Robotic Path Planning in Confined Spaces Using Inverse Kinematic Clustering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional robotic systems fail to generate efficient and collision-free paths in confined spaces, particularly when dealing with multiple target locations, leading to increased assembly time and potential collisions during tasks like fastener installation in complex structures like aircraft wingboxes.
Innovation Solution
A method involving calculating inverse kinematic solutions, clustering these solutions, and generating collision-free paths between clusters using a robot control system that includes a computer and robot controller, which enables the robot to efficiently navigate through confined spaces by identifying reachable and unreachable process points and adjusting its path accordingly.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional algorithms are used to identify the robot path using kinematics, then the path can be generated, but the robot may collide with obstacles and the assembly time increases
Solution Approach 1:
The patent segments the configuration space into discrete cells and the path planning into multiple phases (roadmap construction, path finding). This segmentation allows the system to systematically explore the configuration space while avoiding obstacles, resolving the contradiction between reliable collision avoidance and efficient path generation.
Solution Approach 2:
The patent performs preliminary action by pre-computing a roadmap of the configuration space before actual path execution. This roadmap contains pre-identified collision-free regions and transitions, allowing the robot to quickly query paths during assembly without real-time computation delays, thus improving both reliability and productivity.
2Productivity
If the robot travels to many different locations to install fasteners, then the assembly task is completed, but the path becomes inefficient and time-consuming
Solution Approach 1:
The patent applies dynamics by making the path planning adaptive to the specific task requirements. The system dynamically adjusts the path based on the configuration space roadmap, selecting optimal sequences of configuration cells that minimize travel time while visiting all necessary process points for fastener installation.
Solution Approach 2:
The patent changes parameters by representing the configuration space in terms of discrete cells with specific properties (obstacle presence, accessibility). This parameterization allows efficient computation of optimal paths by transforming the continuous path planning problem into a discrete graph search problem, improving assembly efficiency.
3Adaptability or versatility
If the configuration space is densely populated with process points, then the task coverage is complete, but the path planning complexity increases
Solution Approach 1:
The patent segments the dense configuration space into a grid of discrete cells, where each cell represents a manageable unit. This segmentation reduces the complexity of path planning by transforming a continuous high-dimensional problem into a discrete graph problem, while maintaining complete task coverage through systematic cell exploration.
Solution Approach 2:
The patent substitutes the mechanical path planning process with a computational approach using graph theory and search algorithms. Instead of relying on continuous kinematic calculations, the system uses discrete graph search on the configuration space roadmap, reducing computational complexity while maintaining task coverage.
Data Source
AI summary
A method of automatic path planning for at least one robot within a confined configuration space, the robot including an arm having a plurality of joints and an end effector coupled to the arm. The method includes entering a plurality of process points into a computer, each process point being a location wherein the arm is to be positioned to perform a task, calculating one or more inverse kinematic solutions for each process point, clustering the inverse kinematic solutions into a set of clusters, and generating collision free paths between the clusters in the confined configuration space.