Robot Motion Planning With Cached Swept Volumes for Dynamic Obstacles
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Motion planning systems for robots face inefficiencies in collision detection and path adjustment due to the need for real-time computation of swept volumes, especially in dynamic environments with multiple robots, leading to delays and increased computational complexity.
Innovation Solution
The use of cached swept volume profile data generated during the construction phase to improve motion planning efficiency by precomputing and storing swept volume information, allowing for faster collision detection and path adjustments without the need for on-the-fly calculations during the query phase.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If swept volume computation is performed in real-time during the query phase, then collision detection accuracy is improved, but computational delay increases
Solution Approach 1:
The patent precomputes and caches swept volume profiles for all edges in the probabilistic roadmap during the construction phase. This preliminary action stores the spatial occupation information of robot configurations along each edge, so that during the query phase, collision detection can directly use these precomputed profiles without performing real-time swept volume calculations, thereby eliminating computational delay while maintaining accuracy
Solution Approach 2:
The patent creates cached copies of swept volume profiles and stores them in association with the probabilistic roadmap. These copies represent the spatial occupation patterns of robot motion along each edge, allowing the system to use these pre-generated profiles for rapid collision detection during path queries without re-computing the actual swept volumes
2Speed
If swept volume profiles are precomputed for all edges, then collision detection speed is improved, but memory usage increases
Solution Approach 1:
The patent extracts only the essential swept volume profile information needed for collision detection and stores it in a cached format. By taking out only the necessary spatial occupation data rather than storing complete swept volume computations, the system achieves fast collision detection while managing memory usage efficiently
Solution Approach 2:
The swept volume profiles are precomputed and cached during the construction phase, allowing the query phase to use these profiles directly without re-computation. This preliminary preparation stores the essential collision information in advance, enabling fast detection while the data structure is optimized for memory efficiency
3Measurement precision
If probabilistic roadmap is constructed with dense sampling, then path planning accuracy is improved, but construction time increases
Solution Approach 1:
The patent performs dense sampling and constructs the probabilistic roadmap with comprehensive edge coverage during the construction phase. By performing this preliminary action of creating a dense roadmap, the system establishes an accurate representation of the configuration space that enables precise path planning, while the associated swept volume profiling is also completed in advance to prepare for efficient queries
4Device complexity
If multiple robots are coordinated using individual path planning, then system complexity is reduced, but collision risk increases
Solution Approach 1:
The patent merges the path planning processes of multiple robots by having them share a common probabilistic roadmap structure. Each robot's planned path is represented as a sequence of edges from the same roadmap, and swept volume profiles are cached in association with these shared edges, enabling the system to maintain low individual complexity while detecting and preventing collisions between robots through profile comparison
Data Source
Figure 1
Figure 2A
Figure 2B
AI summary
Techniques and systems are disclosed for using swept volume profile data cached in association with a PRM to improve various aspects of motion planning for a robot. In some implementations, a first map representing possible paths to be travelled by a robot within a physical area is generated. An initial path for the robot within the first map is determined. Data indicating a second map representing a path to be travelled by a movable object within the physical area is obtained. A potential obstruction associated with one or more edges included in the subset of edges is detected. An adjusted path for the robot within the first map is then determined based on the potential obstruction.