Autonomous Route Selection With Pareto Control of Manual Driving
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current routing methods for autonomous driving vehicles do not adequately account for road clearance and transition stress, leading to inefficiencies and safety concerns, particularly in adverse weather or road conditions, which hinders the ability to determine safe and optimal routes in real-time.
Innovation Solution
A vehicle routing method that computes and evaluates alternative routes based on travel time, manual driving time, transition count, and a stressing cost function, using a Pareto optimization approach to select routes that minimize manual driving and transitions, while considering road clearance values and weighting factors for stress analysis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If more information is taken into account and analyzed to ensure safe autonomous driving, then safety is improved, but routing efficiency deteriorates
Solution Approach 1:
The routing problem is segmented into multiple independent criteria (travel time, manual driving time, transition count) that are evaluated separately. Each criterion is optimized independently through Pareto optimization, allowing comprehensive safety analysis without compromising overall routing efficiency.
Solution Approach 2:
The routing system dynamically adapts to varying road clearance conditions by computing alternative routes and evaluating them based on current conditions. The system transitions between different routing strategies depending on whether road clearance is available, optimizing both safety and efficiency for each specific situation.
2Reliability
If road clearance is withdrawn due to adverse conditions, then safety is improved, but manual driving time increases
Solution Approach 1:
The system performs preliminary evaluation of alternative routes before road clearance is withdrawn. By pre-computing multiple route options and their characteristics, the system can quickly switch to manual driving on pre-identified alternative routes when safety concerns arise, minimizing the time loss associated with route re-planning.
Solution Approach 2:
The system changes routing parameters dynamically based on road clearance status. When road clearance is withdrawn, the optimization criteria shift from prioritizing autonomous driving segments to prioritizing routes with minimal manual driving time, allowing flexible adaptation to safety requirements.
3Ease of operation
If transitions from autonomous to manual driving are minimized, then passenger comfort is improved, but routing flexibility deteriorates
Solution Approach 1:
The system adds the dimension of transition count as a separate optimization criterion alongside travel time and manual driving time. By evaluating routes in this multi-dimensional space through Pareto optimization, the system can identify routes that minimize transitions while maintaining adequate routing flexibility for various driving conditions.
Data Source
Figure 1
Figure 2a~2b
Figure 3
AI summary
The present invention is directed to a method and to a device for determining routes for autonomous driving. The following steps are executed: computing K alternative routes from a route source to a route destination; evaluating each one of the K alternative routes, wherein the evaluating comprises determining a Pareto front of the K alternative routes, wherein the Pareto front of the K alternative routes comprises a set of Pareto-optimized routes that are Pareto-optimized with regard to route constraints comprising travel time T, manual driving time TM, and transition count M that indicates a count of transitions from the autonomous driving to a manual driving in a route; if the set of Pareto-optimized routes comprises more than P routes, reducing number of Pareto-optimized routes in the set of Pareto-optimized routes to P; and outputing the Pareto-optimized routes of the set of Pareto-optimized routes as the determined routes for autonomous driving.