scholarly journals Fast deep swept volume estimator

2020 ◽  
pp. 027836492094078
Author(s):  
Hao-Tien Lewis Chiang ◽  
John EG Baxter ◽  
Satomi Sugaya ◽  
Mohammad R Yousefi ◽  
Aleksandra Faust ◽  
...  
Keyword(s):  
Deep Neural Networks ◽  
Volume Estimation ◽  
Robot Kinematics ◽  
Swept Volumes ◽  
Geometry Estimation ◽  
On Line ◽  
Volume Algorithm ◽  
Swept Volume ◽  
Joint Motions ◽  
And Task

Despite decades of research on efficient swept volume computation for robotics, computing the exact swept volume is intractable and approximate swept volume algorithms have been computationally prohibitive for applications such as motion and task planning. In this work, we employ deep neural networks (DNNs) for fast swept volume estimation. Since swept volume is a property of robot kinematics, a DNN can be trained off-line once in a supervised manner and deployed in any environment. The trained DNN is fast during on-line swept volume geometry or size inferences. Results show that DNNs can accurately and rapidly estimate swept volumes caused by rotational, translational, and prismatic joint motions. Sampling-based planners using the learned distance are up to five times more efficient and identify paths with smaller swept volumes on simulated and physical robots. Results also show that swept volume geometry estimation with a DNN is over 98.9% accurate and 1,200 times faster than an octree-based swept volume algorithm.

scholarly journals Task-space regulation of rigid-link electrically-driven robots with uncertain kinematics using neural networks

2021 ◽  
Vol 54 (1-2) ◽  
pp. 102-115
Author(s):  
Wenhui Si ◽  
Lingyan Zhao ◽  
Jianping Wei ◽  
Zhiguang Guan
Keyword(s):  
Neural Networks ◽  
Asymptotic Stability ◽  
Control Method ◽  
Joint Space ◽  
Robot Kinematics ◽  
Rbf Neural Networks ◽  
Task Space ◽  
Actuator Dynamics ◽  
Rigid Link ◽  
On Line

Extensive research efforts have been made to address the motion control of rigid-link electrically-driven (RLED) robots in literature. However, most existing results were designed in joint space and need to be converted to task space as more and more control tasks are defined in their operational space. In this work, the direct task-space regulation of RLED robots with uncertain kinematics is studied by using neural networks (NN) technique. Radial basis function (RBF) neural networks are used to estimate complicated and calibration heavy robot kinematics and dynamics. The NN weights are updated on-line through two adaptation laws without the necessity of off-line training. Compared with most existing NN-based robot control results, the novelty of the proposed method lies in that asymptotic stability of the overall system can be achieved instead of just uniformly ultimately bounded (UUB) stability. Moreover, the proposed control method can tolerate not only the actuator dynamics uncertainty but also the uncertainty in robot kinematics by adopting an adaptive Jacobian matrix. The asymptotic stability of the overall system is proven rigorously through Lyapunov analysis. Numerical studies have been carried out to verify efficiency of the proposed method.

Generating Swept Volumes With Instantaneous Screw Axes

1994 ◽  
Author(s):  
Zeng-Jia Hu ◽  
Zhi-Kui Ling
Keyword(s):  
Moving Object ◽  
Ruled Surfaces ◽  
Screw Axis ◽  
Spherical Surfaces ◽  
Swept Volumes ◽  
Instantaneous Screw Axis ◽  
Swept Volume ◽  
Instantaneous Screw ◽  
Plane Polygon ◽  
Boundary Surfaces

Abstract The instantaneous screw axis is used in the generation of the swept volume of a moving object. The envelope theory is used to determine the boundary surfaces of the swept volume. Specifically, the envelope surfaces generated by a plane polygon, cylindrical and spherical surfaces are presented. Furthermore, the ruled surfaces generated by edges of the moving object are discussed.

Geometric Representation of Translational Swept Volumes and its Applications

1986 ◽  
Vol 108 (2) ◽  
pp. 113-119 ◽  
Author(s):  
M. C. Leu ◽  
S. H. Park ◽  
K. K. Wang
Keyword(s):  
Three Dimensional ◽  
Boundary Representation ◽  
Geometric Representation ◽  
Ray Casting ◽  
Three Dimensional Objects ◽  
Swept Volumes ◽  
Representation Method ◽  
Primitive Object ◽  
Swept Volume ◽  
Cylinders And Spheres

This paper presents a method for representing the geometries of translational swept volumes of three-dimensional objects which can be constructed by the union of three types of primitive objects: blocks, cylinders, and spheres. The representation method involves three major steps. First, the swept volume of each primitive object is modeled by a boundary representation. Second, based on ray-casting and scan-rendering methods, the boundary representation is converted into a ray in–out classification, which represents the rays entering and exiting from the primitive swept volume. Third, the ray in–out classifications for various primitive swept volumes are combined to represent the swept volume of an object constructed from the primitive objects. Examples are given to illustrate how swept-volume representations can be useful in the context of off-line NC and robot program verifications.

scholarly journals An Adaptive Message Passing MPSoC Framework

2009 ◽  
Vol 2009 ◽  
pp. 1-20 ◽  
Author(s):  
Gabriel Marchesan Almeida ◽  
Gilles Sassatelli ◽  
Pascal Benoit ◽  
Nicolas Saint-Jean ◽  
Sameer Varyani ◽  
...  
Keyword(s):  
Message Passing ◽  
Continuous Mapping ◽  
Superior Performance ◽  
Code Migration ◽  
On Line ◽  
Run Time ◽  
Mapping Techniques ◽  
Application Requirements ◽  
Near Future ◽  
And Task

Multiprocessor Systems-on-Chips (MPSoCs) offer superior performance while maintaining flexibility and reusability thanks to software oriented personalization. While most MPSoCs are today heterogeneous for better meeting the targeted application requirements, homogeneous MPSoCs may become in a near future a viable alternative bringing other benefits such as run-time load balancing and task migration. The work presented in this paper relies on a homogeneous NoC-based MPSoC framework we developed for exploring scalable and adaptive on-line continuous mapping techniques. Each processor of this system is compact and runs a tiny preemptive operating system that monitors various metrics and is entitled to take remapping decisions through code migration techniques. This approach that endows the architecture with decisional capabilities permits refining application implementation at run-time according to various criteria. Experiments based on simple policies are presented on various applications that demonstrate the benefits of such an approach.

scholarly journals KABouM: Knowledge-Level Action and Bounding Geometry Motion Planner

2018 ◽  
Vol 61 ◽  
pp. 323-362 ◽  
Author(s):  
Andre Gaschler ◽  
Ronald P. A. Petrick ◽  
Oussama Khatib ◽  
Alois Knoll
Keyword(s):  
Information Gain ◽  
Convex Bodies ◽  
Convex Polyhedra ◽  
Knowledge Level ◽  
Convex Decomposition ◽  
Symbolic Information ◽  
Sensing Actions ◽  
Swept Volumes ◽  
Geometric Knowledge ◽  
Swept Volume

For robots to solve real world tasks, they often require the ability to reason about both symbolic and geometric knowledge. We present a framework, called KABouM, for integrating knowledge-level task planning and motion planning in a bounding geometry. By representing symbolic information at the knowledge level, we can model incomplete information, sensing actions and information gain; by representing all geometric entities--objects, robots and swept volumes of motions--by sets of convex polyhedra, we can efficiently plan manipulation actions and raise reasoning about geometric predicates, such as collisions, to the symbolic level. At the geometric level, we take advantage of our bounded convex decomposition and swept volume computation with quadratic convergence, and fast collision detection of convex bodies. We evaluate our approach on a wide set of problems using real robots, including tasks with multiple manipulators, sensing and branched plans, and mobile manipulation.

scholarly journals Modified Prostate Volume Algorithm Improves Transrectal US Volume Estimation in Men Presenting for Prostate Brachytherapy

Radiology ◽  
2009 ◽  
Vol 250 (1) ◽  
pp. 273-280 ◽  
Author(s):  
Peter J. MacMahon ◽  
Ann-Marie Kennedy ◽  
Darra T. Murphy ◽  
Michael Maher ◽  
Michelle M. McNicholas
Keyword(s):  
Prostate Volume ◽  
Volume Estimation ◽  
Prostate Brachytherapy ◽  
Volume Algorithm

Collision Detection between Rigid Bodies Based on Adaptive Test Lines

2013 ◽  
Vol 756-759 ◽  
pp. 3189-3193
Author(s):  
Xiao Dong Shao ◽  
Wei Gao ◽  
Huan Lling Liu
Keyword(s):  
Collision Detection ◽  
Detection Efficiency ◽  
Rigid Bodies ◽  
Collision Point ◽  
Detection Algorithms ◽  
Adaptive Test ◽  
Rigid Objects ◽  
Continuous Collision Detection ◽  
Volume Algorithm ◽  
Swept Volume

A novel algorithm, which can check the collision point of rigid objects continuously and solve the problem of penetration and crossing in collision detection effectively, is presented in this paper. At each simulation moment, the adaptive test lines (ATLs) are first constructed based on the velocity vector of the moving object and then the intersection between the ATLs and the environment is calculated. The collision happens when the intersection is not empty and the collision point is obtained through crossing-frame processing. By checking the interference between body and ATLs instead of bodies, we greatly improve the detection efficiency. It avoids missing collisions for an object with arbitrary shape or in any motion states. Simulation results show that our algorithm runs faster than the general continuous collision detection algorithms and has similar detection effects to the swept volume algorithm.

A METHOD FOR THE REGISTRATION OF A KNOWN CAD MODEL INTO THE WORKSPACE FRAME OF A ROBOT

2010 ◽  
Vol 19 (03) ◽  
pp. 281-304 ◽  
Author(s):  
DIMITRIS ARISTOS ◽  
SPYROS TZAFESTAS
Keyword(s):  
3D Modeling ◽  
A Priori ◽  
Robot Kinematics ◽  
Coordinate Frame ◽  
Cad Model ◽  
Robot Arm ◽  
Rigid Object ◽  
On Line ◽  
Position And Orientation ◽  
Robotic Applications

In many robotic applications it is required to manipulate a specific rigid object whose CAD model is known a priori, but its position and orientation in space are unknown. This category of tasks includes piercing, painting or iron welding on some specific points of the rigid object's surface. For this kind of tasks to become feasible the CAD data of the rigid object should be registered into the robot's workspace frame, so that the robot arm becomes aware of the position and orientation of the rigid body with reference to its own flange or base coordinate frame. In order to achieve this goal several techniques from the fields of image processing, 3D modeling and robot kinematics should be combined. This paper provides a convenient combination of such methods that can be used for the successful on-line registration of a rigid object's CAD data to the robot's workspace frame.

Incremental interpretation in second language sentence processing

2006 ◽  
Vol 9 (1) ◽  
pp. 71-88 ◽  
Author(s):  
JOHN N. WILLIAMS
Keyword(s):  
Second Language ◽  
Structural Analysis ◽  
Sentence Processing ◽  
Similar Pattern ◽  
Memory Task ◽  
Task Demands ◽  
Syntactic Processing ◽  
On Line ◽  
Incremental Interpretation ◽  
And Task

The degree to which native and non-native readers interpret English sentences incrementally was investigated by examining plausibility effects on reanalysis processes. Experiment 1 required participants to read sentences word by word and to make on-line plausibility judgements. The results showed that natives and non-natives immediately computed the plausibility of the preferred structural analysis, which then affected ease of reanalysis. Experiment 2 required participants to read the same sentences word by word in order to perform a memory task. The natives showed a similar pattern of results to Experiment 1, whereas for the non-natives plausibility effects were delayed. However, the non-natives still appeared to be performing immediate syntactic reanalysis. It is concluded that syntactic processing was person- and task-independent, whereas the incrementality of interpretation was more dependent on task demands for the non-natives than for the natives.

Sign in / Sign up

Export Citation Format

Share Document