Recommendations for a statewide expansion of Ohio's RWIS network were made based on an extensive literature review, product review, cumulative cost comparison, survey of users and administrators, and site visits in Ohio, other states, and Switzerland. Factors used to determine optimal RWIS station deployment included distance between existing RWIS stations in Ohio and surrounding states, declared snow days, and annual snowfall amounts in Ohio. Also recommended is the use of File Transfer Protocol (FTP) to work around proprietary system constraints, and having data from all vendors put into a common database format. Although cost savings and benefits attributed to RWIS and anti-icing in the literature vary widely, there is a consensus that these systems do repay their costs in reduced maintenance costs and reduced accident rates. At a minimum, in addition to the RWIS stations already deployed in the Toledo, Columbus, and Cleveland areas, placement of 14 RWIS stations around the state should be sufficient to meet basic statewide weather prediction and monitoring needs. It may be advantageous for ODOT to deploy more RWIS stations to achieve a denser coverage encompassing all 88 counties in the state. Certain trouble spots, such as bridges, may warrant additional RWIS stations on a case-by-case basis, a concern not addressed in this report. At a minimum, each district should have at least one RWIS station, or two in northern districts. Exact sites should be chosen by district managers or personnel in consultation with meteorology and RWIS experts, subject to approval higher up, to give the district personnel that will use the information an additional incentive to use the RWIS. Installations should be at locations with typical rather than extreme conditions. It is further recommended that each county garage be equipped with a small weather station and simple pavement sensor near the station for the purpose of creating a weather conscious culture at the county garage level. The characteristics of an ideal RWIS station are described; these include pavement temperature sensors, atmospheric weather sensors, accurate forecasting, and traffic volume and speed information. Since the performance of pavement sensors under controlled conditions was not known, a bench test was proposed; results are reported separately in Part II. The benefit of the RWIS network will come only if anti-icing methods are diligently applied using RWIS information. Performance measures for winter operations should be built into the maintenance system.