Building Alarm Filtering With LLM Agents for False Alarm Mitigation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Building management systems often generate false alarms, diverting operator attention and resources away from actual issues, and existing methods lack effective automated means to distinguish between false and true alarms.
Innovation Solution
Implementing a Generative Artificial Intelligence (AI) Large Language Model-based system with primary agents and subagents trained in specific domains to analyze alarms, generate plausible causes, and classify them as false or true alarms based on domain knowledge and collaborative reasoning.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If Building Management Systems issue alarms for detected abnormalities, then operator awareness of potential issues is improved, but operator time and attention are wasted on false alarms
Solution Approach 1:
An AI-based intermediary system is introduced between the alarm generation source and the operator. This intermediary automatically analyzes alarm data, determines whether alarms are false or true, and only presents confirmed true alarms to operators, thereby filtering out false alarms before they consume operator time while maintaining reliable alarm detection
Solution Approach 2:
The system enables self-service by implementing automated alarm analysis and classification capabilities that operate without human intervention. The AI system independently evaluates alarm data, cross-references sensor information, and autonomously determines alarm validity, eliminating the need for operators to manually investigate each alarm
2Reliability
If operators respond to all alarms, then potential issues are addressed, but operator attention is diverted from actual true alarms due to false alarms
Solution Approach 1:
The AI system acts as an intermediary that processes and filters alarm information before reaching operators. It maintains reliability by thoroughly analyzing all alarms through multiple data sources and sensor cross-referencing, while simultaneously improving ease of operation by presenting operators with a streamlined list of only confirmed true alarms, thereby simplifying their workflow
Solution Approach 2:
The system performs preliminary analysis and classification of alarms before they reach operators. By pre-evaluating alarm validity using AI algorithms and sensor data cross-referencing, the system prepares filtered, high-confidence alarm lists in advance, allowing operators to focus immediately on true alarms without preliminary investigation
Data Source
Figure 1
Figure 2A
Figure 2B
AI summary
An alarm is raised. One of a plurality of Generative Artificial Intelligence (AI) Large Language Model-based autonomous primary agents is activated based at least in part on an alarm type of the alarm and performs an initial analysis of the alarm and creates plausible causes for the alarm. The corresponding Generative AI Large Language Model-based autonomous primary agent autonomously assigns each of the plausible causes to one or more Generative AI Large Language Model-based autonomous subagents that perform an analysis of the assigned plausible cause and returns a result back to the Generative AI Large Language Model-based autonomous primary agent, which classifies the alarm as a false alarm or a true alarm.