Autonomous Vehicle Nudge Path Planning Under Lane Space Constraints
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional motion planning for autonomous driving vehicles primarily focuses on reducing speed to avoid collisions without considering options for maneuvering around obstacles, resulting in suboptimal paths that may not be both efficient and safe.
Innovation Solution
A rule-based method is introduced to determine whether to perform a 'nudge' action, which involves path planning to maneuver around obstacles based on lane configuration, current speed, and vehicle width, allowing the vehicle to generate a path that is both efficient and safe by assessing a 'nudge space' within the lane.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the ADV reduces speed to avoid collisions, then safety is improved, but path efficiency deteriorates
Solution Approach 1:
The system dynamically switches between two motion planning modes: speed reduction for safety-critical situations and nudge maneuvering for efficiency optimization. The decision to nudge rather than slow down is made based on real-time assessment of lane configuration, obstacle position, and available space, allowing the system to adapt its collision avoidance strategy to maintain both safety and efficiency
Solution Approach 2:
Instead of resolving the safety-efficiency contradiction along the speed dimension, the system introduces a spatial dimension by performing lateral nudge maneuvers around obstacles. This dimensional shift allows the vehicle to maintain higher speeds while avoiding collisions through sideways movements, thereby improving path efficiency without compromising safety
2Productivity
If the ADV performs nudge maneuver around obstacles, then path efficiency is improved, but system complexity increases
Solution Approach 1:
The motion planning system is segmented into distinct functional modules: a determination module that assesses whether nudge conditions are met, and a path planning module that generates the actual nudge path. This segmentation allows the complex nudge functionality to be implemented as an add-on to conventional motion planning without requiring complete system redesign, thereby managing complexity while achieving improved path efficiency
Solution Approach 2:
The system performs preliminary assessment of lane configuration, boundary types, and available nudge space before committing to a nudge maneuver. By evaluating all necessary conditions in advance and only initiating nudge when conditions are favorable, the system avoids unnecessary complex computations, thereby improving path efficiency while keeping the actual computational burden manageable
3Adaptability or versatility
If the ADV increases nudge space, then maneuverability is improved, but lane boundary safety deteriorates
Solution Approach 1:
The system applies different safety margins and nudge space calculations based on local lane characteristics. When lane boundaries are detected as physical barriers, the system reduces or eliminates nudge space in those directions. When boundaries are virtual lines without physical barriers, the system allows greater nudge space. This localized adaptation of nudge parameters enables improved maneuverability while maintaining boundary safety through context-aware adjustments
Data Source
AI summary
Lane configuration of a lane on which an ADV is driving and a current speed of the ADV are determined. A nudge space is determined based on the lane configuration, the current speed, and a vehicle width of the ADV. Path planning is performed to generate a path to nudge an obstacle, in response to the determining that the nudge space is greater than a predetermined threshold nudge space. The ADV is controlled to drive autonomously according to the planned path to nudge the obstacle.