Building a road weather information network for integrating data from heterogeneous sources

The Road Weather Information System for Canada (RWISC) is being implemented to provide a more efficient, sustainable and safer highway system. This is achieved by integrating a Canada-wide network of environmental system sensors (ESS) and instrumented vehicles which provide a continuous stream of observation data describing road conditions. Provincial computer centres (PC/sup 2/) acquire data from fixed and mobile ESS stations in real-time. The road weather information network (RWIN) will acquire data from PC/sup 2/ using the Canadian meteorological markup language (CMML), an XML based schema. The RWIN provides a centralized repository of all the provincial ESS data, performs quality control (QC) services, sending realtime alerts when deterioration or outage of roadside components is detected, and will deliver QC'ed data in real-time to the province, and their designated recipients in CMML and other formats. RWIN is based on the MSC's data management framework (DMF). The DMF provides integrated access to MSC real-time and archive datasets through an enterprise data interface (EDI). The RWIN component based architecture is scalable and is capable of handling text, model, image, raster and GIS data. Interactive and batch-mode delivery of spatial information is supported while the component architecture provides the flexibility of adding new applications or data sources. This paper describes the architecture for RWIN and the development and integration approaches through collaboration between multiple levels of government and the private sector. The proof-of-concept and prototyping approaches for validating requirements and effectively engaging stakeholders will be highlighted. Specific approaches towards testing and integration, data delivery, security, and protocols for return of RWISC data to contributing stakeholders are discussed. The paper concludes by outlining future phases of the RWIN project, implementation challenges and lessons learned.