 |
Smart Loop Technology Demonstration Project
at Otay Mesa, San Diego
Functional overview of Smart Loop Detection System
 The Link
Processor is connected to Traffic Management Center (TMC) computer via telephone
lines. All Link Processors transfer the information coming from IVS-2000 to the TMC computer.
The TMC computer, upon receiving the vehicle data, stores the information in the database. The
database calls the correlation and assignment functions to perform vehicle origin and destination
matching from one loop to the next set of loops. The database starts the Graphical User Interface
that runs on the 21-inch display screen.
The display is divided into three sections. The first section displays the segments located on Otay
Mesa Road. The segments can be selected in order to display the visual representation of the traffic
along with the average statistics that appear in the second and third sections of the display. The second
section displays an animated representation of traffic with the help of vehicle icons. The lanes and
sensors are drawn on the screen. The sensors can be selected to display the real-time information coming
from the IVS-2000 in section three. The third section of the display has three different tables to present
segment statistics, classification information, and point sensor information on Otay Mesa Road.
The display also has control buttons to start, pause, and quit the system. Current time and date are
displayed. The figure below shows an example of the Otay Mesa Road user interface.
Otay Mesa Road Location of Inductive Loop Instrumentation.
To the far left is the 805 Freeway and to the far right is the Otay Mesa border crossing.
The system installed at Otay Mesa Road consists of 25 inductive loops embedded in the road. Associated
with these loops are four cabinets, which house six or seven IVS-2000s that detect and classify the vehicles. The
Link Processor in each cabinet collects the individual detections from each IVS-2000 and forwards the information
to the Transportation Management Center (TMC) computer. The TMC computer has its own database management functions
to store the incoming data. The TMC computer computes and displays the statistics and real-time information on the user interface.
Vehicle Classification
The OMR Smart Loop System currently classifies 23 different classes of vehicles, shown in Table 1, and one
unknown type. Table 1 also shows the silhouettes of the 23 different classes of vehicles and their respective
average lengths. Vehicle length is a function of the IVS-2000’s classification of the vehicle. An average
vehicle length for each class has been determined. These same silhouettes are used on the GUI. Vehicle
length is a function of the IVS-2000’s classification of the vehicle. An average vehicle length for each
class has been determined.
Vehicle Classifications and Lengths
|
Class 1: Motorcycle
|
5.6 feet
|
 |
|
Class 2: Passenger car
|
17.4 feet
|
 |
|
Class 3: 2-axle, 4-tire single units
|
19.1 feet
|
 |
|
Class 4: Buses
|
41.7 feet
|
 |
|
Class 5: 2-axle, 6-tire single units
|
29.0 feet
|
 |
|
Class 6: 3-axle single units
|
34.0 feet
|
 |
|
Class 7: 4+ axle single units
|
51.2 feet
|
 |
|
Class 8: 4 or less axle single trailers
|
48.0 feet
|
 |
|
Class 9: 5-axle single trailer trucks
|
62.4 feet
|
 |
|
Class 10: 6 or more axle single trailer
trucks
|
71.2 feet
|
 |
|
Class 11: 5 or less axle multi trailer
trucks
|
70.0 feet
|
 |
|
Class 12: 6-axle multi-trailer trucks
|
77.5 feet
|
 |
|
Class 13: 7 or more axle multi-trailer
trucks
|
No Data
|
 |
|
Class 14: Class 2 + trailer
|
27.4 feet
|
 |
|
Class 15: Class 3 + trailer
|
39.1 feet
|
 |
|
Class 16: Class 5 + trailer
|
44.0 feet
|
 |
|
Class 17: Class 6 + trailer
|
63.0 feet
|
 |
|
Class 18: Loaded auto carrier
|
83.1 feet
|
 |
|
Class 19: Empty auto carrier
|
80.0 feet
|
 |
|
Class 20: Bobtail tractor
|
24.0 feet
|
 |
|
Class 21: Combination tractor-trailer
|
64.4 feet
|
 |
|
Class 22: 30-foot bus
|
32.4 feet
|
 |
|
Class 23: 20-foot bus
|
24.0 feet
|
 |
These individual vehicle classifications are grouped into Caltrans-specified categories based on the number
of axles, as shown below. These categories are used for the display on the GUI.
|
Passenger vehicles
|
Class :
2, 3, 14, 15
|
|
2- and 3-axle
|
Class :
4, 5, 6, 8, 16, 20, 22, 23
|
|
4-axle
|
Class :
7
|
|
5 or more axle
|
Class :
9, 10, 11, 12, 13, 17, 18,
19, 21
|
|
Others
|
Class :
1
|
|
Unknown/unclassifiable
|
Class :
24
|
Speed Accuracy
The speed accuracy was tested by using a control vehicle passing through the system crossing over different
sensors of different cabinets. The results of the 10 passes over the sensors are shown below.
The speed accuracy is to within 5 percent and all but one of the trials the controlled vehicle speed
was faster then that of the log file, which is expected.
The Figers below shows the traffic speeds west bound during a 24 hour period. Notice how the truck
traffic is about 10 mph slower then car traffic heading west bound.
Truck and passenger vehicle speed on OMR west bound using a single loop sensor. The Figers below shows the
traffic speeds east bound during a 24 hour period. Again the truck traffic is 10 mph slower then the cars.
Truck and passenger vehicle speed on OMR east bound using a single loop sensor.
Counting Accuracy
The counting accuracy was based on the collected data for the classification. This was based on more than
500 test vehicles. The figure below show the counting accuracy of all the cabinets east- and west-bound.
The east-bound side of cabinet 3 was not functioning properly during our test and hence omitted. The counting
accuracy for the system was above 95 percent accurate.
One can isolate the trucks from the passenger vehicle per lane over a 24 hour period as seen from the figure
below. As expected the trucks and passenger vehicles count are closer together in the slow lane (lane 3) then
in the other two lanes during none compute times.
Truck and passenger vehicle speed east bound isolated by lane. The total traffic for a 24-hour period on
September 11 for east- and west-bound traffic.
Tracking Capabilities
The tracking test was performed using a controlled vehicle through the system on a different day than the
original test. Six runs through the system were performed and, of those, two were completely tracked. Two
runs were tracked for two of the three sections and on two other occasions, the controlled vehicle was tracked
on only one section. On all missed track scenarios, a similar vehicle passed through the sensor during the window
of classification. This window of opportunity is at a fixed time duration for the entire day and does not depend
on any additional information such as traffic light information. The more information that is given to the tracker
about the flow of traffic, the more accurate the estimation of the time at which the vehicle will
The figer below shows the classification of passenger vehicles (category 1) versus trucks (category 2, 3,
and 4). The truck classification is above 95 percent accurate.
|
|
