China
Africa
Explore chapters and articles related to this topic
Robot Path and Motion Planning
Published in Jitendra R. Raol, Ajith K. Gopal, Mobile Intelligent Autonomous Systems, 2016
In fact, motion planning is aimed at providing the robots with the capabilities of automatically deciding and executing a sequence of motions in order that the robot performs a task without colliding with any of the other objects in the robot’s environment. In the path planning, as we have seen in several previous section, the design of only geometric, that is, kinematic details of the locations and orientations of the robot are taken into account, whereas in the trajectory planning the design of the linear and angular velocities is taken into consideration. Autonomous agents like mobile robots that sense, plan and act in real and/or virtual worlds, need path planning and motion-trajectory planning algorithms. These algorithms and systems are utilized for representing, capturing, planning, controlling and rendering motions of several physical objects individually or collective (as discussed in Chapter 17) [3], and the applications are: (i) manufacturing, (ii) mobile robots, (iii) computational biology, (iv) computer-assisted surgery and (v) digital actors. Thus, the goal of motion planning is to compute motion strategies, for example: (a) geometric paths, (b) time-parameterized trajectories and (c) sequence of sensor-based motion commands. Also the aim is to achieve high-level goals, for example: (a) go to destination D without colliding with obstacles, (b) assemble a product P, (c) build map of environment E and (d) detect and find an object O.
Warehouse Automation: An Example
Published in Laxmidhar Behera, Swagat Kumar, Prem Kumar Patchaikani, Ranjith Ravindranathan Nair, Samrat Dutta, Intelligent Control of Robotic Systems, 2020
Laxmidhar Behera, Swagat Kumar, Prem Kumar Patchaikani, Ranjith Ravindranathan Nair, Samrat Dutta
In the case of industrial manipulators where one does not have access to internal motor controllers, motion planning refers to providing suitable joint angle position (or velocity) trajectories needed for taking the robot from one pose to another. In other words, motion planning becomes a path planning problem which is about finding a way to point poses between the current pose and the desired end-effector pose. The problem of generating collision free paths for manipulators with increasingly larger number of links is quite complex and has attracted considerable interest over last couple of decades. Readers can refer to [351] for an overview of these methods. These methods could be primarily divided into two categories - local and global. Local methods start from a given initial configuration and step toward final configuration by using local information of the workspace. Artificial potential field-based methods [352] [353] [354] are one such category of methods where the search is guided along the negative gradient of artificially created vector fields. On the other hand, global methods use search algorithms over the entire workspace to find suitable paths. Some of the examples of global methods are probabilistic roadmaps (PRM) [355] [356]and cell-decomposition based C-Space methods [357] [358]. Rapidly exploring random tree (RRT) [359] is one of the most popular PRM method used for path planning. Many of these state-of-the-art algorithms are available in the form of the open motion planning library (OMPL) [289] which has been integrated into several easy-to-use software packages like Moveit! [288], Kautham [360], and OpenRave [361].
Robot Learning from Human: Hierarchical Interactions
Published in Changliu Liu, Te Tang, Hsien-Chung Lin, Masayoshi Tomizuka, Designing Robot Behavior in Human-Robot Interactions, 2019
Changliu Liu, Te Tang, Hsien-Chung Lin, Masayoshi Tomizuka
As a well-explored field in the past decades, there are already plenty of approaches developed for robotic motion planning, including grid-based [196, 84, 61, 50], sampling-based [105, 93, 92, 66] and optimization-based methods [128, 136, 177, 188].
Online motion planning based on swept volume search with replanning using sequential quadratic programming
Published in Advanced Robotics, 2023
Takuya Iwasaki, Yutaka Takase, Solvi Arnold, Keisuke Takeshita, Kimitoshi Yamazaki
Motion planning is a significant area of study in robotics. Graph-based methods combined with dynamic programming [6] are commonly used traditional approaches in mobile robotics. Although sampling-based methods are a feasible approach for multiple DoF robots [1,2,7,8], they typically require several or dozens of seconds to establish one planning result for a robot with six or more DoFs, necessitating a faster processing method. Notably, methods for enhancing the efficiency of motion planning by referring to past motion experiences have also been proposed. Leven et al. [9] and Kuwata et al. [10] proposed methods that reuse or adjust the results of past probabilistic roadmaps (PRMs). Ferguson et al. [11] proposed methods that reuse or adjust the results of an existing tree method generated by RRT.
Human–robot collaboration: a fabrication framework for the sequential design and construction of unplanned spatial structures
Published in Digital Creativity, 2020
Edvard P. G. Bruun, Ian Ting, Sigrid Adriaenssens, Stefana Parascho
The process of generating new geometry is based on a random sampling approach inspired by the Rapidly-Exploring Random Tree (RRT) class of robotic motion planning algorithms (LaValle and Kuffner 2001). These algorithms are used to randomly and incrementally explore a domain space to find a feasible trajectory, and are particularly successful since their sampling strategy biases them to search unexplored areas of a domain. Thus, the exploratory nature of an RRT sampling approach is perfectly suited to the goal proposed in this project: using randomness to explore a large physical design-space in the process of sequentially constructing an unplanned structure. Note that only the sampling strategy in our project (i.e. generating potential positions for the next piece of a structure to aggregate at each time step) is borrowed from the RRT algorithm. Figure 4 shows examples of 2-D patterns generated in a circular domain by an RRT* algorithm (Karaman and Frazzoli 2011). The final tree structure and path is always different, and is unknown at the start of the process; the tree is grown by drawing a random sample in each iteration and connecting it to an existing branch. The final 3-D form of our structure (see Section 4) is visually reminiscent of this kind of branching output from an RRT algorithm as the solution path is expanded outward.
Connected and automated road vehicles: state of the art and future challenges
Published in Vehicle System Dynamics, 2020
Tulga Ersal, Ilya Kolmanovsky, Neda Masoud, Necmiye Ozay, Jeffrey Scruggs, Ram Vasudevan, Gábor Orosz
Motion planning algorithms can be cast as either path or trajectory generation problems. In the path planning formulation, a path of given length is constructed in the configuration space . The solution does not prescribe how fast the path should be followed, instead the velocity is prescribed by a lower-level feedback controller. In contrast, the trajectory generation problem constructs a trajectory in state space (parameterised by time) along a given time horizon. This enables one to incorporate vehicle dynamics and handle dynamic obstacles. We briefly describe the various algorithms that have been proposed to solve either the path or the trajectory generation problem; however, several recent papers provide longer introductions to these topics [26,27].