Virtual Lane CIPO Detection for Autonomous Vehicles
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current autonomous driving systems lack efficient methods to detect closest in-path objects (CIPOs) that may cause collisions, as they rely on actual lane data and do not effectively account for the vehicle's predicted path or virtual lane based on its state and motion.
Innovation Solution
The system determines a virtual lane based on the autonomous vehicle's current state and motion prediction, using sensor data from cameras, LIDAR, and RADAR, without relying on map data, and classifies objects within this virtual lane as CIPOs, triggering emergency operations if necessary.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the system uses actual lane data from map data to detect objects, then the detection is based on established road boundaries, but it fails to detect objects on the predicted path that are outside the actual lane
Solution Approach 1:
The system performs preliminary action by determining a virtual lane based on the ADV's current operating state and motion prediction before detecting objects. This allows the system to proactively identify potential collision risks on the predicted path, rather than waiting for objects to enter the actual lane. The virtual lane is calculated in advance using sensor data and motion predictions, enabling early detection of CIPOs that may not yet be in the actual lane but could cause collisions.
2Measurement precision
If the system relies on map data for lane information, then it has accurate road boundary data, but it cannot account for dynamic changes in vehicle path due to current state and motion
Solution Approach 1:
The system applies dynamics by determining a virtual lane based on the ADV's current operating state (position, speed, heading) and motion prediction rather than relying solely on static map data. The virtual lane dynamically adjusts to reflect the vehicle's actual trajectory, which may differ from the planned path due to real-time conditions. This dynamic approach allows the system to adapt to changing vehicle states and predict where the vehicle will actually travel, improving detection of objects on the real path.
3Ease of operation
If the system detects objects only in the actual lane, then it follows traditional lane-based detection methods, but it misses objects that are not in the lane but could cause collisions with the moving vehicle
Solution Approach 1:
The system transitions from a static, two-dimensional lane-based detection approach to a dynamic, trajectory-based approach by introducing the virtual lane concept. Instead of only detecting objects within the fixed lane boundaries from map data, the system creates a moving detection zone based on the vehicle's predicted path. This dimensional change from static lane coordinates to dynamic trajectory coordinates allows the system to detect objects in three-dimensional space-time, capturing risks that extend beyond traditional lane boundaries.
Data Source
AI summary
In one embodiment, in addition to detecting or recognizing an actual lane, a virtual lane is determined based on the current state or motion prediction of an ADV. A virtual lane may or may not be identical or similar to the actual lane. A virtual lane may represent the likely movement of the ADV in a next time period given the current speed and heading direction of the vehicle. If an object is detected that may cross a lane line of the virtual lane and is a closest object to the ADV, the object is considered as a CIPO, and an emergency operation may be activated. That is, even though an object may not be in the path of an actual lane, if the object is in the path of a virtual lane of an ADV, the object may be considered as a CIPO and subject to a special operation.