| Title | Continuous Vehicle Classification Data: How Good Is It? |
|---|---|
| Record ID | 10984 |
| Personal Name Creator |
Reel, Richard L., Jr. |
| Source | NATMEC 2000 Conference, Middleton, Wisconsin, 20000827-20000831 |
| Corporate Creator | Florida. Department of Transportation |
| Publisher | Florida Department of Transportation |
| Publication Date | 20000800 |
| Language | English |
| Abstract | Florida has a lengthy history of trying to obtain continuous vehicle classification data. They installed their first piezoelectric axle sensors at a continuous count site in October of 1988. At that time, they had 86 continuous count sites operating around the state, most of which had a pair of loops in each lane. Their thinking, at the time, was that all they would have to do is install a single piezo between each pair of loops, swap the counters, and they'd be collecting continuous classification data. Why is it that the theory is always simpler than the practice? The first obstacle to be overcome was convincing the equipment manufacturers to design a classifier to use a loop-piezo-loop sensor configuration. Their first piezo installation at one of their count sites having two loops per lane taught them a few things. First, simply installing one piezo between two loops isn't that simple a task; second, slot excavation for piezo sensor installation was a time consuming process; and third, they needed to find a better epoxy. Eventually they solved their installation problems, but came to realize that their concern over the durability of the piezo sensors was well founded - most of the piezo sensors failed within 2 years. The cause of failure has been attributed to the deformation of the asphaltic pavement in the wheel paths due to applied wheel loads (also known as "rutting"). Over the past 12 years they've continued to refine their construction techniques and materials, and currently have 248 continuous vehicle classification sites in operation. So how good is the data that is being produced? In 1999, these classifiers generated 132,054 (of a possible 181,040) days of directional class data. Fifteen (15) sites produced no usable vehicle class data, and another 10 sites produced usable data in a single direction only. In all, they flagged 43% if their class data as bad, and 57% as good. 10 p. |
| Rosap ID | dot:4812 |
| Rosap URL | https://rosap.ntl.bts.gov/view/dot/4812 |
| TRT Terms | Vehicles; Classification; Data collection |
| General Subjects | Vehicle classification; Types of vehicles |
| Geographical Coverage |
Florida; United States |
| TRIS Online Accession No |
00811900 |
| Resource type | Tech Report |
| URL | https://ntlrepository.blob.core.windows.net/lib/10000/10900/10984/045ppr.pdf |
| Format | |
| Database | NTL Digital Repository |