Motion Planning for Autonomous Objects Using Guard Area Switching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional motion planning methods for autonomous vehicles are computationally expensive and struggle to provide smooth trajectories in complex scenarios with dynamic obstacles, often requiring frequent re-computation and introducing artifacts or compromising smoothness.
Innovation Solution
A method that continuously computes and updates a time-dependent trajectory by switching between different motion planning procedures based on obstacle detection, using a guard area to determine when to switch between obstacle avoidance, sliding motion, and tracking goals, thereby reducing computational resources and ensuring smooth vehicle motion.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional motion planning methods are used to compute trajectories avoiding obstacles, then obstacle avoidance is achieved, but computational cost increases and trajectory smoothness deteriorates
Solution Approach 1:
The motion planning computation is segmented into multiple procedures (first motion planning procedure for when no obstacle is detected, second motion planning procedure for when obstacle is detected). The system switches between these segmented procedures based on obstacle detection state, avoiding the need to continuously execute computationally expensive obstacle avoidance computations when obstacles are not present.
Solution Approach 2:
The system dynamically switches between different motion planning procedures based on the detected obstacle state. When an obstacle is detected, it transitions to the second procedure; when not detected, it transitions to the first procedure. This dynamic adaptation optimizes computational resources while maintaining reliable obstacle avoidance when needed.
2Stability of the object's composition
If conventional methods use interpolating functions to smooth trajectories, then smoothness is improved, but artifacts are introduced
Solution Approach 1:
The patent introduces a guard area as an intermediary zone around the obstacle. By computing trajectories that pass through this guard area rather than directly to the obstacle center, the system naturally achieves smoother transitions without requiring post-processing interpolation functions that would introduce artifacts.
3Stability of the object's composition
If global optimization methods are used to find optimal trajectories, then smooth solutions are generated, but re-computation frequency increases in high dynamic environments
Solution Approach 1:
The computation is segmented into two procedures with different optimization depths. The first procedure (no obstacle detected) uses a simpler, faster computation suitable for dynamic environments. The second procedure (obstacle detected) uses more comprehensive optimization. This segmentation reduces overall re-computation frequency while maintaining smoothness when obstacles are present.
Solution Approach 2:
The system changes computational parameters based on obstacle detection state. When no obstacle is detected, it uses parameters optimized for speed (first procedure). When an obstacle is detected, it switches to parameters optimized for smoothness and safety (second procedure). This parameter adaptation reduces unnecessary re-computations in dynamic environments.
4Adaptability or versatility
If sample-based methods like RRT are used, then flexibility to deal with complex scenarios is improved, but trajectory smoothness deteriorates
Solution Approach 1:
The system segments the motion planning into two phases: exploration phase using sample-based methods (RRT) to find feasible paths in complex scenarios, and refinement phase that ensures smoothness. By separating these functions into different procedures executed conditionally, the system achieves both flexibility and smoothness without the trade-off present in conventional single-phase methods.
Data Source
AI summary
A method for motion planning for autonomous moving objects (AMO) includes continuously computing or updating a time-dependent trajectory of an AMO to a destination by a computing device avoiding obstacles en route to the destination. The time-dependent trajectory is computed using one or more motion planning computation procedures. The method further includes performing a switching to another motion planning computation procedure when an obstacle is detected or not detected anymore. An obstacle is determined to be detected when a point on the computed trajectory lies within or touches a guard area (GA) around the obstacle. The border of the GA has a certain distance from the detected obstacle.