The MEDWEAR project addresses the lack of interoperability in wearable medical device data used in clinical and rehabilitation research. Currently, wearable manufacturers and research groups rely on heterogeneous data formats, proprietary APIs, and incompatible storage systems, making reproducibility, large-scale integration, and longitudinal analysis difficult. To address this challenge, MEDWEAR developed an Open Research Data (ORD) prototype composed of several interoperable modules:

• A wearable sensor data standard based on extensible JSON schemas.
• Equivalent ROS 2 message definitions for robotics and real-time applications.
• A software platform for multimodal wearable data collection and visualization.
• A streaming and conversion framework for online and offline interoperability.
• A REST API and frontend infrastructure for clinical and research integration.
• Proof of concept in two use cases: a data collection study and a real-time application.

The MEDWEAR defines a set of schemas extending existing standards including Open mHealth, IEEE 11073, and ROS 2 healthcare interfaces to support raw biosignal acquisition and derived digital biomarkers. The project further enables interoperability between wearable devices, robotic systems, clinical databases, and analytics pipelines. Community engagement activities were conducted through the ETH Rehabilitation Science Center (RESC), the Swiss NeuroRehab Initiative, rehabilitation clinics, and partner universities. These

activities included workshops, symposiums, and dissemination events focused on FAIR data practices and wearable data standardization. The resulting infrastructure provides a foundation for FAIR-compliant, reusable, and interoperable wearable data collection across rehabilitation and digital health research environments. MEDWEAR establishes the technical basis for future large-scale clinical integration, reproducible multimodal research pipelines, and long-term ORD infrastructure services within the ETH Domain.