About the Project
Wearable devices are rapidly transforming healthcare by enabling continuous, real-time monitoring of patients outside clinical settings. MEDWEAR is a community-driven initiative to create open standards for structuring and exchanging high-frequency data from medical wearables.
We aim to address the fragmentation of the digital health ecosystem by building upon existing standards—such as Open mHealth, HL7 FHIR, IEEE 11073, and openEHR—to define lightweight, extensible schemas suitable for modern sensing devices and middleware systems like ROS 2 and MQTT.
Why Standardization Matters
Although wearable technologies have become integral to modern healthcare—supporting use cases from chronic disease management to preventive care—the lack of common data formats hinders integration with EHRs, AI systems, and clinical decision tools. Over 80% of healthcare data remains unstructured and difficult to reuse.
Standardizing wearable data supports interoperability, clinical utility, patient safety, and research reproducibility. MEDWEAR provides schemas and APIs that support high-frequency data, real-time streaming, and device-agnostic modeling to facilitate broad adoption and integration.
Collaborating Entities
- ETH Zurich - Open Research Data Initiative
- SCAI Lab, ETH Zurich
- DART Lab, Lake Lucerne Institute
- UZH
- SUPSI
Challenges in Wearable Health Data
- Fragmented systems with incompatible formats
- Lack of EHR integration
- Data privacy and regulatory constraints
- Low patient engagement and digital divide
- Limited clinical adoption and ROI evidence
- Barriers to real-time, high-frequency data streaming
Comparative Overview of Standards
We evaluate four key initiatives that inspire the MEDWEAR schema design:
- Open mHealth: Simple, modular JSON schemas for structured, contextual data.
- HL7 FHIR: Granular RESTful resources widely used in clinical settings.
- IEEE 11073-PHD: Device-specific standards for personal health devices, well-suited for discrete measures.
- openEHR: Robust clinical models with semantic precision.