Article ID: | iaor2013695 |
Volume: | 26 |
Issue: | 1 |
Start Page Number: | 99 |
End Page Number: | 124 |
Publication Date: | Jan 2013 |
Journal: | Transportation Research Part C |
Authors: | Bottero M, Dalla Chiara B, Deflorio F P |
Keywords: | statistics: inference |
A wireless sensor network (WSN) is a net of small sensor nodes, communicating using wireless technology to collect data. It combines distributed sensing and wireless communication, integrated in a self‐powered small device with limited computation and memory functions. In this research, a WSN for traffic monitoring was installed and tested in the area of a logistic platform, the freight village of Turin. The sensor network layout was designed to detect all vehicles entering and leaving the area as well as the zones to which they are relating. One peculiarity of the logistic centre installation is related to the sensors’ locations on the roadway: as it was not possible to install sensors in optimal locations, characterised by almost‐constant vehicle speed, scarce lane changing, and stationary vehicle pattern, the detection system accuracy requires assessment using vehicle count and classification. After a statistical analysis of system performance under different traffic conditions, a method to analyse and correct detection data is then proposed to reach satisfactory accuracy even in atypical installations.