AI Processing Anomaly Detection via Cross-Result Comparison

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current methods fail to effectively detect anomalies in AI processing results, particularly in environments with uncertainties and unexperienced events, leading to inaccurate predictions and misjudgments.

Innovation Solution

An anomaly detection apparatus that estimates high-level information through initial processing and compares it with subsequent processing results, using an evaluation unit to detect anomalies based on the comparison.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If AI processing is used to replace human judgment in complex tasks, then processing efficiency is improved, but processing accuracy deteriorates in uncertain environments

Engineering Contradiction:

Improveprocessing efficiencyVSAvoidprocessing accuracy

Core Design Contradiction:

ProductivityVSMeasurement precision

Solution Approach 1:

The patent introduces an anomaly detection apparatus as an intermediary system between the AI processing system and the final decision-making process. This mediator detects anomalies in AI processing results by comparing them against expected patterns and historical data, thereby maintaining high processing efficiency while improving accuracy in uncertain environments through additional verification

Inventive Principle:

Principle #24Intermediary (Mediator)

2Adaptability or versatility

If AI models are trained with more diverse data to handle unexperienced events, then adaptability is improved, but processing time increases

Engineering Contradiction:

Improvehandling unexperienced eventsVSAvoidprocessing time

Core Design Contradiction:

Adaptability or versatilityVSLoss of time

Solution Approach 1:

The patent implements preliminary anomaly detection mechanisms that are prepared in advance through training on diverse data patterns. The anomaly detection apparatus pre-learns various anomaly patterns and can quickly identify them during operation, thereby improving adaptability to unexperienced events without significantly increasing processing time during actual deployment

Inventive Principle:

Principle #10Preliminary action

3Reliability

If multiple verification methods are introduced to ensure AI processing accuracy, then reliability is improved, but device complexity increases

Engineering Contradiction:

Improveprocessing reliabilityVSAvoidsystem complexity

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The patent segments the verification process into distinct functional modules within the anomaly detection apparatus, including data acquisition modules, analysis modules, and determination modules. This segmentation allows each component to perform a specific verification function, improving reliability through comprehensive checking while managing system complexity through modular design

Inventive Principle:

Principle #1Segmentation

Data Source

PatentUS12197200B2Anomaly detection apparatus, anomaly detection method, and recording medium

Publication Date: 2025.01.14 NEC CORP

AI summary

Provided is an anomaly detection apparatus 10 capable of detecting anomalies of processing results. The anomaly detection apparatus 10 estimates first processing result information in accordance with first processing, using data regarding a target, compares second processing result data, for the data, obtained by second processing that differs from the first processing with the first processing result information, and detects an anomaly of the second processing according to the comparison result.