Dual Sensor Processing Paths for Reliable AD Environment Recognition

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current automated driving systems face challenges in validating their reliability and safety, particularly in predicting and quantifying sporadic errors in environment detection, which are sensitive to software and parameter changes, and lack standardized quantitative proof for acceptance criteria.

Innovation Solution

The system decomposes the automated driving system into environment detection, trajectory planning, and driving action subsystems, with independent processing channels for environment detection sensors, allowing for validation of robustness against raw data and employing a dual signal processing path for redundancy and error correction, ensuring maximum reliability in the safety zone.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Extent of automation

If the automated driving system uses environment detection sensors to perceive the driving environment, then the system can achieve automated driving functionality, but the system cannot provide standardized quantitative proof for acceptance criteria and reliability validation

Engineering Contradiction:
Improveautomated driving functionalityVSAvoidvalidation of acceptance criteria
Core Design Contradiction:
Extent of automationVSReliability

Solution Approach 1:

The patent segments the automated driving system into multiple independent signal processing paths (first path and second path), where each path processes sensor data independently. This segmentation enables separate validation of each path's reliability and allows for quantitative assessment of system robustness by comparing results across paths.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent creates a copy of the environment detection system with independent signal processing paths that replicate the same detection functions. By comparing outputs from the original system and its copy, the system can validate detection accuracy and provide quantitative proof of reliability without requiring extensive real-world testing.

Inventive Principle:
Principle #26Copying

2Productivity

If the system processes sensor data through a single signal processing path, then the processing is simple and fast, but the system lacks robustness against sporadic errors and cannot demonstrate electrical/electronic robustness

Engineering Contradiction:
Improvesignal processing speedVSAvoidrobustness against sporadic errors
Core Design Contradiction:
ProductivityVSReliability

Solution Approach 1:

The patent applies different processing characteristics to different signal processing paths. The first path is optimized for speed and standard processing, while the second path is designed with enhanced validation capabilities for robustness assessment. This local differentiation allows simultaneous optimization of processing speed and reliability validation.

Inventive Principle:
Principle #3Local quality

Solution Approach 2:

The patent implements a second signal processing path that performs redundant processing beyond what is strictly necessary for basic functionality. This excessive action in the form of duplicate processing enables statistical validation of sporadic errors while the first path maintains optimal processing speed for normal operation.

Inventive Principle:
Principle #16Partial or excessive action

3Reliability

If the system uses complex validation methods to prove reliability, then the acceptance criteria can be met, but the validation process becomes extremely complex and difficult to standardize

Engineering Contradiction:
Improveacceptance criterion validationVSAvoidvalidation process complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The validation system uses itself to validate its own reliability. The multiple signal processing paths serve dual purposes: they perform the primary function of environment detection and simultaneously serve as validation mechanisms for each other. This self-service approach simplifies the validation process by eliminating the need for external complex validation systems.

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The patent implements feedback loops where the output of one signal processing path is used to validate the other path's results. This mutual feedback mechanism provides automated reliability assessment and generates quantitative validation data without requiring external intervention, thereby reducing validation complexity.

Inventive Principle:
Principle #23Feedback

Data Source

PatentEP3945338B1Environment recognition device for an automated driving system
Publication Date: 2024.12.18 ZF FRIEDRICHSHAFEN AG
  • EP3945338B1 patent drawingFigure 1
  • EP3945338B1 patent drawingFigure 2
  • EP3945338B1 patent drawingFigure 3

AI summary

Signal processing path (P2) for environment detection (SEE) of an automated driving system (AD) comprising initial interfaces (INT) to environment detection sensors (RAD, CAM, LID) to obtain raw data from a safety area (S) of the driving system (AD), a control unit (ECU2) comprising a processing channel for each of the environment detection sensors (RAD, CAM, LID), executed in the individual processing channels to determine environment models from the respective raw data, to combine detected objects by performing logical operations and, depending on the combination, to determine control and/or regulation signals for trajectory planning (THINK) of the driving system (AD), wherein the control unit (ECU2) generates a copy of the control and/or regulation signals.