Alarm Spuriousness Detection Using Machine Learning Vectors
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current alarm systems in industrial facilities, such as oil and gas fields, face challenges in accurately distinguishing between genuine and spurious alarms due to environmental factors, leading to unnecessary safety protocols, increased operational costs, and personnel fatigue, as existing techniques lack the necessary intelligence to effectively detect spurious alarms.
Innovation Solution
A method and system utilizing a machine learning model to determine the spuriosity of alarms by generating input vectors from sensor and maintenance data, including environmental and trigger parameters, to calculate a spuriosity index, thereby distinguishing between true and false alarms.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If alarm systems use sensors to detect hazardous substances, then detection capability is improved, but false alarms increase due to environmental factors
Solution Approach 1:
The patent introduces an intermediary system consisting of a processor and machine learning model that sits between the sensor detection layer and the alarm generation layer. This intermediary analyzes multiple input vectors including environmental parameters, sensor data, and maintenance history to determine spuriosity indices, thereby filtering out false alarms while preserving true detection capability
Solution Approach 2:
The patent changes the parameters used for alarm determination from simple sensor threshold values to a comprehensive set of parameters including environmental conditions (temperature, humidity, pressure), sensor performance metrics, and maintenance history. This multi-parameter approach allows the system to distinguish between genuine hazards and environmental interference
2Reliability
If spurious alarms trigger safety protocols, then safety is improved, but operational costs increase
Solution Approach 1:
The system performs preliminary analysis by calculating spuriosity indices for all alarm conditions before activating safety protocols. The machine learning model evaluates multiple input vectors in advance to predict whether an alarm is spurious, allowing the system to prevent unnecessary safety protocol activation while maintaining readiness for genuine hazards
3Reliability
If frequency of spurious alarms increases, then alarm system sensitivity is improved, but personnel fatigue increases
Solution Approach 1:
The alarm system performs self-service by automatically analyzing its own performance through the machine learning model that processes maintenance history, sensor data, and environmental parameters. The system self-evaluates spuriosity indices and makes autonomous decisions about alarm validity, reducing the burden on personnel while maintaining high sensitivity
Data Source
AI summary
This disclosure relates to method and system for determining probability of an alarm generated by an alarm system. The method may include receiving sensor data and maintenance data. The sensor data may include one or more environmental parameters and one or more trigger parameters, and the alarm is generated based on the one or more trigger parameters. The method may further include generating one or more input vectors based on the sensor data and the maintenance data, and determining a spuriosity index of the alarm based on the one or more input vectors using a machine learning model. The machine learning model may be created using historical sensor data and historical maintenance data, and the spuriosity index is indicative of the probability of the alarm.