Community-Driven Sensor Metadata Ecosystem for Exposure Health

PROJECT SUMMARY ABSTRACT Environmental exposures heavily influence human health. To gain more precise insight into environmental exposures' effects on health, scientists must integrate and assimilate sensor data and other types of research data across domains and varied temporal and spatial scales. However, we lack informatics tools that enable researchers to share, find, assess, and re-use environmental data sources. The central problem is the lack of associated metadata that enables appropriate use and re-use. There is a lack of consensus on metadata standards and conceptual models representing the exposure domain. Investigators and curators of real-world data collections are often not incentivized or responsible for metadata discovery for data re-use. Moreover, we lack accessible tools that enable scientists to discover, harmonize, store, and publish metadata. In prior work, we developed the Exposure Health Informatics Ecosystem (EHIE), a scalable, standards-based, open-source informatics infrastructure. EHIE supports semantically consistent, metadata-driven, event-based management of exposure data for research. Here, we propose to leverage our expertise in metadata and event-based modeling to advance methods for exposure health research entailing sensors. In aim 1, we will develop a logical model for representing diverse types of sensor metadata, and in doing so, discover and harmonize sensor and device metadata in a community-driven approach. In aim 2, we design and evaluate a user-facing metadata repository that will encompass a library of data, sensors/ devices, and interfaces that support metadata submission and browsing. In aim 3, we will develop specifications, logical models, and the functions needed to transform any exposure health data into an event-based format, enhanced with spatial and temporal coordinates. In aim 4, we will develop prototype workflows that research teams can use as models for using the tools we create to generated metadata-enhanced, semantically consistent exposure health data. Across these four aims, we engage the environmental exposure health research community through an expert panel, user-centered design, public comment, and a communications plan. This effort builds upon prior research and development efforts of an experienced informatics team in conjunction with the University of Utah's Center of Excellence in Exposure Health Informatics. The expected outcome is the development of community-driven, shared, and generalizable tools for metadata and data management, supporting reproducible environmental exposure health research.