Autonomous Driving VIL Verification with Virtual Road Synchronization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing vehicle-in-the-loop (VIL) systems for autonomous driving function verification are limited in simulating diverse road environments, requiring frequent changes to real test roads, which is inconvenient and costly, and struggle to verify autonomous driving performance in complex scenarios without physically altering the test road environment.
Innovation Solution
A VIL system-based autonomous driving performance verification method that creates a virtual road environment using a simulator, converting the real position and posture of an autonomous vehicle into a virtual one, allowing verification without the need for a corresponding real test road, using a road environment database, simulation engine, rendering module, emulation module, calculation module, and interlocking module to generate and synchronize virtual sensor data and position information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a real test road is used for verification, then the autonomous vehicle can verify function in a real environment, but the system cannot verify in various virtual road environments without changing the physical road
Solution Approach 1:
The patent creates virtual copies of road environments through detailed 3D modeling and mapping technologies. Virtual road environments are constructed by copying real-world road geometry, topology, and semantic information from point cloud data and map databases, allowing verification in diverse virtual scenarios without physical road changes
Solution Approach 2:
The patent introduces a calculation module as an intermediary that transforms real vehicle position and posture data into virtual coordinate systems. This mediator enables seamless integration between real vehicle operations and virtual environment verification, allowing the vehicle to verify functions in virtual roads while physically remaining on the same test road
2Adaptability or versatility
If the test road is changed to verify different road environments, then various road scenarios can be verified, but huge temporal and financial burdens are incurred
Solution Approach 1:
Instead of physically changing roads, the system creates multiple virtual copies of different road environments by processing point cloud data and map information. These virtual road environments can be rapidly switched and configured without any physical infrastructure changes, enabling verification across diverse road scenarios including highways, urban streets, and rural roads
Solution Approach 2:
The system pre-processes road environment data by collecting point cloud information, constructing 3D models, and building virtual road environments in advance. This preliminary action allows rapid switching between different virtual road scenarios during verification without time-consuming physical road changes
3Adaptability or versatility
If the test road is changed to verify different road environments, then various road scenarios can be verified, but huge financial costs are incurred
Solution Approach 1:
The patent uses virtual copying of road environments through digital modeling instead of physical road construction or modification. Virtual road environments are generated from existing point cloud data and map information, eliminating the need for expensive physical infrastructure changes while enabling verification in diverse road scenarios
Solution Approach 2:
The system replaces the mechanical approach of physically changing test roads with an information-based virtual environment generation system. By substituting physical road modification with digital model creation and coordinate transformation, the patent eliminates huge financial costs associated with test road construction and modification
4Productivity
If virtual sensor data is generated through emulation, then verification can proceed without real sensor input, but the accuracy of position and posture conversion must be maintained
Solution Approach 1:
The system implements feedback mechanisms where the calculated virtual position and posture are continuously compared with expected values from multiple data sources including point cloud matching, map coordinate systems, and vehicle sensor data. This feedback loop ensures accuracy maintenance while enabling rapid virtual environment verification
Solution Approach 2:
The emulation module creates accurate virtual copies of sensor data by processing real sensor inputs through mathematical models that replicate sensor behavior. This copying approach generates realistic virtual sensor data that maintains the statistical and spatial characteristics of real sensor measurements, enabling accurate verification without real-time sensor dependency
Data Source
Figure 1
Figure 2
Figure 3
AI summary
There is provided a VIL system-based autonomous driving function verification method. According to embodiments of the disclosure, a VIL system enables an autonomous vehicle to verify an autonomous driving function by interlocking with a virtual road environment in any other place, without having to go to a real test road corresponding to a simulated virtual road environment. Accordingly, an autonomous driving function can be rapidly verified based on a VIL system with respect to various virtual road environments without changing a driving place, so that speed and convenience in development of autonomous driving technology can be enhanced.