Weighted Confidence Estimation for Driving Assistance

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing driving assistance systems face challenges in reliability due to detection inaccuracies and errors from sensor limitations, leading to potentially confusing or unsafe system responses.

Innovation Solution

A method for a driving assistance system that uses weighted confidence estimates to improve prediction reliability by associating weights with basic confidence estimates based on detection error impacts, allowing for more accurate assessments and reduced probabilities of wrong predictions.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If additional sensor equipment is provided to enhance system reliability, then detection accuracy and prediction reliability improve, but hardware complexity and costs increase

Engineering Contradiction:
Improveprediction reliabilityVSAvoidhardware complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent changes the parameter of confidence estimation by introducing weighted confidence estimates that account for detection errors. Instead of adding sensors, the system modifies how existing sensor data is evaluated by associating weights with basic confidence estimates based on the impact of potential detection errors, thereby improving prediction reliability through parameter optimization rather than hardware expansion

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The patent creates a virtual model of detection uncertainty by generating weighted confidence estimates that replicate the effect of having multiple sensor sources. The weighting mechanism copies the reliability assessment function that would otherwise require additional sensor equipment to provide, allowing the system to evaluate prediction reliability without physical duplication of sensing capabilities

Inventive Principle:
Principle #26Copying

2Measurement precision

If additional sensor equipment is provided to enhance system reliability, then detection accuracy improves, but costs increase

Engineering Contradiction:
Improvedetection accuracyVSAvoidhardware costs
Core Design Contradiction:
Measurement precisionVSQuantity of substance

Solution Approach 1:

The patent improves detection accuracy by changing the parameter of confidence weighting. By associating different weights with basic confidence estimates based on detection error impact, the system optimizes the use of existing sensor data to achieve higher effective detection accuracy without the cost of additional sensors

Inventive Principle:
Principle #35Parameter changes

3Ease of operation

If sensor data inaccuracies are not considered, then system operation is simpler, but prediction reliability deteriorates

Engineering Contradiction:
Improvesystem operation simplicityVSAvoidprediction reliability
Core Design Contradiction:
Ease of operationVSReliability

Solution Approach 1:

The patent applies preliminary action by pre-calculating and associating weights with basic confidence estimates before predictions are made. The system prepares confidence weighting information in advance based on detection error characteristics, so that when predictions are generated, the reliability assessment is already embedded in the weighted confidence estimates, maintaining operational simplicity while improving reliability

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent introduces feedback by using detection error information to adjust confidence estimates. The system feeds back knowledge about potential sensor inaccuracies into the confidence calculation process, allowing predictions to be weighted according to their expected reliability, thereby improving prediction accuracy without complicating the core prediction logic

Inventive Principle:
Principle #23Feedback

Data Source

PatentEP2865576B1Composite confidence estimation for predictive driver assistant systems
Publication Date: 2018.07.04 HONDA RES INST EUROPE
  • EP2865576B1 patent drawingFigure 1
  • EP2865576B1 patent drawingFigure 2
  • EP2865576B1 patent drawingFigure 3

AI summary

The invention relates to a driving assistance system (100) including a prediction subsystem (110) in a vehicle. According to a method aspect of the invention, the method comprises the steps of accepting a set of basic environment representations (120); allocating a set of basic confidence estimates (122); associating weights to the basic confidence estimates (122); calculating (128) a weighted composite confidence estimate for a composite environment representation; and providing the weighted composite confidence estimate as input for an evaluation of a prediction (130, 132) based on the composite environment representation.