Quantitative characterization of the vehicle motion environment (VME)

Abstract

A concept is presented for creating a measurement system that can quantify the specific motions which vehicles exhibit as they move in traffic, under the full array of traffic operations. Such quantification is seen as crucial to the development of automatic collision prevention systems and has spinoff utility for the study of many other issues in human factors and vehicle and highway engineering. This study has addressed the experimental and analytical challenges involved in wide-area sensing, large-volume data processing, and both deterministic and statistical analyses of the data which will characterize this so-called, "Vehicle Motion Environment" (VME). The basic concept which appears to be feasible for such measurements involves a remote sensor which is installed at the roadside, probably on a tall pole, and which produces electro-optic images of the traffic stream and converts them into a permanent data file of the quantified trajectory for each motor vehicle passing through the field of view. The paper covers the performance specifications for the VME measurement system plus considerations for the measurement package and the subsequent processing needed for deriving the variables of interest. Various applications of the VME system are also addressed.