Mi4A-HomeData. MicroChip4Age dataset on domestic behaviours in a multi-user environment

Abstract The Mi4A-HomeData dataset contains multimodal sensor data collected in a real smart home environment to monitor daily activities of residents. The data were acquired between April 3 and April 16, 2024, from two individuals living in the instrumented environment . The sensing infrastructure includes motion (PIR), door/window open–close, temperature/humidity/pressure, smart plug (energy consumption), wearable activity bands (steps and battery), and Bluetooth Low Energy (BLE) anchor devices providing RSSI-based localization. Sensors are distributed across different rooms (bathroom, bedroom, kitchen, and living room) . The dataset is organized into two main JSON files: ground_truth_daily_behaviour.jsonContains labeled daily activities (sleeping, eating, showering, medication intake, leaving home, brushing teeth). Each record includes: user_id start_date end_date activity_type devices_data.jsonContains time-stamped sensor measurements, with structure depending on the device type: Motion and open/close sensors: timestamp state sensor_id Temperature and humidity sensors: timestamp temperature humidity pressure sensor_id Smart plug: timestamp current_a current_w total_voltage total_consumption today_consumption sensor_id BLE anchor devices (localization): timestamp rssi anchor_id user_id Wearable activity band: timestamp steps battery user_id All records are timestamped and linked to unique sensor or user identifiers, enabling temporal alignment and multimodal analysis. Acknowledgements This dataset has been partially supported by grant PDC2023-145863-I00 funded by MICIU/AEI/10.13039/501100011033, and by "European Union NextGenerationEU/PRTR".