Autonomous Driving Map Reliability Verification Using Sensor Maps
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicles rely on stored maps that may be unreliable due to inaccurate vehicle pose data or environmental changes, leading to potential safety hazards if the planned trajectory crosses unintended areas.
Innovation Solution
The system generates an estimated map using sensor data from cameras and lidar sensors, which is then compared to the stored map to determine consistency. This process involves image and lidar data conversion, feature extraction, and confidence value calculation to assess the reliability of the stored map and planned trajectory.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the autonomous vehicle uses stored map data for navigation, then the vehicle can operate with pre-planned routes, but the reliability of navigation decreases when the stored map is inaccurate or outdated
Solution Approach 1:
The system performs preliminary actions by generating an estimated map from current sensor data before comparing it with the stored map. This advance verification allows the system to identify discrepancies between the stored map and actual environment, ensuring navigation reliability by detecting outdated or inaccurate map information before it affects routing decisions
Solution Approach 2:
The system implements feedback by comparing the estimated map (generated from current sensor data) with the stored map data. This comparison provides feedback on the accuracy and currency of the stored map, allowing the system to identify when stored map information has become unreliable due to environmental changes or inaccuracies
2Measurement precision
If the autonomous vehicle relies on accurate vehicle pose data, then the planned trajectory can be precise, but safety hazards increase if the pose data is inaccurate causing the vehicle to cross unintended areas
Solution Approach 1:
The system introduces an intermediary verification layer by comparing the planned trajectory against both the stored map and the estimated map generated from current sensor data. This intermediary check detects discrepancies in vehicle pose accuracy by identifying when the planned path would cross areas that differ between maps, preventing safety hazards before they occur
Solution Approach 2:
The system applies preliminary anti-action by proactively identifying potential safety hazards through map comparison before the vehicle executes the planned trajectory. By detecting inconsistencies between stored and estimated maps that indicate pose inaccuracies, the system can prevent the vehicle from following an incorrect path that would lead to crossing unintended areas
3Measurement precision
If the system verifies map reliability by comparing stored maps with sensor-generated estimated maps, then navigation accuracy improves, but the computational complexity and processing time increase
Solution Approach 1:
The system extracts only the essential elements needed for verification by comparing specific features and characteristics between the stored map and estimated map, rather than processing all sensor data. This selective extraction approach maintains navigation accuracy through targeted comparison while reducing overall computational complexity
Solution Approach 2:
The system performs partial verification by focusing the map comparison on critical areas and features relevant to navigation safety, rather than verifying every detail of the maps. This partial action approach provides sufficient navigation accuracy for safe operation without the full computational burden of complete map verification
Data Source
AI summary
Techniques for verifying a reliability of map data are discussed herein. In some examples, map data can be used by a vehicle, such as an autonomous vehicle, to traverse an environment. Sensor data (e.g., image data, lidar data, etc.) can be received from a sensor associated with a vehicle and may be used to generate an estimated map and confidence values associated with the estimated map. When the sensor data is image data, images data from multiple perspectives or different time instances may be combined to generate the estimated map. The estimated map may be compared to a stored map or to a proposed vehicle trajectory or corridor to determine a reliability of the stored map data.