Brick-DICL: Dynamic In-Context Learning for Automated Brick Schema Classification

2026-06-16

AI Summary

This work addresses the challenges in Brick ontology classification within building management systems (BMS), where vendor heterogeneity and inconsistent metadata lead to a proliferation of classes, insufficient domain knowledge in large language models (LLMs), and high costs of manual validation. To tackle these issues, we propose Brick-DICL, a two-stage dynamic in-context learning framework that introduces dynamic context learning to Brick classification for the first time. Our approach leverages metadata-RAG to enrich domain knowledge and class-RAG to narrow the candidate space, while integrating multi-LLM collaborative prediction with a confidence-based filtering mechanism to automatically flag low-confidence results for human review. Experiments on multiple real-world building datasets demonstrate that Brick-DICL significantly outperforms existing methods, achieving high-accuracy, vendor-agnostic, and format-agnostic automated classification, thereby advancing standardization and interoperability in building systems.

Abstract

Building Management Systems (BMS) are essential for optimizing energy efficiency and operational performance in modern buildings. However, the lack of standardization across BMS points from different manufacturers creates significant barriers to integration and data utilization. While the Brick schema offers a standardized ontology for building systems, mapping BMS points to appropriate Brick classes presents three critical challenges: (i) the extensive number of Brick classes (936 in the latest version), (ii) limited domain-specific knowledge in large language models (LLMs), and (iii) substantial manual effort required for verification. To address these challenges, we propose Brick-DICL, a two-stage dynamic in-context learning framework for automated Brick schema classification. Brick-DICL consists of two primary components: metadata-RAG, which retrieves relevant examples to enhance LLMs' domain knowledge, and class-RAG, which narrows down potential Brick classes to address the large classification space. Additionally, we implement a multi-LLM filtering mechanism that compares predictions across multiple models, flagging low-confidence classifications for human review. As a result: (i) General: Brick-DICL is applicable to any building management system regardless of manufacturer or metadata format; (ii) Novel and Powerful: as the first dynamic in-context learning approach for Brick schema classification, Brick-DICL achieves significant classification accuracy improvements on building datasets, outperforming existing methods; (iii) Efficient: our multi-LLM filtering strategy reduces manual verification effort, enabling rapid digital building onboarding. Extensive experiments demonstrate Brick-DICL's effectiveness across diverse building datasets, accelerating the path toward standardized, interoperable building management systems.

Problem

• Brick schema
• Building Management Systems
• schema classification
• ontology mapping
• standardization

Innovation

• Dynamic In-Context Learning
• Brick Schema Classification
• RAG
• Multi-LLM Filtering
• Building Management Systems