VR-IoT-DigitalTwin-AnomalyDetection-Reproducibility-Package

This dataset provides the complete reproducibility package supporting the experimental results of the study. The package contains curated residential IoT sensor streams, ground-truth fault event annotations derived from controlled fault injection, anomaly event logs, statistical summaries, and evaluation artifacts used to assess unsupervised anomaly detection performance under a strict time-ordered (leak-free) protocol. In addition, the repository includes documentation describing the temporal data split strategy, event-level evaluation methodology, annotation rules, and spatial mapping between sensors and rooms within a VR-based digital twin. Algorithmic pseudocode and configuration files are provided to enable independent understanding and methodological replication without exposing proprietary source code or identifiable residential information. Raw sensor data and VR assets are partially anonymized to preserve participant privacy while maintaining scientific validity. The dataset is intended for reproducibility verification, methodological transparency, and future comparative research in spatial anomaly detection, digital twins, and human-centered diagnostic systems for smart residential environments.