reaching as much workspace as possible with their hands is an important issue because the locations of target objects may range from the floor to above the robot’s head. At the same time, to adapt to the constantly-changing household environment, inverse kinematics for the whole body must be solved in real time. In this paper, to achieve real-time motion generation for a humanoid robot, we propose a method of separating the inverse kinematics calculation into simpler problems. Using regression to estimate the torso orientation, we independently solve inverse kinematics for the lower body and both arms. First, using the target pose of both hands as input, we calculate the orientation of the torso and determine the target position of the center of mass considering the reachability of both arms. At each control step, we calculate the joint angles of the lower body from the position of the center of mass, feet poses, and torso orientation. Then, we calculate the joint angles of both arms. In experiments, we apply the proposed method to a human-size humanoid robot for reaching low-height positions while hunkering down. The proposed inverse kinematics solver is ten times faster than the numerical solution using the Jacobian matrix. We also verify the applicability of the proposed method using a sequence of random target positions for the hands as input.

Real-time user motion generation and correction using RGBD sensor

Real-time user motion generation and correction using RGBD sensor

Dynamical System Based Robotic Motion Generation With Obstacle Avoidance

The problem of real-time motion generation to a target with obstacle avoidance is considered. A second-order dynamical system having the target as a unique globally asymptotically stable equilibrium is modified in the presence of obstacles by an additive signal whose design is based on the prescribed performance control methodology. It is proved that obstacle avoidance and dynamical system stability are guaranteed and that the target remains asymptotically stable. Simulations are utilized to reveal the cases where the proposed scheme outperforms the modulated dynamical system and the constrained optimization priority framework. Experimental results further validate the theoretical findings.

Learning potential functions from human demonstrations with encapsulated dynamic and compliant behaviors

We consider the problem of devising a unified control policy capable of regulating both the robot motion and its physical interaction with the environment. We formulate this control policy by a non-parametric potential function and a dissipative field, which both can be learned from human demonstrations. We show that the robot motion and its stiffness behaviors can be encapsulated by the potential function's gradient and curvature, respectively. The dissipative field can also be used to model desired damping behavior throughout the motion, hence generating motions that follows the same velocity profile as the demonstrations. The proposed controller can be realized as a unification approach between "realtime motion generation" and "variable impedance control", with the advantages that it has guaranteed stability as well as does not rely on following a reference trajectory. Our approach, called unified motion and variable impedance control (UMIC), is completely time-invariant and can be learned from a few demonstrations via solving two (convex) constrained quadratic optimization problems. We validate UMIC on a library of 30 human handwriting motions and on a set of experiments on 7-DoF KUKA light weight robot.

Real-Time Motion Generation and Control Systems for High Wheel-Legged Robot Mobility

A wheel-legged mobile robot (WLMR) has both leg and wheel structures. WLMRs have adaptability advantages because they can change locomotion methods depending on the terrain. However, the location of a WLMR's center of gravity (CoG) is very high; thus, almost all existing WLMRs move statically. In this paper, whole body motion generation and various control systems are studied to facilitate higher WLMR mobility. To this end, a zero moment point (ZMP) is introduced as a stability index. In addition, WLMRs are modeled as single point mass linear inverted pendulums. Subsequently, online CoG pattern generation methods are proposed; one is a preview control approach, and a second is an approach that realizes the desired ZMP pattern using a zero-phase low-pass filter. It is then possible to generate the desired CoG patterns more easily and faster than with a preview control approach. The CoG patterns based on the single point model are constructed via the resolved-momentum-control approach. Finally, the effectiveness of the whole body motion pattern generated by the proposed methods is validated by simulations and experiments.

Biological inspiration used for robots motion synthesis

This work presents a biologically inspired method of gait generation. Bipedal gait pattern (for hip and knee joints) was taken into account giving the reference trajectories in a learning task. The four coupled oscillators were taught to generate the outputs similar to those in a human gait. After applying the correction functions the obtained generation method was validated using ZMP criterion. The formula suitable for real-time motion generation taking into account the positioning errors was also formulated. The small real robot prototype was tested to be able walk successfully following the elaborated motion pattern.

Real time falling animation with active and protective responses

Combined with motion capture and dynamic simulation, characters in animation have realistic motion details and can respond to unexpected contact forces. This paper proposes a novel and real-time character motion generation approach which introduces a parallel process, and uses an approximate nearest neighbor optimization search method. Besides, we employ a support vector machine (SVM), which is trained on a set of samples and predicts a subset of our ‘return-to’ motion capture (mocap) database in order to reduce the search time. In the dynamic simulation process, we focus on designing a biomechanics based controller which detects the balance of the characters in locomotion and drives them to take several active and protective responses when they fall to the ground in order to reduce the injuries to their bodies. Finally, we show the time costs in synthesis and the visual results of our approach. The experimental results indicate that our motion generation approach is suitable for interactive games or other real-time applications.

Robot gait synthesis using the scheme of human motions skills development

This work presents biologically inspired method of gait generation. Referring to the periodic signals generated by biological CPG, the coupled oscillators with correction functions are used to produce leg joint trajectories. The method of oscillators parameters evaluation using genetic algorithm is described. The correction functions are introduced. The gait formula suitable for real-time motion generation taking into account the positioning errors is given. Generated gait can be used as the basis for modifications to produce more complex motions. The problem of suitability of human gait for bipeds with body proportions different than that in human body is discussed.

放送現業・コンテンツ制作 モーションジェネレータによるバーチャルサーカス リアルタイムアニメーション生成システムの試み

Realistic virtual characters generated by computer graphics are increasingly required in broadcasting programs. We propose an algorithm that implements interactive, real-time motion generation for the creation of virtual acrobats that perform complex circus acrobatics using a motion generation engine to support production. To facilitate control, we designed a graphical user interface and a minimum set of inputs which give a very effective result in motion. For fast computation, we designed each character and other objects with minimum dynamics construction to perform each motion. Because a circus acrobat usually requires large space, the motion capture system cannot be used. Furthermore, traditional key-frame animation cannot apply the external forces that are important to the motion. To solve these problems, our algorithm uses dynamics simulation and is able to create a variety of motion.

Elastic Strips: A Framework for Motion Generation in Human Environments

Robotic applications are expanding into dynamic, unstructured, and populated environments. Mechanisms specifically designed to address the challenges arising in these environments, such as humanoid robots, exhibit high kinematic complexity. This creates the need for new algorithmic approaches to motion generation, capable of performing task execution and real-time obstacle avoidance in high-dimensional configuration spaces. The elastic strip framework presented in this paper enables the execution of a previously planned motion in a dynamic environment for robots with many degrees of freedom. To modify a motion in reaction to changes in the environment, real-time obstacle avoidance is combined with desired posture behavior. The modification of a motion can be performed in a task-consistent manner, leaving task execution unaffected by obstacle avoidance and posture behavior. The elastic strip framework also encompasses methods to suspend task behavior when its execution becomes inconsistent with other constraints imposed on the motion. Task execution is resumed automatically, once those constraints have been removed. Experiments demonstrating these capabilities on a nine-degree-of-freedom mobile manipulator and a 34-degree-of-freedom humanoid robot are presented, proving the elastic strip framework to be a powerful and versatile task-oriented approach to real-time motion generation and motion execution for robots with a large number of degrees of freedom in dynamic environments.

Elastic Strips: A Framework for Motion Generation in Human Environments

Robotic applications are expanding into dynamic, unstructured, and populated environments. Mechanisms specifically designed to address the challenges arising in these environments, such as humanoid robots, exhibit high kinematic complexity. This creates the need for new algorithmic approaches to motion generation, capable of performing task execution and real-time obstacle avoidance in high-dimensional configuration spaces. The elastic strip framework presented in this paper enables the execution of a previously planned motion in a dynamic environment for robots with many degrees of freedom. To modify a motion in reaction to changes in the environment, real-time obstacle avoidance is combined with desired posture behavior. The modification of a motion can be performed in a task-consistent manner, leaving task execution unaffected by obstacle avoidance and posture behavior. The elastic strip framework also encompasses methods to suspend task behavior when its execution becomes inconsistent with other constraints imposed on the motion. Task execution is resumed automatically, once those constraints have been removed. Experiments demonstrating these capabilities on a nine-degree-of-freedom mobile manipulator and a 34-degree-of-freedom humanoid robot are presented, proving the elastic strip framework to be a powerful and versatile task-oriented approach to real-time motion generation and motion execution for robots with a large number of degrees of freedom in dynamic environments.