Freight Data Mining Strategy Using Socio-economic Variables for Metropolitan Planning

Infrastructure investment decisions consider future infrastructure demand projections from freight models, the quality of which depends on fidelity of input freight data. The Freight Analysis Framework Version 2.2 (FAF^2.2) being a primary source of freight data for infrastructure planning provides commodity origin-destination flows for 114 zones within the USA. Freight disaggregation approaches using demographic and economic variables can be used to obtain county-level freight distribution. Freight distributions at both the levels: federal and county are insufficient for incorporating the effect of freight-related traffic on metropolitan-level transportation infrastructure. This paper describes a clusters-based freight data mining strategy using socio-economic variables to aggregate the most granular representations of freight flows called Traffic Analysis Zones (TAZs) in the Mobile Metropolitan Area (MMA). Such aggregation results in an intermediate level of freight distribution between the county and traffic zone levels, at a resolution meaningful for metropolitan-level planning.