ATM Anomaly Detection With Automated False-Positive Resolution

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing systems face challenges in efficiently detecting and resolving anomalies across a fleet of automated teller machines (ATMs), which are often time-consuming and resource-intensive.

Innovation Solution

A centralized computing platform with anomaly detection logic and a resolution model processes ATM transaction data to identify and resolve anomalies by preprocessing for false positives, applying detection rules, trending anomalies, and automatically executing corrective actions.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If manual detection and resolution methods are used for ATM anomalies, then human expertise can identify complex issues, but the process becomes time-consuming and resource-intensive

Engineering Contradiction:

Improveanomaly detection accuracyVSAvoidtime to resolve anomalies

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The system enables automated self-diagnosis and self-resolution of ATM anomalies through machine learning models that automatically detect, classify, and resolve issues without human intervention, eliminating the time loss associated with manual detection while maintaining high accuracy through intelligent algorithms

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

Manual human expertise is replaced with automated machine learning-based detection and resolution systems that process ATM data, identify anomalies, and execute resolutions programmatically, significantly reducing resolution time while maintaining or improving detection accuracy through consistent algorithmic application

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Reliability

If comprehensive anomaly detection is performed across the entire ATM fleet, then all anomalies can be identified, but the computational resources and processing time increase significantly

Engineering Contradiction:

Improveanomaly detection coverageVSAvoidcomputational resources

Core Design Contradiction:

ReliabilityVSUse of energy by moving object

Solution Approach 1:

The ATM fleet is segmented into groups or individual units that can be processed independently by the machine learning models, allowing parallel processing that reduces overall computational burden while maintaining comprehensive detection coverage across the entire fleet

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The system applies anomaly detection with targeted precision rather than exhaustive analysis of all possible parameters, using machine learning models to focus computational resources on the most relevant features and patterns that indicate anomalies, thereby reducing overall resource consumption while maintaining reliable detection

Inventive Principle:

Principle #16Partial or excessive action

3Measurement precision

If multiple anomaly detection rules are applied to improve accuracy, then detection precision increases, but the system complexity increases

Engineering Contradiction:

Improveanomaly identification accuracyVSAvoiddetection system complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

A single machine learning model is designed to perform multiple detection functions simultaneously, evaluating various anomaly patterns and rules within one unified framework, which maintains high detection accuracy while reducing system complexity compared to multiple separate detection systems

Inventive Principle:

Principle #6Universality (Multi-functionality)

Solution Approach 2:

Multiple anomaly detection rules and criteria are merged into an integrated machine learning evaluation process that combines various detection logic into a unified model, achieving comprehensive anomaly identification with reduced complexity through consolidated processing

Inventive Principle:

Principle #5Merging (Combining)

4Productivity

If automated resolution actions are executed without human review, then resolution speed increases, but the risk of incorrect actions increases

Engineering Contradiction:

Improveanomaly resolution speedVSAvoidresolution action accuracy

Core Design Contradiction:

ProductivityVSReliability

Solution Approach 1:

The system incorporates feedback mechanisms where machine learning models continuously learn from the outcomes of automated resolution actions, adjusting and refining their decision-making to improve accuracy over time, enabling reliable automated execution without human review while maintaining high resolution speed

Inventive Principle:

Principle #23Feedback

Solution Approach 2:

The machine learning models are trained in advance with extensive data and scenarios to pre-learn the correct resolution actions for various anomaly types, enabling them to execute accurate resolutions automatically without real-time human intervention, thus achieving both speed and reliability

Inventive Principle:

Principle #10Preliminary action

Data Source

PatentUS20250363874A1Automated Teller Machine (ATM) Anomaly Detection and Resolution

Publication Date: 2025.11.27 BANK OF AMERICA CORP

AI summary

Arrangements for detecting and resolving automated teller machine (ATM) anomalies are provided. A computing platform may receive information related to one or more automated teller machines (ATMs) that may include one or more anomalies. The computing platform may preprocess the information to remove one or more false positives from the information. The computing platform may apply anomaly detection logic to the preprocessed information to identify one or more anomalies. The computing platform may output one or more anomaly codes that correspond to the identified one or more anomalies. The computing platform may identify and subsequently execute one or more actions to resolve the one or more anomalies.