AMR Path Planning with Adaptive Sampling and Two-Way Search
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current techniques for autonomous mobile robot (AMR) route planning and navigation, such as 2D grid A* and Rapidly Exploring Random Tree (RRT), are computationally expensive and unsuitable for continuous multi-planning scenarios.
Innovation Solution
The proposed path planning system and method incorporate a Probabilistic Road Map (PRM) with an adaptive sampling technique and a rapid two-way search process, enabling efficient global path computation for AMR navigation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If conventional path planning methods (2D grid A*, RRT) are used, then path quality is maintained, but computational complexity and processing time increase significantly
Solution Approach 1:
The path planning process is divided into two independent phases: roadmap construction (offline) and path query (online). The roadmap is pre-computed by sampling configuration space and connecting nearby samples, separating the computationally intensive exploration phase from the query phase, thereby reducing real-time computational complexity while maintaining path quality
Solution Approach 2:
The roadmap is constructed in advance before actual path queries are needed. By pre-computing the probabilistic road map structure and storing it for future use, the system eliminates redundant computation during real-time navigation, significantly reducing processing time while preserving path optimality
2Reliability
If conventional path planning methods are used, then complete path exploration is achieved, but processing time and latency increase
Solution Approach 1:
The system dynamically adapts between offline roadmap construction and online path querying based on operational context. The probabilistic sampling approach dynamically explores configuration space during roadmap construction, ensuring completeness while allowing the system to switch to rapid query mode for real-time operations, reducing overall processing time
Solution Approach 2:
The probabilistic road map maintains continuous coverage of the configuration space through systematic sampling and connection of nearby points. This continuous structure allows for efficient interpolation and pathfinding during queries, ensuring path completeness while enabling rapid response times through the pre-established topological relationships
3Adaptability or versatility
If destination position changes frequently, then navigation adaptability is improved, but computational latency increases causing movement jitter
Solution Approach 1:
The probabilistic road map is constructed in advance, creating a reusable navigation framework that remains valid even when destination positions change. This preliminary structure allows the system to rapidly compute new paths by querying the existing roadmap without re-computing the entire configuration space, eliminating jitter caused by frequent recalculations
Solution Approach 2:
The navigation system separates the static roadmap structure from the dynamic query process. When destination positions change, only the query phase needs to be executed, not the entire path planning process. This segmentation allows rapid adaptation to new destinations while maintaining computational efficiency, preventing movement jitter
Data Source
Figure 1
Figure 2A
Figure 2B
AI summary
Techniques are disclosed to facilitate multi-agent path planning and to enable navigation for robotics systems. The path planning may include a Probabilistic Road Map (PRM), Deterministic Road Map (DRM) or any other kind of road maps with an adaptive sampling technique having a rapid two-way search to compute a global path. With the road map and Two-way search strategy, a parallel or simultaneous check for feasibility at each iteration may be performed. The road map can be saved for future inquiries.