Automated Driving Discrepancy Scoring and Algorithm Adaptation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing methods for determining a driver's driving behavior in relation to an automated driving system and adapting control algorithms for vehicle fleets lack precision and effectiveness in quantifying discrepancies and optimizing system performance.

Innovation Solution

A method that records driver control commands and vehicle trajectories during manual driving, simulates automated driving system trajectories, and calculates a score value to measure discrepancies between driver and automated driving behaviors. This method also adapts control algorithms based on statistically evaluated score values from multiple vehicles.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If driver behavior is compared to automated driving system behavior using existing methods, then a qualitative assessment can be obtained, but precise quantification of discrepancies is lacking

Engineering Contradiction:
Improveprecision of driving behavior discrepancy measurementVSAvoidcomplexity of behavior comparison system
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent transforms qualitative driving behavior comparisons into quantitative measurements by introducing score values that represent the probability of trajectory correspondence. This parameter transformation enables precise measurement of discrepancies between driver and automated system behaviors through statistical evaluation of multiple trajectory probabilities.

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The patent introduces simulated trajectories as an intermediary element that bridges the comparison between actual driver behavior and automated driving system behavior. By generating multiple simulated trajectories with associated probabilities, the system creates a measurable intermediate representation that enables precise discrepancy quantification without direct complex comparison.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Productivity

If control algorithms are adapted for vehicle fleets without systematic evaluation, then deployment is faster, but effectiveness in optimizing system performance is reduced

Engineering Contradiction:
Improvespeed of control algorithm deploymentVSAvoideffectiveness of system performance optimization
Core Design Contradiction:
ProductivityVSReliability

Solution Approach 1:

The patent implements a feedback mechanism where score values from multiple vehicles are collected, statistically evaluated, and used to identify positions with relevant discrepancies. This feedback loop enables systematic adaptation of control algorithms based on actual performance data, ensuring both speed and effectiveness in fleet-wide optimization.

Inventive Principle:
Principle #23Feedback

Solution Approach 2:

The patent performs preliminary statistical evaluation of score values across the vehicle fleet before deploying adapted control algorithms. By pre-identifying positions with relevant discrepancies and preparing optimized algorithms in advance, the system enables rapid deployment without sacrificing optimization effectiveness.

Inventive Principle:
Principle #10Preliminary action

3Reliability

If statistical evaluation of score values is performed for all positions, then comprehensive optimization is achieved, but computational resources and time are excessive

Engineering Contradiction:
Improvecompleteness of optimization coverageVSAvoidtime for statistical evaluation
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The patent extracts and focuses computational resources on only those positions where relevant discrepancies are identified through statistical evaluation of score values. By filtering out positions without significant discrepancies, the system achieves comprehensive optimization where needed while minimizing unnecessary computational expenditure on positions that do not require optimization.

Inventive Principle:
Principle #2Taking out (Extraction)

Data Source

PatentEP4351947B1Method for determining a driving behaviour, and method for adapting control algorithms of automated driving systems
Publication Date: 2025.06.04 MERCEDES BENZ GROUP AG
  • EP4351947B1 patent drawingFigure 1~2
  • EP4351947B1 patent drawingFigure 3
  • EP4351947B1 patent drawingFigure 4

AI summary

The invention relates to a method for determining a driving behaviour of a driver in relation to a driving behaviour of an automated driving system of a vehicle (2). According to the invention, actuating commands (SB) from the driver and a manually navigated trajectory (r) of the vehicle (2) are detected during manual driving operation of the vehicle (2). In a first calculation path (R1), determination is performed as to which automated actuating commands the automated driving system would generate in a respectively current actual position of the vehicle (2) on the manually navigated trajectory (r) if the automated driving system were active. In a second calculation path (R2), a trajectory (ß, ß', ß0 to ß2) of the vehicle (2) which the vehicle (2) would cover with the automated driving system active is simulated. Depending on the detected actuating commands (SB) from the driver and the automated actuating commands and/or the manually navigated trajectory (r) and the simulated trajectory (ß, ß', ß0 to ß2), at least one score value (SW1 to SW4) is determined as a measure of a discrepancy between the driving behaviour of the driver and the driving behaviour of the automated driving system. The invention further relates to a method and a device (12) for adapting control algorithms of automated driving systems of vehicles (2) of a vehicle fleet.