InsightBuild: LLM-Powered Causal Reasoning in Smart Building Systems
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2025
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866913937982750720 |
|---|---|
| author | Neogi, Pinaki Prasad Guha Mohammadshirazi, Ahmad Ramnath, Rajiv |
| author_facet | Neogi, Pinaki Prasad Guha Mohammadshirazi, Ahmad Ramnath, Rajiv |
| contents | Smart buildings generate vast streams of sensor and control data, but facility managers often lack clear explanations for anomalous energy usage. We propose InsightBuild, a two-stage framework that integrates causality analysis with a fine-tuned large language model (LLM) to provide human-readable, causal explanations of energy consumption patterns. First, a lightweight causal inference module applies Granger causality tests and structural causal discovery on building telemetry (e.g., temperature, HVAC settings, occupancy) drawn from Google Smart Buildings and Berkeley Office datasets. Next, an LLM, fine-tuned on aligned pairs of sensor-level causes and textual explanations, receives as input the detected causal relations and generates concise, actionable explanations. We evaluate InsightBuild on two real-world datasets (Google: 2017-2022; Berkeley: 2018-2020), using expert-annotated ground-truth causes for a held-out set of anomalies. Our results demonstrate that combining explicit causal discovery with LLM-based natural language generation yields clear, precise explanations that assist facility managers in diagnosing and mitigating energy inefficiencies. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2507_08235 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | InsightBuild: LLM-Powered Causal Reasoning in Smart Building Systems Neogi, Pinaki Prasad Guha Mohammadshirazi, Ahmad Ramnath, Rajiv Machine Learning Artificial Intelligence Smart buildings generate vast streams of sensor and control data, but facility managers often lack clear explanations for anomalous energy usage. We propose InsightBuild, a two-stage framework that integrates causality analysis with a fine-tuned large language model (LLM) to provide human-readable, causal explanations of energy consumption patterns. First, a lightweight causal inference module applies Granger causality tests and structural causal discovery on building telemetry (e.g., temperature, HVAC settings, occupancy) drawn from Google Smart Buildings and Berkeley Office datasets. Next, an LLM, fine-tuned on aligned pairs of sensor-level causes and textual explanations, receives as input the detected causal relations and generates concise, actionable explanations. We evaluate InsightBuild on two real-world datasets (Google: 2017-2022; Berkeley: 2018-2020), using expert-annotated ground-truth causes for a held-out set of anomalies. Our results demonstrate that combining explicit causal discovery with LLM-based natural language generation yields clear, precise explanations that assist facility managers in diagnosing and mitigating energy inefficiencies. |
| title | InsightBuild: LLM-Powered Causal Reasoning in Smart Building Systems |
| topic | Machine Learning Artificial Intelligence |
| url | https://arxiv.org/abs/2507.08235 |