Constrained Motion Planning Research Articles

Constrained Motion Planning Networks X

Constrained motion planning is a challenging field of research, aiming for computationally efficient methods that can find a collision-free path on the constraint manifolds between a given start and goal configuration. These planning problems come up surprisingly frequently, such as in robot manipulation for performing daily life assistive tasks. However, few solutions to constrained motion planning are available, and those that exist struggle with high computational time complexity in finding a path solution on the manifolds. To address this challenge, we present Constrained Motion Planning Networks X (CoMPNetX). It is a neural planning approach, comprising a conditional deep neural generator and discriminator with neural gradients-based fast projection operator. We also introduce neural task and scene representations conditioned on which the CoMPNetX generates implicit manifold configurations to turbo-charge any underlying classical planner such as sampling-based motion planning methods for quickly solving complex constrained planning tasks. We show that our method finds path solutions with high success rates and lower computation times than state-of-the-art traditional path-finding tools on various challenging scenarios.

Constrained Motion Planning of 7-DOF Space Manipulator via Deep Reinforcement Learning Combined with Artificial Potential Field

During the on-orbit operation task of the space manipulator, some specific scenarios require strict constraints on both the position and orientation of the end-effector, such as refueling and auxiliary docking. To this end, a novel motion planning approach for a space manipulator is proposed in this paper. Firstly, a kinematic model of the 7-DOF free-floating space manipulator is established by introducing the generalized Jacobian matrix. On this basis, a planning approach is proposed to realize the motion planning of the 7-DOF free-floating space manipulator. Considering that the on-orbit environment is dynamical, the robustness of the motion planning approach is required, thus the deep reinforcement learning algorithm is introduced to design the motion planning approach. Meanwhile, the deep reinforcement learning algorithm is combined with artificial potential field to improve the convergence. Besides, the self-collision avoidance constraint is considered during planning to ensure the operational security. Finally, comparative simulations are conducted to demonstrate the performance of the proposed planning method.

A Holistic Approach to Human-Supervised Humanoid Robot Operations in Extreme Environments.

Nuclear energy will play a critical role in meeting clean energy targets worldwide. However, nuclear environments are dangerous for humans to operate in due to the presence of highly radioactive materials. Robots can help address this issue by allowing remote access to nuclear and other highly hazardous facilities under human supervision to perform inspection and maintenance tasks during normal operations, help with clean-up missions, and aid in decommissioning. This paper presents our research to help realize humanoid robots in supervisory roles in nuclear environments. Our research focuses on National Aeronautics and Space Administration (NASA’s) humanoid robot, Valkyrie, in the areas of constrained manipulation and motion planning, increasing stability using support contact, dynamic non-prehensile manipulation, locomotion on deformable terrains, and human-in-the-loop control interfaces.

Constrained motion planning of free-float dual-arm space manipulator via deep reinforcement learning

Space manipulator has complex kinetic and dynamic properties due to its free-float base. Moreover, the motion planning of a dual-arm manipulator is even more challenging due to the extra motion constraints between arms. To this end, a deep reinforcement learning based motion planning method is proposed to solve the complex constrained motion planning problem of free-floating dual-arm space manipulator. Firstly, the Generalized Jacobian Matrix is used to establish the kinematic model of the manipulator system. Based on this model, the Deep Deterministic Policy Gradients algorithm is adopted to realize the complex constrained motion planning. Meanwhile, the self-collision avoidance and velocity constraint of the end-effector are taken into consideration in the planning process to improve the robustness of the algorithm. Finally, trajectories of individual joint are obtained under both motionless and spinning scenarios, and simulation tests are conducted to demonstrate the performance of the proposed planning methods.

Towards Tethered Tool Manipulation Planning with the Help of a Tool Balancer

Handling and maneuvering tools across a robot workspace is a challenging task that often requires the implementation of constrained motion planning. In the case of wired or tethered tools, their maneuvering becomes considerably harder by the tool cable. If the cable presence is not considered, the robot motions may make the cable become entangled with the robot arms or elements of its workspace, causing accidents or unnecessary strain on the robot and the tool. Furthermore, the behavior of the tool cable during manipulation and its degree of entanglement around the robot are difficult to predict. The present paper introduces a constrained manipulation planner for dual-armed tethered tool manipulation involving tool re-grasping. Our solution employs a tool balancer to straighten the tool cable and facilitate the cable deformation problem. The planner predicts the cable states during manipulation and restricts the robot motions in order to avoid cable entanglements and collisions while performing tool re-posing tasks. Furthermore, the planner also applies orientational constraints to limit the cable bending, reducing the torque and stress suffered by the robot due to the cable tension. Simulations and real-world experiments validated the presented method.

Tangent bundle RRT: A randomized algorithm for constrained motion planning

SUMMARYThe Tangent Bundle Rapidly Exploring Random Tree (TB-RRT) is an algorithm for planning robot motions on curved configuration space manifolds, in which the key idea is to construct random trees not on the manifold itself, but on tangent bundle approximations to the manifold. Curvature-based methods are developed for constructing tangent bundle approximations, and procedures for random node generation and bidirectional tree extension are developed that significantly reduce the number of projections to the manifold. Extensive numerical experiments for a wide range of planning problems demonstrate the computational advantages of the TB-RRT algorithm over existing constrained path planning algorithms.

Dynamic walking and whole-body motion planning for humanoid robots: an integrated approach

This paper presents a general method for planning collision-free whole-body walking motions for humanoid robots. First, we present a randomized algorithm for constrained motion planning, that is used to generate collision-free statically balanced paths solving manipulation tasks. Then, we show that dynamic walking makes humanoid robots small-space controllable. Such a property allows to easily transform collision-free statically balanced paths into collision-free dynamically balanced trajectories. It leads to a sound algorithm which has been applied and evaluated on several problems where whole-body planning and walk are needed, and the results have been validated on a real HRP-2 robot.

Splines and polynomial tools for flatness-based constrained motion planning

This article addresses the problem of trajectory planning for flat systems with constraints. Flat systems have the useful property that the input and the state can be completely characterised by the so-called flat output. We propose a spline parametrisation for the flat output, the performance output, the states and the inputs. Using this parametrisation the problem of constrained trajectory planning can be cast into a simple quadratic programming problem. An important result is that the B-spline parametrisation used gives exact results for constrained linear continuous-time system. The result is exact in the sense that the constrained signal can be made arbitrarily close to the boundary without having intersampling issues (as one would have in sampled-data systems). Simulation examples are presented, involving the generation of rest-to-rest trajectories. In addition, an experimental result of the method is also presented, where two methods to generate trajectories for a magnetic-levitation (maglev) system in the presence of constraints are compared and each method's performance is discussed. The first method uses the nonlinear model of the plant, which turns out to belong to the class of flat systems. The second method uses a linearised version of the plant model around an operating point. In every case, a continuous-time description is used. The experimental results on a real maglev system reported here show that, in most scenarios, the nonlinear and linearised models produce almost similar, indistinguishable trajectories.

Heuristic solution for constrained 7-DOF motion planning in 3D scanning application

This paper proposes a heuristic algorithm for near-optimal handling of the redundancy in robot mechanisms, applied to a 3D scanning platform consisting of a 6-DOF articulated robot arm which moves a laser sensor around a workpiece placed on a rotary table. The proposed method handles both static constraints, like avoiding joint limits, kinematic singularities or collisions, modelled using a configuration map viewed as a grayscale image, and dynamic constraints like velocity and acceleration. The algorithm is compared with Dijkstra algorithm, which gives the optimal solution, but is very slow, and with two other heuristics which are very fast, but give suboptimal solutions.

Constrained motion planning of nonholonomic systems

This paper addresses the constrained motion planning problem for nonholonomic systems represented by driftless control systems with output. The problem consists in defining a control function driving the system output to a desirable point at a given time instant, whereas state and control variables remain over the control horizon within prescribed bounds. The state and control constraints are handled by extending the control system with a pair of state equations driven by the violation of constraints, and adding regularizing perturbations. For the regularized system a Jacobian motion planning algorithm is designed, called imbalanced. Solutions of example constrained motion planning problems for the rolling ball illustrate the theoretical concepts.

Modeling and control of the air system of a turbocharged gasoline engine

In order to reduce CO 2 emissions in the automotive industry, there is a trend to decrease the size of the engines and add turbocharging technologies to increase the engine efficiency while keeping performances. In this context, numerous studies have focused on the improvement of turbocharger control strategies. The strategy proposed in this paper is based on constrained motion planning and feedback linearization applied to a simplified model of the system. It takes advantage of the whole system bandwidth and requires a limited calibration effort. Thanks to its structure, it offers an interesting potential for generalization to more complex turbocharging systems. Extensive experimental tests are reported for a four-cylinder gasoline engine.

A Heuristic Approach for Constrained Real Time Motion Planning of a Redundant 7-DOF Mechanism for 3D Laser Scanning

This article presents a heuristic algorithm for motion planning of a short range 7-DOF Laser Scanning System consisting of a 6-DOF vertical robot arm and a rotary table holding the workpiece. The redundancy of the mechanism is exploited for specifying certain constraints such as imposing a smooth motion of the rotary table while respecting the acceleration and speed limits and avoiding undesirable configurations of the robot arm such as the kinematic singularities, out-of-range conditions and collisions between the mechanical elements of the system (robot, table, workpiece and laser probe). The paper also presents the control system of the rotary table, and a communication protocol which allows the integration of the above-mentioned planning algorithm into the 3D scanning system.

Real-time motion planning for multibody systems

The solution of constrained motion planning is an important task in a wide number of application fields. The real-time solution of such a problem, formulated in the framework of optimal control theory, is a challenging issue. We prove that a real-time solution of the constrained motion planning problem for multibody systems is possible for practical real-life applications on standard personal computers.

Predicting planar motion behavior under contact uncertainty

Performing complex assembly tasks with robots requires fine-motion planners able to cope with uncertainty and contact motions, and this is a recognized difficult issue. This paper proposes a method to predict the behavior of motions under contact uncertainty in order to check the feasibility of paths generated by gross-motion planning algorithms from a nominal model of the environment. This pragmatical approach enables the extension of gross-motion planning techniques to constrained-motion planning problems, ensuring the feasibility of the task despite the uncertainties. The approach has been implemented for assembly tasks in the plane with three degrees of freedom.

A method of progressive constraints for manipulation planning

We present a method for manipulation planning derived from a systematic approach to constrained motion planning. We call it a method of progressive constraints. It replaces an original problem by a series of progressively constrained ones, and makes a solution of the current problem converge toward a solution of the original problem. The manipulation constraints are analyzed and it is shown how they can be introduced progressively. The progressive constraints framework can thus be used for solving manipulation planning problems. The general algorithm proposed for constrained motion planning can be derived from a standard geometric path planner. We use the variational dynamic programming path planner, in which at every iteration a current path is used to generate a submanifold that is searched for a better path.

Algorithmic motion planning in robotics

A survey is presented of an approach to motion planning that emphasizes object-oriented, exact, and discrete (or combinatorial) algorithmic techniques in which worst-case asymptotically efficient solutions are being sought. Following a statement of the problem, motion planning in static and known environments is treated. The discussion covers general solutions, lower bounds, the projection method, the retraction method, the expanded obstacles, the single-component approach, and a mobile convex object moving in a 2D polygonal space. Variants of the motion-planning problem are then considered, namely, optimal motion planning, adaptive and exploratory motion planning, motion planning in the presence of moving obstacles, constrained motion planning, motion planning with uncertainty, and general task planning. >