Tracking Beacons and AI Models for Post-Transfer Data Anomaly Detection

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Ensuring data security and maintaining control over data movement and access after transmission to a third-party entity is challenging, particularly in terms of geo-location, data routing, and access management.

Innovation Solution

Implementing tracking beacons in data sets prior to transmission, coupled with Machine Learning (ML) models trained on historical patterns, to detect anomalies in data movement and access by analyzing signals from these beacons.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If data is transmitted to a third-party entity outside the internal computing network, then data sharing and collaboration are enabled, but control and surveillance over the data become difficult to maintain

Engineering Contradiction:

Improvedata sharing capabilityVSAvoiddata control

Core Design Contradiction:

Adaptability or versatilityVSReliability

Solution Approach 1:

Tracking beacons are embedded in the data set before transmission to the third-party entity. This preliminary action enables continuous monitoring of the data's location, access, and usage throughout its journey outside the internal network, maintaining control without restricting sharing capabilities.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The tracking beacons continuously provide feedback signals about the data set's status, location, and access events. This feedback mechanism allows the originating entity to monitor and detect anomalies in real-time, ensuring ongoing control over the shared data.

Inventive Principle:

Principle #23Feedback

2Reliability

If traditional monitoring methods are used to track data after transmission, then some level of surveillance is possible, but the system lacks intelligence to differentiate between normal and anomalous data movement

Engineering Contradiction:

Improvedata surveillance capabilityVSAvoidanomaly detection accuracy

Core Design Contradiction:

ReliabilityVSDifficulty of detecting and measuring

Solution Approach 1:

The machine learning models are trained on historical data patterns to autonomously distinguish between normal and anomalous data movement behaviors. This self-service capability eliminates the need for manual rule-setting and enables intelligent, adaptive anomaly detection that improves over time.

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The system transitions from static monitoring thresholds to dynamic, learning-based parameters. The ML models continuously adapt their detection criteria based on observed patterns, enabling accurate differentiation between legitimate data access and potential security threats.

Inventive Principle:

Principle #35Parameter changes

3Difficulty of detecting and measuring

If machine learning models are trained on historical data patterns, then intelligent anomaly detection is achieved, but the system complexity and computational resources increase

Engineering Contradiction:

Improveanomaly detection capabilityVSAvoidsystem complexity

Core Design Contradiction:

Difficulty of detecting and measuringVSDevice complexity

Solution Approach 1:

The machine learning models are trained in advance on historical data patterns before deployment. This preliminary training phase enables the models to learn normal behavior patterns offline, reducing the computational burden during real-time operation and simplifying the deployed system's complexity.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system uses tracking beacons that create simplified representations of data movement events. These beacon signals serve as copies or proxies for complex data access patterns, enabling efficient monitoring without requiring analysis of the entire data set.

Inventive Principle:

Principle #26Copying

4Reliability

If tracking beacons are embedded in data sets, then continuous monitoring of data location and access is enabled, but the data set size and transmission overhead increase

Engineering Contradiction:

Improvedata tracking capabilityVSAvoiddata set size

Core Design Contradiction:

ReliabilityVSQuantity of substance

Solution Approach 1:

The tracking functionality is extracted as a separate, minimal beacon component embedded within the data set. This extraction allows monitoring capabilities to be added without requiring duplication of the entire data set, minimizing the increase in data size and transmission overhead.

Inventive Principle:

Principle #2Taking out (Extraction)

Data Source

PatentUS20260017368A1Artificial intelligence-based detection of anomalies in movement or accessing of electronic data post-data transmission

Publication Date: 2026.01.15 BANK OF AMERICA CORP

AI summary

Artificial-Intelligence (AI) in the form of Machine Leaning models are implemented to detect anomalies in data movement and/or access post-data transmission. Specifically, tracking beacons, which are configured to transmit signals periodically and/or when encountering a hop in a data route, are inserted in data sets prior to transmitting the data sets to a third-party entity. Tracking beacons may also be configured such that transmission of a signal from the beacon triggers acquisition and initiates communication of access logs. In response to receiving signals from the tracking beacons and/or access logs, ML models which have been trained to detect anomalies in data movement and/or access based on historical movement and/or access patterns of the same/similar data sets are executed to detect any such anomalies based, at least on information included within and/or derived from the tracking beacon signal and/or the access logs.