TopoBrick: Agentic Topology Sampling of Exogenous Variables for Zero-Shot Building IoT Forecasting

Source

• Source name: arXiv
• Source kind: paper
• Authors / organization: Xiachong Lin, Du Yin, Arian Prabowo, Hao Xue, Wen Hu, Imran Razzak, Matthew Amos, Sam Behrens, Flora D. Salim
• Venue: arXiv: cs.AI
• Published: 2026-07-07
• Collected: 2026-07-08
• URL: http://arxiv.org/abs/2607.06349v1
• PDF: https://arxiv.org/pdf/2607.06349v1
• DOI: UNKNOWN

Plain-Language Brief

The paper introduces TopoBrick, a framework for zero-shot forecasting of Internet-of-Things (IoT) data in buildings, which does not require training and takes into account the physical topology, spatial hierarchy, and operational context. It uses an agentic topology sampler to select relevant exogenous variables for each target based on a building knowledge graph.

What the Document Says

• TopoBrick is a framework designed for zero-shot forecasting of Building IoT data.
• It does not require training and takes into account the physical topology, spatial hierarchy, and operational context.
• The framework uses building knowledge graphs to construct a structural skeleton.
• An agentic topology sampler is employed to select target-specific exogenous variables.
• The selected variables are organized based on their deployment-time availability.
• Past-known sensor states and future-known calendar, schedule, and meteorological exogenous variables are distinguished.
• TopoBrick performs better than strong zero-shot foundation model baselines and competes with fully trained building-specific models in three real-world buildings.
• Ablation studies show that topology-aware sampling is more reliable than random, ontology-only, or fixed-hop selection, particularly for physically coupled HVAC (Heating, Ventilation, and Air Conditioning) and weather-driven sensing variables.

Key Concepts Mentioned

• TopoBrick
• zero-shot forecasting
• Building IoT
• building knowledge graphs
• agentic topology sampler
• exogenous variables
• HVAC

Claims Made by the Document

• TopoBrick outperforms strong zero-shot foundation model baselines.
• TopoBrick remains competitive with fully trained building-specific models.

Evidence Used

• Across three real-world buildings, TopoBrick outperformed the baselines and was competitive with fully trained models.

Methods / Approach

UNKNOWN

Limitations Stated by the Authors

UNKNOWN

Limitations Noted by the Agent

Agent interpretation — not from the source.

UNKNOWN

Open Questions

UNKNOWN

Notes for Future Q&A

May include interpretation. Labeled where it does.

UNKNOWN

Raw Abstract / Excerpt

Building sensors are embedded in physical topology, spatial hierarchy, and operational context, yet existing forecasters often treat them as isolated time series or rely on fixed covariate sets. We present TopoBrick, a training-free framework for zero-shot building IoT (Internet-of-Things) forecasting. TopoBrick uses building knowledge graphs to construct a compact structural skeleton and employs an agentic topology sampler to select target-specific exogenous variables. The selected variables are organized by deployment-time availability, separating past-known sensor states from future-known calendar, schedule, and meteorological exogenous variables. Across three real-world buildings, TopoBrick outperforms strong zero-shot foundation-model baselines and remains competitive with fully trained building-specific models. Ablations show that topology-aware sampling is more reliable than random, ontology-only, or fixed-hop selection, especially for physically coupled HVAC and weather-driven sensing variables.