TEHI-2026 Research Artifacts: Experimental Code, Models, Cross-Hardware Data, and Prototype Materials

<p dir="ltr">This research artifact accompanies the study titled “An Edge-AI Enabled Multi-Parameter Remote Patient Monitoring System with Subject-Independent Wearable Fall Detection.” The package contains the experimental notebooks, source code, trained machine-learning models, converted SisFall dataset, cross-hardware MPU6050 data, data-collection materials, prototype images and video, and embedded deployment investigation files.

The artifacts support the ECG beat classification and wearable fall-detection experiments, including participant-independent evaluation, causal inference, model quantization, cross-hardware testing, and ESP32-based implementation. The package is provided to support research reproducibility, experimental inspection, and further development of Edge AI, TinyML, wearable sensing, and remote patient monitoring systems. The deployment investigation also documents the memory limitations encountered when attempting complete Transformer inference on the classic ESP32 platform.</p>