Estimating Traffic Volume in Developing Countries using Socioeconomic and Built Environment Factors

Abstract: Many developing countries experience a significant lack of up-to-date traffic volume data on their road networks due to the limited resources and funding for regular traffic monitoring and technological upgrades. This leads to multiple challenges in transportation planning, maintenance, and rehabilitation, assessing safety and environmental impacts, etc. that result in congestion, higher road crashes, inefficient maintenance, environmental pollution, and many more. This study focused on developing a regression model to predict the Average Daily Traffic (ADT) of roadways where timely updated ADT counts are unavailable. For the study, two A-class roads in Sri Lanka were selected with a total length of 317.78 km and 430.57 km respectively. These two roads traverse different administrative districts in the country which are different in land use, population density, economic activities, etc. Then, the selected roadways were segmented into uniform road sections in length based on the availability of ADT data. For each selected road section, the type of area, land use, vehicle population of the district, availability of public transport, major economic activities, population density, number of B-class roads connected, road width, number of lanes, and number of intersections variables were collected. The road sections where ADT counts were available were used to develop the multiple regression model with a 95% level of confidence (p < 0.95). Initially, the data set was divided into training and testing data and the testing data set was used to validate the model. Population density, number of intersections, and area type variables were found to be statistically significant in the model with an R-squared value of 0.784 indicating a reasonable relationship between the dependent and independent variables. This approach helps to estimate ADT on roads where transport planners experience a lack of timely updated traffic volume data and to make informed decisions on transport planning, policy-making, safety improvements, and regular maintenance and rehabilitation.