BMS Autoconfiguration for ML-Based Equipment Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing Building Management Systems (BMS) face inefficiencies in data aggregation and integration of equipment due to manual onboarding processes that require repeated identification of spatial hierarchies, leading to misclassification and operational issues.
Innovation Solution
A system architecture that utilizes machine learning models to automatically discover and classify equipment within a BMS, integrating them without repeated spatial hierarchy identification, and includes a user interface that restricts operations to specific levels of the spatial hierarchy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual onboarding processes are used to integrate equipment into BMS, then equipment can be integrated with spatial hierarchy information, but repeated identification of spatial hierarchies is required leading to time delays and misclassification
Solution Approach 1:
The system performs preliminary actions by automatically discovering equipment and pre-classifying it into spatial hierarchies using machine learning models before formal integration is required. This eliminates the need for repeated manual identification of spatial hierarchies during subsequent integration operations, reducing both time delays and classification errors.
Solution Approach 2:
The equipment integration process becomes self-service through automated discovery mechanisms that independently identify and classify equipment without requiring manual intervention. The system autonomously performs spatial hierarchy identification and equipment classification, eliminating repetitive manual operations and reducing integration time while maintaining accuracy.
2Loss of information
If manual onboarding processes are used for equipment integration, then spatial hierarchy information can be captured, but repeated identification requirements increase operational complexity
Solution Approach 1:
The manual mechanical process of repeatedly identifying and inputting spatial hierarchy information is replaced with an automated electronic discovery and classification system. Machine learning models automatically capture and retain spatial hierarchy information, eliminating the need for manual data entry operations and reducing process complexity while preventing information loss.
Solution Approach 2:
Machine learning models serve as intermediaries between equipment discovery and spatial hierarchy integration. These models automatically process equipment data, classify equipment into appropriate spatial hierarchies, and retain the information without requiring manual intervention, thereby simplifying the onboarding process while ensuring information is captured and retained.
3Productivity
If automated machine learning classification is used for equipment discovery, then integration speed is improved, but system complexity increases
Solution Approach 1:
The machine learning classification system serves multiple functions simultaneously: it discovers equipment, classifies equipment into spatial hierarchies, validates classifications, and integrates equipment into the BMS. This multi-functionality consolidates what would otherwise require separate systems into a single unified platform, improving integration speed while managing overall system complexity through functional consolidation.
Data Source
AI summary
A Building Management System (BMS) can include one or more memory devices storing instructions thereon that, when executed by one or more processors, cause the one or more processors to receive a selection of a building, perform a search on a network of the building for a plurality of pieces of equipment or a plurality of datapoints associated with the plurality of pieces of equipment, discover at least one piece of equipment of the plurality of pieces of equipment or one or more datapoints of the plurality of datapoints, update the user interface to display a graphical representation of the at least one piece of equipment or the one or more datapoints, and prompt a machine learning model to generate at least one classification for the at least one piece of equipment or the one or more datapoints.