Autonomous Robot Trajectory Generation for Obstacle Avoidance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous driving mobile robots face challenges in generating optimal local routes with minimal calculation time while avoiding obstacles, often resulting in inefficient trajectories or isolation due to previous planning steps and high computational requirements.
Innovation Solution
The autonomous driving mobile robot is equipped with a state determination module and a trajectory generation module in a PC that determines the robot's state and generates stationary or avoidance trajectories, allowing it to stop near the destination and escape isolation by setting collision or isolation avoidance destination points, reducing repetitive movements and calculation time.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If incremental planning method is used to generate local route, then robot can move without modifying global planning route, but robot may be isolated or generate inefficient trajectory due to previous time step selections
Solution Approach 1:
The patent performs arrival determination before completing the full trajectory generation. By checking whether the robot is already at or near the destination at the beginning of the process, it prevents unnecessary trajectory calculations and avoids the isolation problem that occurs when incremental planning repeatedly generates suboptimal paths. This preliminary check resolves the contradiction by ensuring reliable route generation only when necessary.
Solution Approach 2:
The patent implements a stationary trajectory generation step that generates only the necessary portion of the trajectory needed to reach the destination, rather than generating a complete trajectory from start to finish. When the robot is already at or near the destination, this partial action approach avoids the inefficiency of generating full trajectories that would result in repetitive avoidance movements, thus improving movement efficiency while maintaining reliability.
2Loss of time
If sampling-based optimal tree method is used for real-time route planning, then calculation time is reduced, but local minima problem occurs making it difficult to find optimal trajectory
Solution Approach 1:
The patent performs arrival determination before trajectory generation to check if the robot is already at or near the destination. This preliminary action prevents the need for complex sampling-based calculations when unnecessary, reducing calculation time significantly while avoiding the local minima problem by not attempting to generate trajectories when the robot is already at the goal.
Solution Approach 2:
The patent extracts and handles the arrival determination as a separate, independent step before the main trajectory generation process. By separating this critical check from the sampling-based optimal tree calculations, it eliminates the local minima problem that would otherwise occur during unnecessary trajectory computations, while maintaining fast real-time performance when the robot is near the destination.
3Reliability
If conventional local route planning methods are used to consider robot state and environment at each time step, then complete trajectory can be generated, but a lot of calculation time is required
Solution Approach 1:
The patent performs arrival determination before initiating the trajectory generation process. This preliminary check ensures that complete trajectory generation is only performed when necessary (when the robot is not already at or near the destination), thereby maintaining trajectory completeness reliability while significantly reducing the average calculation time by avoiding redundant computations.
Solution Approach 2:
The patent implements a stationary trajectory generation step that generates only the necessary portion of the trajectory when the robot is already at or near the destination, rather than generating a complete trajectory from start to finish. This partial action approach maintains the reliability of having a complete plan when needed while reducing calculation time by avoiding excessive computations when the robot is already at the goal.
Data Source
AI summary
The present disclosure relates to an autonomous driving robot that can move without user intervention. The autonomous driving mobile robot able includes a PC that receives vision data from a vision device, sets a route, generates operation commands, and thereby controls a motor drive board and a power supply board, in which the PC is configured with a state determination module and a trajectory generation module and a state determination module determines whether the robot is at a location less than or equal to a first threshold distance from a destination is performed after determining a position and direction of the robot. When the robot is located at the location less than or equal to the first threshold distance from the destination, the trajectory generation module generates a stationary trajectory where the speed of the robot decreases and becomes zero as the robot approaches the destination.