Autonomous Driving Behavior Control for Human-Like Maneuver Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous motor vehicles are perceived as difficult to understand by human users and other road users, leading to reduced acceptance and potential critical traffic situations, especially when interacting with non-autonomous vehicles.
Innovation Solution
A method for automated driving that adjusts target variables influencing driving behavior based on detected surroundings, using a system with target variables that have variable levels, allowing for 'humanized' driving behavior by selecting appropriate driving maneuvers that balance competing objectives such as minimizing acceleration, lane changes, and maintaining safe distances.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If autonomous vehicles use fixed, predetermined driving maneuvers for each road section, then the control system is simple and reliable, but the driving behavior appears robotic and difficult to understand for human users
Solution Approach 1:
The patent applies dynamics by making the driving behavior adaptable and flexible rather than fixed. The control system dynamically selects from multiple alternative driving maneuvers (first, second, and third maneuvers) based on real-time evaluation of target variable levels. This allows the autonomous vehicle to exhibit varied, human-like driving behavior that is easier to understand, while maintaining a structured control architecture that manages complexity through hierarchical decision-making.
2Reliability
If autonomous vehicles consistently apply the same driving rules, then safety and predictability are improved, but interaction with human-driven vehicles becomes rigid and may lead to critical traffic situations
Solution Approach 1:
The patent implements parameter changes by evaluating multiple target variables (such as travel time, fuel consumption, emissions, and driver comfort) and adjusting the selection of driving maneuvers based on their current levels. The system can shift between conservative safe maneuvers and more aggressive efficient maneuvers depending on the evaluated parameters, enabling flexible adaptation to different traffic situations while maintaining safety through structured evaluation frameworks.
3Productivity
If autonomous vehicles optimize for multiple competing target variables simultaneously, then overall performance is improved, but the decision-making process becomes more complex and harder to interpret
Solution Approach 1:
The patent applies segmentation by dividing the complex decision-making process into distinct, evaluable components. Multiple target variables (travel time, fuel consumption, emissions, comfort) are segmented and evaluated separately, each with its own level assessment. This modular approach allows the system to consider multiple competing objectives simultaneously while maintaining interpretability through structured, independent evaluation of each parameter before integrating them into the final maneuver selection.
Data Source
Figure 1
Figure 2
AI summary
The invention relates to a method for automated, in particular autonomous, driving of a motor vehicle (60). The method comprises providing one or more target variables (20, 30, 40, 50) for influencing the driving behavior of the motor vehicle (60), wherein the target variables (20, 30, 40, 50) have a variable level (22, 32, 42, 52). The method further comprises detecting the environment of the motor vehicle (60) and generating driving behavior of the motor vehicle (60) depending on the detected environment of the motor vehicle (60) and the levels (22, 32, 42, 52) of the target variables (20, 30, 40, 50). The procedure also involves adjusting the levels (22, 32, 42, 52) of the target variables (20, 30, 40, 50) depending on the driving behavior of the motor vehicle (60) generated.