Surprising Event Detection in ADAS Using Sensor Fusion

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Advanced Driver Assistance Systems (ADAS) face challenges in effectively testing and updating their performance due to the inability to simulate infinite combinations of driving situations, weather conditions, and environmental changes, leading to reduced effectiveness under unpredictable conditions and potential sensor failures.

Innovation Solution

A system and method that include an event detector to identify surprising events using a combination of input signals from vehicle operators and object detection sensors, generating an event data file that is analyzed off-board to update the ADAS system, allowing for the collection and analysis of rare events and sensor inconsistencies, and issuing software and firmware patches to improve system performance.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If ADAS systems are tested under controlled conditions, then system reliability is improved, but the ability to handle unpredictable real-world conditions deteriorates

Engineering Contradiction:

Improvesystem reliabilityVSAvoidability to handle unpredictable conditions

Core Design Contradiction:

ReliabilityVSAdaptability or versatility

Solution Approach 1:

The system collects real-world event data from vehicles equipped with ADAS, analyzes it off-board to identify surprising events and sensor inconsistencies, and uses this feedback to continuously improve system reliability through updated algorithms and sensor calibration

Inventive Principle:

Principle #23Feedback

Solution Approach 2:

The system performs preliminary analysis of event data off-board before deploying updated ADAS systems, allowing comprehensive testing and validation of improvements in a controlled environment before real-world deployment

Inventive Principle:

Principle #10Preliminary action

2Measurement precision

If comprehensive event data is collected from multiple sources, then measurement precision is improved, but device complexity increases

Engineering Contradiction:

Improveevent detection precisionVSAvoiddata collection system complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

An off-board analysis system serves as an intermediary that receives, processes, and analyzes comprehensive event data from multiple vehicle sources, centralizing the complexity of data integration and analysis outside the individual vehicles

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The system merges data from multiple sensors and vehicles into unified event data files, combining diverse data sources to improve measurement precision while managing complexity through integrated processing

Inventive Principle:

Principle #5Merging (Combining)

Data Source

PatentUS10427655B2Systems and methods for detecting surprising events in vehicles

Publication Date: 2019.10.01 HARMAN INT IND INC

AI summary

A driver assistance system is presented, including an event detector which comprises a set of rules defining a surprising event based on signals reflecting a vehicle operator input signals from an object detection sensor. An event data file generator is configured to generate an event data file according to rules comprised in the event detector, the event data file comprising a video signal received from a camera, a signal from at least one dynamic vehicle sensor, and target object information received from the object detection sensor. The event data file generator is further configured to initiate data file generation responsive to a surprising event being detected by the event generator, and wherein the contents of the data file are specified by the rules comprised in the event detector. In this way, surprising events may be collected and analyzed off-board in order to generate updates for the driver assistance system.