Causal Relationship Models for Building Fault Root Cause Identification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional building management systems (BMS) face challenges in efficiently identifying and addressing the root cause of faults, often reporting downstream effects rather than the source, due to the complexity of managing disparate subsystems and protocols.
Innovation Solution
A computerized method and system that utilizes a causal relationship model to traverse and determine the root cause of faults by identifying causally-related building objects, retrieving historical data, and analyzing time data to pinpoint the earliest abnormality, thereby excluding inconsistencies and providing transparent management across different building subsystems.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional BMS report fault indications from downstream effects, then the system can detect faults, but the root cause identification is delayed and inefficient
Solution Approach 1:
The system pre-establishes causal relationship models that map potential root causes to their downstream effects before faults actually occur. When a fault is detected, the system can immediately query the pre-built model to identify the root cause, eliminating the need for time-consuming analysis of system logs and relationships.
Solution Approach 2:
The causal relationship model acts as an intermediary layer between fault detection and root cause identification. Instead of directly analyzing complex system relationships when a fault occurs, the system uses the pre-computed causal model to translate downstream effect observations into root cause identifications, significantly reducing diagnostic time.
2Adaptability or versatility
If BMS manages multiple disparate subsystems and protocols, then the system can monitor comprehensive building functions, but the complexity of identifying fault causes increases
Solution Approach 1:
The causal relationship model serves as a universal framework that can represent causal relationships across multiple disparate subsystems and protocols (HVAC, lighting, security, etc.). By using a unified model structure that accommodates different device types and communication protocols, the system simplifies fault analysis complexity while maintaining comprehensive subsystem integration.
Solution Approach 2:
The causal relationship model acts as an intermediary abstraction layer that translates complex, protocol-specific device relationships into a unified causal framework. This allows the system to manage multiple disparate subsystems without proportionally increasing fault analysis complexity, as the model handles the heterogeneity translation.
3Reliability
If BMS reports downstream effects rather than root causes, then the system can provide fault notifications, but the effectiveness of fault resolution is reduced
Solution Approach 1:
The system pre-computes and stores causal relationship models that link downstream effects to their root causes. When faults are detected and notified, the same pre-built models are used to immediately identify and report the root causes alongside the notifications, enabling maintenance personnel to address the actual problem source rather than just symptoms.
Solution Approach 2:
The system implements a feedback mechanism where fault notifications are enhanced with root cause information derived from causal relationship models. This feedback loop provides maintenance personnel with both the symptom (downstream effect) and the underlying cause, significantly improving fault resolution efficiency while maintaining reliable fault notification delivery.
Data Source
AI summary
A computerized method for evaluating and reporting a cause of a performance change in a building management system is shown and described. The method includes receiving an indication of a fault for building equipment of the building management system and determining a root cause for the fault by traversing a causal relationship model including the building equipment and other devices of the building management system.