Confidence-Determined Road Model for Vehicle Path Planning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Advanced driver-assistance systems (ADAS) and automated driving systems (ADS) face challenges in accurately estimating road models due to uncertainties in vehicle pose and digital map data, leading to inaccuracies in path planning.
Innovation Solution
A vehicle road model system that determines vehicle pose uncertainty and transforms digital map data into an ego-vehicle coordinate system, creating confidence areas around static elements to represent the reliability of road model estimates, thereby propagating pose uncertainties into relevant road model uncertainties.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If vehicle pose estimation is performed using localization data and digital map, then the road model can be continuously updated, but the estimated pose contains uncertainty due to sensor noise and digital map tolerances
Solution Approach 1:
The system performs preliminary transformation of the digital map into the ego-vehicle coordinate system using the estimated pose, then subsequently determines confidence areas around static elements based on pose uncertainty. This preliminary transformation allows the road model to be continuously updated while accounting for uncertainty in a structured manner.
Solution Approach 2:
The patent introduces a new dimensional aspect to the road model by adding confidence areas that represent uncertainty levels. Instead of only providing positional information, the system now provides spatial information with associated confidence levels, transforming the road model from a simple geometric representation to a probabilistic spatial model.
2Loss of information
If confidence areas are determined around static elements based on pose uncertainty, then road model uncertainties are quantified, but the system complexity increases
Solution Approach 1:
The system segments the uncertainty representation by creating discrete confidence areas around individual static elements (lane markings, traffic signs, barriers) rather than treating uncertainty as a single global parameter. This segmentation allows uncertainty to be quantified and managed at the element level, making the complex uncertainty information more tractable and useful for path planning.
Solution Approach 2:
The confidence area determination unit acts as an intermediary between the pose estimation module and the path planning module. It transforms the abstract concept of pose uncertainty into concrete spatial regions (confidence areas) that can be directly used by path planning algorithms, serving as a bridge that translates uncertainty information into actionable data.
Data Source
Figure 1
Figure 2
Figure 3a~3b
AI summary
The present disclosure relates to a method performed by a vehicle road model system (1) for providing a confidence-determined road model (5) for a vehicle (2). The vehicle road model system determines (1001) based on derived localization data, a position and orientation, pose (20), of the vehicle in view of a digital map (3), wherein the vehicle pose is associated with an uncertainty pertinent the localization data and/or the digital map. The vehicle road model system further transforms (1002) - based on the vehicle pose - at least a portion of the digital map into an ego-vehicle coordinate system (4). Moreover, the vehicle road model system determines (1003) based on the vehicle pose and the digital map, a road model (5) in the ego-vehicle coordinate system, which road model (5) comprises one or more static elements (6) positioned in vicinity and ahead of the vehicle pose. Furthermore, the vehicle road model system determines (1004) - based on the uncertainty and the one or more static elements - a respective confidence area (7) surrounding the one or more static elements, wherein respective confidence area represents a region which with a predeterminable confidence level encompasses a real-world position of respective one or more static elements. The disclosure also relates to a vehicle road model system in accordance with the foregoing, a vehicle comprising such a vehicle road model system, and a respective corresponding computer program product and non-volatile computer readable storage medium.