Autonomous Vehicle Dispatch Using Proactive ODD Capability Trips
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicle dispatch systems face challenges in accurately determining the operational capabilities of different vehicle types due to incomplete or inaccurate operational domain data (ODD), which can lead to safety issues and inefficient trip routing.
Innovation Solution
The system generates and monitors trips to demonstrate specific capabilities of autonomous vehicles, using vehicle movement data to augment and correct ODD data, thereby improving the accuracy of vehicle selection for trips based on their actual performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If the dispatch system uses available ODD data to determine vehicle capabilities, then the system can operate with limited data, but the accuracy and completeness of vehicle capability determination deteriorates
Solution Approach 1:
The system proactively generates trips for autonomous vehicles to demonstrate capabilities before actual dispatch decisions are made. This preliminary data collection through generated trips allows the system to build accurate ODD profiles in advance, resolving the contradiction by ensuring capability data is available before it is needed for dispatch operations
Solution Approach 2:
Autonomous vehicles perform self-demonstration of capabilities by executing generated trips and collecting performance data. The vehicles essentially service their own capability profiling by autonomously completing trips that test their operational domains, eliminating the need for external manual testing while improving data accuracy
2Measurement precision
If the system generates and monitors trips to demonstrate vehicle capabilities, then the accuracy of ODD data improves, but the time and resources required for data collection increases
Solution Approach 1:
The system dynamically adjusts trip generation parameters such as trip frequency, route complexity, and capability testing focus based on the current state of ODD data completeness. When certain capabilities are already well-documented, the system reduces trip generation for those parameters and focuses resources on under-documented capabilities, optimizing the time-cost versus accuracy-gain tradeoff
Solution Approach 2:
The system generates trips selectively rather than continuously, performing partial action only when needed to fill specific gaps in ODD data. Trip generation is triggered by identified deficiencies in capability data rather than operating continuously, reducing overall time loss while maintaining sufficient data accuracy for safe dispatch decisions
3Reliability
If trips are tailored to the capabilities of individual vehicle types, then safety is enhanced, but the complexity of trip planning and vehicle selection increases
Solution Approach 1:
The system segments the fleet into distinct vehicle types with standardized capability profiles based on demonstrated ODD data. Each vehicle type has a predefined set of capabilities and limitations, allowing the dispatch system to match trips to vehicle types rather than evaluating individual vehicle performance in real-time. This segmentation reduces planning complexity while maintaining safety through appropriate capability-matching
Data Source
AI summary
A method for dispatching trips to a plurality of autonomous vehicles is disclosed. Operational design domain (ODD) parameters for an autonomous vehicle of a first vehicle type are derived based on vehicle movement data. The vehicle movement data describes a plurality of trips executed by a set of autonomous vehicles of the first vehicle type. A first ODD parameter of the ODD parameters is identified that meets a criterion. A first new trip is generated based on the identified ODD parameter. An autonomous vehicle is selected to execute the first new trip. A request to execute the first new trip is then sent to the selected autonomous vehicle.