Statistiek en verkeer*

Abstract

SummaryStatistics and traffic.Traffic defined as the collectivity of moving (andparked) vehicles, is the direct consequence of transportation needs as they occur in society as a result of economic and social activities. Hindrances to the free flow of traffic result in decrease and even stagnance of human activities. The correct functioning of traffic facilities is imperative to modern human society.To solve traffic engineering problems, present and future traffic loads must be determined and traffic behaviour in relation to geometric road design must be known. As the volume and character of human activity fluctuates and changes in time, traffic loads will vary in volume and character. Any parameter used to indicate traffic load (volume, speed, etc.) has no well determined value but rather should be evaluated by describing its frequency distribution (histogram). The traffic engineer should not design for the average value, because the facility would not render the expected services whenever a value higher than average would occur. Nor should he design for the highest value, because then the facility will be underloaded, and therefore uneconomical, except when this exceptionally high value occurs. This is illustrated in case of determining the “design speed” for a curve in a road and of determining the “design‐volume” for a road‐crosssection.To determine the histograms of the various design‐factors numerous observations of actual traffic are required. The survey programs should be well correlated with the data which are required for designwork to limit scope and cost. Often, sampling will render data of sufficient accuracy; the required accuracy being set by the design engineer. The statistician will render essential services in determining the size of samples, the survey programs and supervising the correct processing of data.The statistical problems involved in traffic‐behaviour research work are even more challenging because many of the factors involved are not known and should be detected through large series of observations.