Obstacle Filtering for Autonomous Driving Planning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous driving vehicles face inefficiencies in planning due to the large number of obstacles in their environment, which increases processing time and reduces planning efficiency without impacting safety.
Innovation Solution
The method involves identifying a subset of obstacles in a lane based on predetermined criteria, expanding their dimensions to match the lane width to prevent nudging, and planning a trajectory around these expanded obstacles, thereby reducing the number of obstacles to process and improving planning efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If all perceived obstacles are considered in trajectory planning, then safety is maintained, but planning efficiency deteriorates due to increased processing time
Solution Approach 1:
The patent segments the set of all perceived obstacles into multiple subsets based on their spatial relationship to the ADV. Specifically, obstacles are divided into: (1) obstacles in the current lane that require full consideration, (2) obstacles in adjacent lanes that may be filtered, and (3) obstacles far from the vehicle that can be excluded. This segmentation allows the planning system to focus computational resources on critical obstacles while ignoring less relevant ones, thereby improving planning efficiency without compromising safety.
Solution Approach 2:
The patent applies different levels of obstacle consideration to different spatial regions. Obstacles in the current lane receive full attention with detailed trajectory planning, while obstacles in adjacent lanes receive reduced attention through filtering based on distance and relevance criteria. This local differentiation of obstacle processing quality optimizes the balance between safety and computational efficiency by allocating more processing resources to high-risk areas.
2Productivity
If the number of obstacles to process is reduced, then planning efficiency improves, but the risk of missing critical obstacles increases
Solution Approach 1:
The patent performs preliminary filtering of obstacles before the main trajectory planning process. By pre-identifying and excluding obstacles that are far from the ADV or located in lanes that will not be used, the system reduces the computational load for subsequent planning while ensuring that all potentially critical obstacles remain in the consideration set. This preliminary action prevents the loss of important obstacles while improving overall processing efficiency.
Solution Approach 2:
The patent implements a feedback mechanism where the planning system continuously monitors the filtered obstacle set and adjusts filtering criteria based on the vehicle's current state and intended maneuvers. If the ADV plans to change lanes, the filtering criteria dynamically adjust to include obstacles in the target lane, ensuring that critical obstacles are not missed while maintaining efficiency during straight-line driving.
Data Source
Figure 1
Figure 2
Figure 3A
AI summary
A system and a method for filtering obstacles to reduce the number of obstacles for an autonomous driving vehicle (ADV) to process in a given planning phase. The ADV can identify a first set of obstacles based on a set of criteria in a first lane where the ADV is travelling, filter out the remaining obstacles in the first lane, and expand each identified obstacle to a width of the first lane from the view of the ADV so that the ADV cannot nudge any of the first set of identified obstacles. When switching from the first lane to a second lane, the ADV can identify a second set of obstacles in the second lane using the same set of criteria, and expand each obstacle in the second set of obstacles to a width of the second lane while keep tracking the first set of identified obstacles. When the lane switching is completed, the ADV can stop tracking the first set of identified obstacles.