Automated Driving Scenario Interpretation for Computational Load Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current automated driving systems face high computational efforts in interpreting environment data, which can lead to inefficiencies and increased processing demands.
Innovation Solution
The method breaks down complex scenarios into elementary situations, each with assigned monitoring areas and release criteria, allowing only relevant data to be processed, reducing the computational load by focusing on specific areas relevant to the current situation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If all environmental data is processed for automated driving, then comprehensive situation awareness is achieved, but computational effort increases significantly
Solution Approach 1:
The patent segments the environment into multiple monitoring areas (front, rear, left, right) and further divides each monitoring area into elementary situations (e.g., intersection, pedestrian crossing, curve). This segmentation allows the system to process only relevant data for each specific situation rather than analyzing all environmental data uniformly, thereby reducing computational effort while maintaining comprehensive situation awareness.
Solution Approach 2:
The patent applies local quality by assigning different evaluation criteria and processing depths to different monitoring areas and elementary situations. For example, the front monitoring area receives more detailed analysis than the rear area, and specific elementary situations like intersections trigger more rigorous evaluation than open roads. This localized processing approach optimizes computational resources based on the actual risk and importance of each area.
2Reliability
If comprehensive environmental data is evaluated, then safety is improved, but processing time increases
Solution Approach 1:
The patent performs preliminary classification of environmental data into monitoring areas and elementary situations before detailed analysis. By pre-organizing data into structured categories with assigned priorities and processing requirements, the system prepares the evaluation framework in advance, enabling faster real-time processing while maintaining comprehensive safety checks for all relevant situations.
3Measurement precision
If detailed environmental analysis is performed, then detection precision is improved, but computational load increases
Solution Approach 1:
The patent applies partial action by performing detailed analysis only on monitoring areas and elementary situations that are currently relevant to the vehicle's operation. For example, when the vehicle is approaching an intersection, only the front and side monitoring areas receive detailed analysis, while other areas use standard processing. This selective detailed analysis maintains detection precision where needed while reducing overall computational load.
Data Source
Figure 1
Figure 2
Figure 3a
AI summary
The invention relates to a method for automated driving of a motor vehicle (50), comprising the following steps: providing environmental data (7), recognizing objects (113) in the environmental data (7), selecting a current scenario (12, 100) by a scenario interpretation device (3) based on the recognized objects (113), wherein a scenario (12, 100) has at least one elementary situation (13, 101, 101.1 - 101.X) to which at least one monitoring area (110.1 - 110.7) and a release criterion (114) are assigned as attributes (108), furthermore, successively for each of the elementary situations (13, 101, 101.1 - 101.X) of the current scenario (12, 100): estimating a target point (111.1 - 111.7), planning a trajectory (112.1 - 112.7) to the target point (111.1 - 111.7) by a maneuver planning device (4), queries for information about the monitoring area (110.1 - 110.7) in the environment perception device (2), the scenario interpretation device (3) verifies the presence of at least one release criterion (114), and if at least one release criterion (114) is met, automated driving along the trajectory (112.1 - 112.7) to the destination point (111.1 - 111.7) by a controller (5). The invention further relates to an associated device (1).