When conducting traffic studies, data must be provide a clear picture to be useful. Depending on what data is important and what can be done with it, there are many different ways and types of data that can be collected. The following sections outline and describe each of the data types (Raw Per-Vehicle, Binned, Count, Sensor) you are able to collect using Diamond equipment.
|
The premise behind Raw Per-Vehicle data is that to gather
the information needed to completely identify and classify each individual vehicle, the time, and record of that vehicle. Essentially, Per-Vehicle data takes an information “picture” of each vehicle and stores it individually. Per-Vehicle data has the advantage of no loss of resolution so as to be able to process and view reports and studies without any sacrifice in accuracy or precision. The following is an example of Per-Vehicle data for a passenger car and a semi-truck:
|
|
With the information provided above, you can identify each individual vehicle. This identification can be done using a variety of sensors and sensor arrays (see below) all providing Per-Vehicle data. There are some differences between the calculations of the data according to the sensor array types. The following arrays have a description of their strengths and weaknesses.
|
Presence - Axle - Presence |
Two Presence style sensors and one axle sensor can classify vehicle axles, speed, and lengths. It is primarily used in permanent installation applications on high speed and low speed sites. This configuration is the most commonly used today on Interstate and highway ATR (Automatic Traffic Recorder) sites for both binned and Per-vehicle data.
Pros:
Cons:
|
|
Axle - Presence - Axle |
Two Axle style sensors and one Presence sensor can classify vehicle axles, speed, and lengths. This configuration is also used in permanent installation applications and occasionally in temporary installation sites. Depending on sensor types, it is useful in high speed and low speed applications.
Pros:
Cons:
|
|
Axle - Axle |
Two Axle sensors can classify vehicle axles, speed, and lengths with the exception of vehicle overall length (overall length calculated by adding all axle spacing lengths when a presence sensor is not used). The number of axles are also calculated using the axle classification definitions without the presence sensor information. This configuration is used primarily in temporary classification of low to medium speed applications.
Pros:
Cons:
|
With all Diamond classifiers, Per-Vehicle data is calculated and stored within the classifier to be transferred later to a computer for databasing and analysis. To make sure classification was captured correctly with the classifier, all Diamond classifiers have a Display showing the accuracy of each vehicle for real time verification out in the field.
Per-vehicle data once transferred into a PC can be converted to any classification data or count data (chart 1 shows the data path conversion). The conversion to classification data can be customized to any user or predefined class tables.
Binned Classification is data stored in “bins” as to later reference
|
it for analysis or viewing. The bins that data is stored in are
defined by the type of data and the bin definitions. The types of
bins are Axle, Speed, Gap, Headway, Length, Speed by Axle, and
Speed by Length. These bins are collected at set intervals
determined by the user when a study is setup. The bins are
individually defined by classification tables that are user
setable before data collection as well. As an example the
following is one interval of speed bins of one lane:
Lane 1 : 09:45 - 10:00 (15min interval)
|
Speed in MPH |
1-14 |
15-24 |
25-29 |
30-34 |
35-39 |
40-44 |
45-49 |
50-54 |
55-59 |
60-74 |
75-99 |
|
Count |
2 |
7 |
15 |
36 |
32 |
13 |
3 |
1 |
0 |
0 |
0 |
While the example above only shows a total of 11 separate bins, 30 user definable bins can be created for each bin type. As a standard, the FHWA has outlined the Scheme F for regular use in the United States. It contains the most commonly used definitions for classification data and provides a standard in which individual state traffic monitoring agencies can report their data.
Per-Vehicle data can be converted into Classification data at any time which makes Per-vehicle data very valuable. However, Binned Classification data has the advantage of saving memory for the traffic recorders as it takes up a fraction of the size as Per-vehicle data does. On large interstates and highways, the amount of data collected sometimes requires that Classification data be collected to save space in the traffic recorder. Binned Classification has the advantage of having to retrieve the data less frequently than with Per-Vehicle data. Most ATR sites and permanent installs collect data using the Binned Classification setting.
The sensor configurations used to collect Binned Classification data is the same as Per-Vehicle. Diamond equipment processes the information into Per-Vehicle data and then stores it as Binned Classification data for the most accurate results.
|
Count data is the simplest of data as it is just a total and does not contain any other information. There are two types of Count data: Volume Count data with Intervals and Volume Count data without intervals. Volume Count data with Intervals records a tally for a set period of time defined by the user. Once the time has elapsed, a new interval begins and the count
starts over. Volume Count data without intervals continually tally’s counts once it begins and does not stop until it has completed its study by either a user or programmed stop time.
Count data is commonly used for studies where traffic frequency, and growth rate is of interest. Cities typically use count data to extrapolate ADT (average daily traffic) and seasonal trends in traffic volumes. Since accuracy is not limited by speed or vehicle type, Count data can be collected at all speeds and on all roadway types. One sensor is typically used (road tube for temporary, inductive loops or piezos for permanent sites) to collect basic count data, but for directional and lane subtraction, two or more sensors are needed.
|
|
|
Sensor data is the individual recorded sensor events that have a time stamp to identify an activation or state change. Sensor data is used to create all other data types in counters and classifiers. The sensor data recorded in a classifier has a time stamp with a resolution of a hundred thousandths. Sensor data can be collected using any sensor array settings used by Classification or count data.
An example of sensor data is below:
|
|
Sensor data is unprocessed and is mostly unusable until processed into Per-Vehicle, Binned Classification, or Volume Count data. While sensor data would seem ideal in traffic gathering because of its versatility it does have some drawbacks.
Diamond Classifiers and Counters are capable of doing on site calculations to make the data verifiable on site. If only sensor data is gathered, classification data cannot be verified or validated until it is processed by a computer. All Diamond classifiers process the information within the unit as to provide instant feedback and allow for easy validation of the data that is collected.
Another drawback of Sensor time stamp data is that while it is precise, in order to save space in the classifier memory, resolution is lost and therefore not as accurate as Per-Vehicle or Binned Classification data stored within the classifier. We recommend that novice users avoid collecting sensor data for these reasons.
© 1991 - 2004 Diamond Traffic Products. All Rights Reserved.
