Automated Vehicle Safe State Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In highly automated driving, existing systems fail to react effectively to errors, particularly in situations involving longer braking distances and higher speeds, as they rely on pre-calculated emergency trajectories and fixed fallback strategies, which are not adaptable to changing conditions.
Innovation Solution
A method and apparatus for operating a vehicle that selects a safe state from a plurality of options based on parameters such as traffic situation, sensor availability, hardware status, and driver health, allowing the vehicle to adapt and safely stop or navigate in fully automated mode, even after an emergency trajectory is calculated.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Power
If a pre-calculated emergency trajectory is used, then the computing power requirement is reduced, but the system cannot react to events occurring after the emergency trajectory is calculated
Solution Approach 1:
The patent implements dynamic recalculation of emergency trajectories at multiple evaluation points along the planned path, rather than using a single pre-calculated trajectory. This allows the system to adapt to new events occurring during execution while maintaining computational efficiency through hierarchical evaluation.
Solution Approach 2:
The system pre-calculates multiple emergency trajectories for different evaluation points along the planned path before execution begins. This preliminary preparation enables rapid response to errors without requiring intensive real-time computation when an error occurs.
2Device complexity
If a fixed fallback strategy is implemented, then the system complexity is reduced, but the system cannot adapt to different traffic situations and sensor availability
Solution Approach 1:
The patent changes the parameters used for fallback strategy selection based on the current situation, including sensor availability, traffic conditions, and error type. The system evaluates multiple parameters (sensor functionality, traffic situation, planned path characteristics) to dynamically select appropriate safe states rather than using a single fixed strategy.
Solution Approach 2:
The fallback strategy is segmented into multiple discrete safe states (e.g., stop in current lane, stop in breakdown lane, stop in rest area) rather than a single monolithic strategy. This segmentation allows the system to select the most appropriate safe state based on specific situational parameters.
3Ease of manufacture
If the vehicle stops on the roadway in all error situations, then the implementation is simplified, but safety is compromised in situations where halting is not the safest option
Solution Approach 1:
The patent applies different safe state selection criteria based on local conditions (traffic situation, sensor availability, error type, planned path characteristics). Rather than a universal stop-everywhere approach, the system tailors the safe state selection to the specific local context, evaluating multiple factors to determine the safest option for each situation.
Data Source
AI summary
A method for operating a vehicle, the vehicle being guided in fully automated fashion, and if an error is detected during the fully automated guidance, a safe state being selected from a plurality of safe states as a function of one parameter, the vehicle being guided in fully automated fashion into the selected safe state. Also described is an apparatus for operating a vehicle, as well as to a computer program.