AV Fleet Simulation for Routing and Facility Placement Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicle fleets face inefficiencies in navigating streets due to unoptimized metrics such as pick-up locations, drop-off locations, routes, and facility positions, which are exacerbated by the large number of vehicles and time-consuming data collection processes.
Innovation Solution
A simulated environment is created where autonomous vehicles can navigate and record synthetic sensor data, allowing a server to analyze and optimize routing updates and facility placements to improve fleet-wide metrics like power consumption and trip times using machine-learning models and heatmaps.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If autonomous vehicles operate without centralized optimization, then each vehicle can navigate independently, but fleet-wide inefficiencies occur due to unoptimized metrics such as pick-up locations, drop-off locations, routes, and facility positions
Solution Approach 1:
The patent creates a simulated environment that copies real-world geographic regions, streets, and autonomous vehicle operations. This virtual replica allows for parallel simulation of multiple routing versions and collection of cost metrics without interfering with actual fleet operations, thereby resolving the contradiction between achieving fleet-wide optimization and avoiding time-consuming real-world data collection.
Solution Approach 2:
The system performs preliminary simulations and optimizations in the virtual environment before implementing routing updates in the real fleet. By pre-evaluating different routing versions and determining optimal pick-up locations, drop-off locations, and facility positions in advance, the system achieves fleet-wide efficiency improvements without causing delays in actual operations.
2Measurement precision
If multiple routing versions are tested in the real fleet, then optimal routing can be determined, but operational disruptions and safety risks increase
Solution Approach 1:
The patent uses a virtual copy of the fleet operating in a simulated environment to test multiple routing versions. This allows for precise measurement and comparison of different routing strategies' cost metrics without disrupting real fleet operations or compromising safety, as the simulations run independently in parallel with actual fleet deployment.
Solution Approach 2:
The system cushions against potential operational disruptions by conducting all routing version tests in a protected virtual environment before deploying any changes to the real fleet. This preliminary testing phase acts as a buffer, ensuring that only thoroughly validated routing updates are implemented, thereby maintaining fleet operation stability while achieving routing optimization accuracy.
3Productivity
If fleet-wide optimization is implemented across thousands of vehicles, then overall efficiency improves, but the complexity of managing and coordinating all vehicles increases
Solution Approach 1:
The patent segments the fleet optimization problem into independent simulation instances that can be processed separately. Each routing version is evaluated independently in the virtual environment, and results are aggregated to determine the optimal strategy. This segmentation approach allows for fleet-wide optimization without requiring complex real-time coordination of all vehicles, as the simulation framework handles the coordination automatically.
Solution Approach 2:
The virtual environment serves as an intermediary layer between individual vehicle operations and fleet-wide optimization goals. This intermediary simulation framework absorbs the coordination complexity, allowing real vehicles to operate independently while still achieving fleet-wide efficiency improvements through the intermediary's optimization algorithms and routing update distributions.
Data Source
AI summary
The subject disclosure relates to techniques for optimizing autonomous vehicle activities at a fleet level. A process of the disclosed technology can include measuring a first set of cost metrics for a first plurality of autonomous vehicles (AVs) associated with a first vehicle routing version, wherein the first set of cost metrics comprises energy consumption values for two or more of the first plurality of AVs, measuring a second set of cost metrics for a second plurality of AVs associated with a second vehicle routing version, wherein the second set of cost metrics comprises energy consumption values for two or more of the second plurality of AVs, and comparing the first set of cost metrics with the second set of cost metrics to determine routing updates configured to reduce fleetwide energy consumption.