Origin Destination Matrix Estimation for Increasing Traffic Survey Efficiency

Abstract

In pavement management, traffic data is important. Time series-based traffic projection method, which is used widely by the Indonesian Road Agency, requires a lot of data. Consequently for the purpose of network-wide pavement management system, traffic data to be collected are in each particular link roads in every single years. These activities require large of resources. However, time-series projection method are not sensitives to the changes of physical and operational road network and the change in land use pattern. It is important therefore to improve the traffic projection method. One method that is sensitive to changes in the land use area and road network is a four-stage transportation modeling, where data requirements are a) road network characteristics, b) land use pattern and c) Origin Destination Matrix (ODM). The advantage of this method is that the results are more accurate and the amount of data required is less. The main problem is in providing data ODM for the base year. This paper attempts to determine the optimal number of traffic survey locations, using a case study of East Java Road Network, Indonesia. The results show that reducing the number of traffic data points can be done while improving the accuracy of traffic projections.