Vehicle Motion Planning With Occluded Region Scoring
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous and semi-autonomous vehicles face increased computational demands due to the rapid growth of occluded regions in their environment, which are not adequately addressed by existing motion planning systems.
Innovation Solution
A system that includes sensors to detect surroundings, a perception module to generate a map, an occlusion computing module to identify occluded regions, a filtering module to remove unimportant occlusions, a scoring module using neural networks to assess the importance of filtered occlusions, and a motion planning module to modify the vehicle's trajectory based on these scores.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the motion planning system processes all occluded regions in the map, then the completeness of environmental perception is improved, but the computational load increases significantly
Solution Approach 1:
The system segments occluded regions into different priority levels based on their importance to the vehicle's trajectory. The scoring module divides occluded regions into multiple categories (e.g., high priority, medium priority, low priority) based on factors such as proximity to the trajectory, size of the occluded region, and potential impact on safe navigation. This segmentation allows the motion planning system to process only the most critical occluded regions in detail while handling less critical regions with reduced computational resources.
Solution Approach 2:
The system applies different processing qualities to different occluded regions based on their local importance. High-priority occluded regions that are close to the vehicle's trajectory or represent potential hazards receive detailed processing and analysis, while low-priority occluded regions receive minimal processing or are filtered out entirely. This local quality approach ensures computational resources are concentrated on the most relevant regions rather than uniformly processing all occluded areas.
2Productivity
If the system filters out more occluded regions, then the computational efficiency is improved, but the risk of missing important hazards increases
Solution Approach 1:
The system performs preliminary scoring and filtering of occluded regions before the main motion planning process. The scoring module pre-evaluates all occluded regions and assigns importance scores based on predefined criteria (distance to trajectory, occlusion size, object type, etc.). This preliminary action identifies and filters out low-priority occluded regions that are unlikely to impact safe navigation, allowing the motion planning system to focus computational resources on high-priority regions while maintaining safety.
Solution Approach 2:
The system incorporates feedback mechanisms where the importance scores of occluded regions are continuously updated based on the vehicle's changing trajectory, speed, and environmental context. As the vehicle moves and its trajectory is adjusted, the scoring module re-evaluates occluded regions to ensure that newly emerging hazards are not missed. This feedback loop maintains safety by dynamically adapting the filtering criteria to the current operational context.
Data Source
AI summary
A system for a vehicle includes sensors that sense surroundings of the vehicle. A perception module generates a map of the surroundings of the vehicle. The map comprises objects around the vehicle. An occlusion computing module computes occluded regions in the map. The occluded regions are regions around the vehicle that are occluded by one or more of the objects around the vehicle. A filtering module filters none, or one or more of the occluded regions from the map. The map comprises a plurality of filtered occluded regions after the filtering. A scoring module scores the filtered occluded regions based on importance of the filtered occluded regions to a trajectory of the vehicle. A motion planning module modifies the trajectory of the vehicle based on importance scores of the filtered occluded regions. A propulsion module propels the vehicle according to the modified trajectory.