Drivable Area Boundary Segmentation for Road User Trajectory Prediction

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Autonomous vehicles struggle to accurately predict the trajectories of uncertain road users, such as pedestrians and cyclists, due to the lack of meaningful information about smaller discrete target locations within large drivable areas, leading to potential collisions and hazardous situations.

Innovation Solution

The system segments drivable area boundaries into logical segments that represent potential goals for actors, using a threshold-based concavity analysis to identify edges and form segments that reflect pedestrian destinations, enabling accurate trajectory prediction and collision avoidance.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Device complexity

If drivable areas are represented as large segments, then the representation is simplified and easier to process, but meaningful information about smaller discrete target locations cannot be inferred

Engineering Contradiction:
Improvedrivable area representation complexityVSAvoidtarget location information
Core Design Contradiction:
Device complexityVSLoss of information

Solution Approach 1:

The drivable area boundary is segmented into multiple logical segments based on concavity analysis. Each logical segment represents a potential target location for uncertain road users. This segmentation divides the large drivable area into meaningful smaller units while maintaining the overall simplified representation structure.

Inventive Principle:
Principle #1Segmentation

2Measurement precision

If drivable area boundaries are analyzed in detail to identify discrete target locations, then trajectory prediction accuracy is improved, but processing complexity and computational load increase

Engineering Contradiction:
Improvetrajectory prediction accuracyVSAvoidboundary analysis complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The system changes the parameter used for boundary analysis from general geometric features to specific concavity measurements. By calculating concavity values and comparing them against thresholds, the system efficiently identifies logical segments without requiring complex detailed analysis of the entire boundary geometry.

Inventive Principle:
Principle #35Parameter changes

3Productivity

If traditional drivable area representation is used, then processing is faster and simpler, but collision avoidance with uncertain road users becomes less reliable

Engineering Contradiction:
Improveprocessing speedVSAvoidcollision avoidance reliability
Core Design Contradiction:
ProductivityVSReliability

Solution Approach 1:

The system performs preliminary segmentation of the drivable area boundary into logical segments before trajectory prediction. This pre-processing step creates a structured representation that enables more reliable collision avoidance by identifying potential target locations in advance, while keeping the overall process efficient through threshold-based classification.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP4275014B1Methods and system for predicting trajectories of uncertain road users by semantic segmentation of drivable area boundaries
Publication Date: 2025.11.19 VOLKSWAGEN GROUP OF AMERICA INVESTMENTS LLC
  • EP4275014B1 patent drawingFigure 1
  • EP4275014B1 patent drawingFigure 2A
  • EP4275014B1 patent drawingFigure 2B

AI summary

Methods and systems for controlling navigation of an autonomous vehicle for traversing a drivable area are disclosed. The methods include receiving information relating to a drivable area that includes a plurality of polygons, identifying a plurality of logical edges that form a boundary of the drivable area, sequentially and repeatedly analyzing concavities of each the plurality of logical edges until identification of a first logical edge that has a concavity greater than a threshold, creating a first logical segment of the boundary of the drivable area. This segmentation may be repeated until each of the plurality of logical edges has been classified. The method may include creating and adding (to a map) a data representation of the drivable area that comprises an indication of the plurality of logical segments, and adding the data representation to a road network map comprising the drivable area.