InsightBuild: LLM-Powered Causal Reasoning in Smart Building Systems

Pinaki Prasad Guha Neogi, Ahmad Mohammadshirazi, Rajiv Ramnath

Keywords: Large Language Models, Fine-tuning, Smart Buildings, Energy Systems, Time-Series Forecasting, Interpretable AI

Abstract: 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 \& mitigating energy inefficiencies.

Submission Number: 17