Manipulator Motion Planning With Global-Local Collision Avoidance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing collision-free motion planning methods for robots struggle to adapt to changing environments, particularly when considering robot dynamics, leading to frequent stops and unnatural movements due to the limitations of global path planning methods and probabilistic kinodynamic approaches.
Innovation Solution
A method combining continuous global planning with local reactive planning, where a global planner repeatedly determines a target path independently and a local planner continuously adjusts the robot's movement, allowing for seamless adaptation to dynamic environments without stopping or reducing speed.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If global path planning methods are used for collision-free motion planning, then a complete path from starting point to target point can be determined, but the robot must stop or reduce speed when replanning is needed due to moving obstacles
Solution Approach 1:
The motion planning system is segmented into two independent planners: a global planner that determines the overall target path and a local planner that handles real-time collision avoidance. This segmentation allows each planner to operate independently at different time scales, enabling continuous movement while maintaining safety.
Solution Approach 2:
The global planner performs preliminary path planning in advance, establishing a target path before the robot begins movement. This preliminary action allows the robot to move along a pre-planned safe path while the local planner handles real-time adjustments, eliminating the need to stop for replanning.
2Reliability
If kinodynamic motion planning with probabilistic methods is used to account for robot dynamics, then robot dynamics can be considered in motion planning, but the computation time becomes unpredictable
Solution Approach 1:
The planning system separates dynamics consideration into the local planner, which operates at a higher frequency with simplified dynamics models, while the global planner focuses on geometric path planning. This segmentation allows deterministic computation times for each planner while still accounting for robot dynamics where critical.
Solution Approach 2:
The system changes the level of dynamics detail considered at different planning levels. The global planner uses simplified or ignored dynamics for path generation, while the local planner incorporates dynamics more explicitly for trajectory adjustment, allowing computation time to remain predictable at each level.
3Productivity
If local reactive methods are used for motion planning in changing environments, then fast response to moving obstacles is achieved, but the robot can get stuck in local minima
Solution Approach 1:
The system segments planning into global and local components where the global planner ensures path completion by providing an overall safe path from start to goal, while the local planner provides fast response to obstacles. This segmentation prevents the local planner from getting stuck in local minima by always having a global path to follow.
Solution Approach 2:
The global target path acts as an intermediary that guides the local reactive planner. When the local planner encounters obstacles or potential local minima, it can reference the global path to determine appropriate corrective actions, ensuring progress toward the goal while maintaining fast local response capabilities.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Method for collision-free motion planning of a first manipulator (14) from a starting point (26) to a target point (28). The method comprises the following steps: repeatedly determining a target path of the first manipulator (14) to a target point (28) using a global planner (34); continuously determining a motion of the first manipulator (14) using a local planner (36) based on the current target path of the global planner; and executing the motion by the first manipulator (14) in parallel with the determination by the global and local planners (34, 36).