Measures of Pavement Performance Must Consider the Road User

In 1960, Bill Carey and Paul Irick developed the Present Serviceability Index (PSI) as a user-based performance measure to define pavement quality and failure at the AASHO Road Test, a controlled load experiment that cost $300 million in 2014 dollars (1). The Canadians used the same approach in creating their Riding Comfort Index, but on a 0 to 10 scale. The PSI method was adopted and used worldwide to define pavement quality until the early 1990's when FHWA arbitrarily adopted International Roughness Index (IRI). It was intended as a measure of quality for HPMS (Highway Performance Monitoring System) data because IRI was touted to be standard and universal by the World Bank. PSI is still used by many agencies around the world but most state DOTs felt forced to follow FHWA and adopt IRI. The IRI is not standard state-to-state and more importantly the levels of "acceptability" and "failure," which must be set to define performance, vary from state-to-state. The US Federal MAP-21 requires state DOTs to do broader "performance" management and develop acceptable pavement performance measures (2). PSI is tied to road user response but IRI is not. This paper examines these indexes and how they derived. It contends that PSI can serve all levels of need while IRI does not, because it is not understood by highway users and legislators. PSI reflects human rider response and IRI does not close that gap.