Pre-computed Kinematic Roadmaps for Multi-Agent Collision Avoidance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing robotic navigation systems face challenges in efficiently planning and coordinating paths for multiple robotic devices in complex environments, particularly in large warehouses with many agents, due to limitations in existing multi-agent path finding (MAPF) solutions that restrict kinematics and geometry, leading to suboptimal route planning and potential collisions.
Innovation Solution
The system generates a roadmap with a discrete planning graph, converting it into a Boolean equation to optimize path planning, allowing for efficient coordination of robotic devices by discretizing paths into uniform cost segments and using SAT solvers to find feasible and optimal solutions, which can handle diverse robotic footprints and environments.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If existing multi-agent path finding (MAPF) solutions are used to plan paths for multiple robotic devices, then route planning can be performed, but the solutions restrict kinematics and geometry leading to suboptimal planning and potential collisions
Solution Approach 1:
The patent segments the continuous path planning problem into discrete planning graph nodes and edges, where each node represents a specific pose and edges represent feasible transitions. This discretization allows the system to handle complex kinematic constraints while maintaining computational efficiency, resolving the contradiction between reliability in collision avoidance and adaptability to diverse kinematics.
Solution Approach 2:
The patent introduces an additional dimensional layer by creating a planning graph that incorporates both spatial position and orientation (pose) information. This transforms the traditional 2D path planning into a higher-dimensional state space problem, enabling more accurate collision avoidance while accommodating diverse robotic device geometries and kinematic constraints.
2Productivity
If complex path planning algorithms are used to handle diverse robotic footprints and environments, then optimal routes can be found, but the computational complexity increases
Solution Approach 1:
The patent performs preliminary actions by pre-computing the planning graph structure, including all valid poses and transitions, before actual path planning occurs. This pre-processing step creates a reusable roadmap that can be queried efficiently for multiple agents, reducing the computational complexity of real-time path planning while maintaining optimality for diverse robotic footprints.
Solution Approach 2:
The patent creates a simplified copy or abstraction of the physical environment in the form of a planning graph data structure. This graphical representation captures the essential geometric and kinematic constraints without requiring complex physical simulations during path planning, thereby improving computational efficiency while preserving planning accuracy.
3Ease of operation
If continuous path planning methods are used, then smooth trajectories are achieved, but coordination of multiple robotic devices becomes difficult
Solution Approach 1:
The patent segments continuous trajectories into discrete graph edges connecting pose nodes, where each edge represents a valid motion segment. This segmentation enables independent planning of individual agent paths through the shared planning graph, making multi-agent coordination straightforward while allowing smooth trajectories to be constructed by interpolating between discrete graph nodes when needed.
Data Source
AI summary
A computing device can determine a roadmap having a path for a robotic device in an environment associated with starting and ending poses. The computing device can generate a plurality of trajectories from the starting pose, where each trajectory can include a steering position and a traction velocity directing the robotic device during a planning time interval. For each trajectory of the plurality of trajectories, the computing device can determine a score for the trajectory indicative of advancement from the starting pose toward the ending pose after simulating the steering position and the traction velocity for the planning time interval. The computing device can select, and then store, a nominal trajectory from among the scored plurality of trajectories. The computing device can receive a first request to provide a route though the environment and can send a first response that includes the stored nominal trajectory.