Evaluation of Roadside Surveys for Freight Transport Data at Border Crossings

摘要

Roadside freight surveys are one of the main types of survey used to observe road freight transport. These surveys require many staff, resulting in high costs, limited survey periods, and variable sampling rates. As a result, data may be incomplete and/or unreliable. This paper presents an examination of roadside survey data of freight transport at a border crossing in Thailand to determine the best day in the week for gathering data and the appropriate sampling rate. The data were collected at the border between Thailand and Laos at Friendship Bridge Number Three in the north-eastern past of Thailand using the face-to-face interview method for all trucks at the outbound border of Thailand during open hours of the border for four weeks. The objective of this analysis is to determine the best day for survey data and the optimum sampling rate. The results indicate that Wednesday?s survey data can represent all data better than the other days of the week, both in terms of the number of trucks passing the survey point and the captured weight of freight. Of the remaining days, the best representative day is Thursday. A low sampling rate (20%) has the ability to provide an estimated total weight of freight passing through the survey point which comes close to the actual weight. However, the estimated freight weight is an aggregate from many origins and thus cannot provide the details of transport data. The freight weight classified by freight origin analysis revealed that many origins were disappearing from the captured data of low sampling rate. To solve this problem, the survey needs a high sampling rate for low volume traffic, perhaps as high as 90%.