Project Overview
HARNode addresses the limitations of current HAR research tools: high cost, complex setup, and poor synchronisation, by providing a fully open hardware and software platform built from off-the-shelf components. Each node achieves ~1 ms time synchronisation via NTP and can stream data in real time to a central server. Designed for rapid field deployment, HARNode’s modular, 3D-printed housing with Velcro straps allows flexible body placement and supports over 100 nodes simultaneously. The system’s open-source design enables researchers to build, adapt, and extend it with minimal technical overhead, supporting evidence-based studies of sensor placement and activity recognition.
Our Goal
The goal of HARNode is to democratise large-scale, synchronised motion sensing for human activity recognition. By simplifying setup, reducing cost, and improving scalability, HARNode empowers researchers to perform fast, reproducible, multi-sensor field studies and to optimise sensor configurations for applications such as gait analysis, exoskeleton control, and wearable computing.
Paper
HARNode has been published as a scientific research paper at the OpenWearables Workshop 2025.
