Autonomous Vehicle Trajectory Planning With Fallback Stopping Costs
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicles face challenges in safely stopping or pulling over when new trajectory updates are not received in time, as existing systems lack effective fallback planning mechanisms that consider potential stopping locations and costs.
Innovation Solution
The method involves identifying potential stopping locations and states, determining fallback costs using a backward induction approach, and planning a trajectory that accounts for these costs to ensure safe stopping, by generating a nominal trajectory influenced by fallback considerations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If the autonomous vehicle follows a nominal trajectory to reach the destination, then the vehicle can proceed efficiently towards its goal, but if a new trajectory is not received in time, the vehicle cannot safely stop or pull over
Solution Approach 1:
The system performs preliminary action by pre-planning a fallback trajectory that includes a fallback portion with safe stopping instructions before the vehicle actually needs to stop. The fallback trajectory is generated in advance with an initial section extracted from the nominal trajectory, and the vehicle is instructed to follow this prepared fallback plan if trajectory updates are delayed, enabling safe stopping without compromising the efficiency of nominal trajectory following during normal operation.
2Ease of operation
If the system generates a separate fallback trajectory independently of the nominal trajectory, then the fallback portion can be planned without constraints, but the fallback portion cannot influence the nominal trajectory to improve stopping outcomes
Solution Approach 1:
The system merges the fallback trajectory planning with the nominal trajectory by extracting the initial section of the nominal trajectory to form the first portion of the fallback trajectory. This combination allows the fallback portion to be influenced by the nominal trajectory's route and timing information, enabling the fallback planning to optimize stopping locations along the intended path while maintaining the benefits of independent fallback trajectory generation for safety-critical decisions.
3Device complexity
If the fallback portion simply follows the geometry of the nominal trajectory with modified speed profile, then the planning is simple, but the vehicle cannot stop in the best possible circumstances
Solution Approach 1:
The system applies dynamics by making the fallback trajectory adaptive rather than static. Instead of simply modifying the speed profile of the nominal trajectory, the system generates a dynamic fallback trajectory that can adjust the vehicle's path, speed, and stopping location based on real-time conditions. The fallback portion is planned to bring the vehicle to a stop at the best possible location by considering current state and potential future states, enabling optimal stopping outcomes while maintaining computational efficiency through structured planning approaches.
Data Source
AI summary
Aspects of the disclosure provide for controlling an autonomous vehicle. For example, a set of potential stopping locations may be identified based on a current location of an autonomous vehicle. A set of potential states may be identified for the autonomous vehicle. A fallback cost for each potential state of the set of potential states to reach each potential stopping location of the set of potential stopping locations may be determined. A trajectory for an autonomous vehicle may be planned based on a route to a destination and the determined fallback costs. The autonomous vehicle may be controlled according to at least a portion of the trajectory.