Variations in traffic count data

Abstract

Traffic counts are the most basic input into transport planning studies, yet there is the potential for errors to propagate within the data collection process that can result in uncertainty in the final collected data. If left unaccounted for, these errors have the potential to skew transport planning and traffic engineering decision making. Multi-million Rand infrastructure projects are born out of studies based on this data. This paper explores variations in traffic flow data observed from a variety of traffic count surveys. It considers measurement errors that are inherent in manual and automatic counts, and explores the impact of outliers and seasonality adjustments, as well as the concept of an average day in traffic data summaries. Local and international data sources have been applied in a case study to demonstrate the likely outcome of variations in analytical results when using erroneous traffic count data. The paper also draws conclusions about the potential implications to policy and decision-making and offers suggestions about the role that planning authorities in all spheres of government should play in controlling the quality of data used in transportation studies.