Building Equipment Fault Prediction Using ML Time Windows
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Building management systems often fail to detect equipment faults in a timely manner, leading to prolonged discomfort and increased energy costs, as they typically identify issues only after they have significantly impacted the building environment.
Innovation Solution
A method using machine learning models to predict faults in building equipment by analyzing historical measurement data, generating confidence scores, and performing automated actions to prevent or mitigate faults before they occur, including identifying the root cause of predicted faults.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional building management systems monitor equipment until faults occur, then equipment faults are detected, but detection happens too late causing prolonged discomfort and increased energy costs
Solution Approach 1:
The system performs preliminary actions by training machine learning models on historical equipment data to predict future faults before they occur. The model analyzes patterns in operational parameters and generates predictions about potential equipment failures, enabling maintenance teams to take preventive action before actual faults develop, thus resolving the contradiction between detection accuracy and time loss.
2Reliability
If building equipment operates at capacity to compensate for breakdowns, then building comfort is maintained, but energy consumption increases
Solution Approach 1:
The system predicts equipment faults before they occur, allowing maintenance to be scheduled proactively. This prevents unexpected breakdowns that would force other equipment to operate at maximum capacity, thereby maintaining building comfort reliability while avoiding the excessive energy consumption associated with compensatory operation of remaining equipment.
3Difficulty of detecting and measuring
If traditional systems analyze potential causes after temperature increase, then root cause is identified, but substantial time has already passed
Solution Approach 1:
The machine learning model performs preliminary analysis of equipment operational patterns to predict faults before they manifest as observable problems like temperature increases. By identifying at-risk equipment in advance, the system enables proactive maintenance actions that resolve issues before they impact building comfort, eliminating the time loss associated with post-fault analysis and response.
4Reliability
If machine learning models predict faults early, then preventive maintenance can be performed, but system complexity increases
Solution Approach 1:
The patent introduces machine learning models as intermediary components that bridge existing building management systems and predictive maintenance capabilities. These models process historical equipment data and generate fault predictions without requiring complete system redesign, thus improving equipment operation reliability while limiting the increase in system complexity through modular integration.
Data Source
AI summary
A method for predicting time periods in which faults are likely to occur for a piece of building equipment. The method includes receiving a plurality of measurements for one or more points that are associated with a piece of building equipment, the plurality of measurements measured during a first time period; executing a machine learning model using the plurality of measurements as an input to generate fault data for a plurality of time periods subsequent to the first time period; selecting a second time period from the plurality of time periods responsive to an assessment of the fault data for the plurality of time periods indicating a fault will likely occur in the piece of building equipment during the second time period of the plurality of time periods; and performing an automated action responsive to the selection of the second time period.