surveillance and control systems to create smart corridors that increase the effectiveness
of the transportation network. The infrastructure improvements and new technologies are,
in turn, being integrated with communications and displays in smart cars and public
access areas (such as shopping centers) to form intelligent transportation systems.
Vehicle detectors are an integral part of these modern traffic control systems. The types of
traffic flow data, as well as their reliability, consistency, accuracy, and precision, and the
detector response time are some of the critical parameters to be evaluated when choosing
a vehicle detector. These attributes become even more important as the number of
detectors proliferate and the real-time control aspects of ITS put a premium on the
quantity and quality of traffic flow data, as well as the ease of data interpretation and
integration into the existing traffic control system.
Current vehicle detection is based predominantly on inductive loop detectors (ILDs)
installed in the roadway subsurface. When properly installed and maintained, they can
provide real-time data and a historical database against which to compare and evaluate
more advanced detector systems. Alternative detector technologies being developed
provide direct measurement of a wider variety of traffic parameters, such as density
(vehicles per mile per lane), travel time, and vehicle turning movement. These advanced
detectors supply more accurate data, parameters that are not directly measured with
previous instruments, inputs to area-wide surveillance and control of signalized
intersections and freeways, and support of motorist information services. Furthermore,
many of the advanced detector systems can be installed and maintained without
disrupting traffic flow. The less obtrusive buried detectors will continue to find
applications in the future, as for example, where aesthetic concerns are dominant or
procedures are in place to monitor and repair malfunctioning units on a daily basis.
Newer detectors with serial outputs currently require specific software to be written to
interpret the traffic flow parameters embedded in the data stream. Since each detector
manufacturer generally uses a proprietary serial protocol, each detector with a unique
protocol requires corresponding software. This increases the installation cost or the real
purchase price of the detector. Furthermore, not every detector outputs data on an
individual vehicle basis. While some do, others integrate the data and output the results
over periods that range from tens of seconds to minutes, producing parameters that are
characteristic of macroscopic traffic flow. The traffic management agency must thus use
caution when comparing outputs from dissimilar detectors.
In performing the technology evaluations and in analyzing the data, focus was placed on
the underlying technology upon which the detectors were based [1,2]. It was not the
purpose of the program to determine which specific detectors met a set of requirements,
but rather whether the sensing technology they used had merit in measuring and reporting
traffic data to the accuracy needed for present and future applications. Obviously, there
can be many implementations of a technology, some of which may be better exploited
than others at any time. Thus, a technology may show promise for future applications, but
the state-of-the-art of current hardware or software may be hampering its present
deployment. The detectors that were used in the technology evaluations during the field
tests are listed in Table 1.
Technology
Manufacturer
Model
Output Data
U-1
Ultrasonic
Doppler
Sumitomo
SDU-200 (RDU101)
Count, speed
U-2
Ultrasonic
Presence
Sumitomo
SDU-300
Count,
presence
U-3
Ultrasonic
Presence
Microwave Sensors
TC-30C
Count,
presence
M-1
Microwave
Detector,
Motion,
Medium
Beamwidth
Microwave Sensors
TC-20
Count
M-2
Microwave
Detector,
Motion,
Medium
Beamwidth
Microwave Sensors
TC-26
Count, speed
binning
M-4a
Microwave
Detector,
Motion,
Narrow
Beamwidth
Whelen
TDN-30
Count, speed
M-5
Microwave
Detector,
Motion,
Wide Beamwidth
Whelen
TDW-10
Count, speed
M-6
Microwave
Radar,
Narrow
Beamwidth
Electronic Integrated
RTMS-X1
Systems
Count,
presence
speed,
occupancy
IR-1
Active IR,
Laser Radar
Schwartz
Electro-Optics
780D1000
(Autosense I)
Count,
presence,
speed
IR-2
Passive IR
Eltec
842
Count,
Presence
presence
IR-3
Passive IR
Pulse Output
Eltec
833
Count
IR-4 b
Imaging IR
Grumman
Traffic Sensor
Presence,
speed
VIP-1
Video Image
Processor
Econolite
AUTOSCOPE
2003
VIP-2
Video Image
Processor
Computer
Recognition
Systems
Traffic Analysis
System
VIP-3 e
Video Image
Processor
Traficon
CCATS -VIP 2
VIP-4b
Video Image
Processor
Sumitomo
IDET-100
VIP-5 c
Video Image
Processor
EVA
2000
A-1f
Passive Acoustic
Array
AT&T
SmartSonic
TSS-1
Count
MA-1
Magnetometer
Midian
Electronics
Self Powered
Vehicle Detector
Count,
presence
L-1 b
Microloop
3M
701
Count,
presence
Tube-Type
Vehicle
Counter
Timemark
Delta 1
Count
T-1
a M-3 was designated for a microwave radar detector that was not received.
b Used at Tucson Arizona test site only.
c Used in Phoenix Arizona 7/94 test only.
d Count, presence, occupancy, speed, classification based on length. Some
provide headway, density, and alarm functions.
e Used at all Arizona test sites.
f Used in Phoenix 11/93 and Tucson tests.
Not all detectors were available at all sites as shown in the footnotes to the table. A
summary of the advantages and disadvantages of the detector technologies is given in
Table 2. Some of them are application specific, implying that a particular technology
may be suitable for some but not all applications. A factor not addressed in this table is
detector cost. This issue is again application specific. For example, a higher cost detector
may be appropriate for an application requiring specific data or multiple detection zones
(suitable for multiple lane coverage) that are incorporated into the more expensive
detector.
Table 2. Advantages and Disadvantages of Candidate Detector Technologies
Technology
Advantages
Disadvantages
Ultrasonic
May be sensitive to
temperature and air
turbulence
Microwave Doppler
Good in inclement
weather
Directly measures
vehicle speed
Good in inclement
weather
Detects stopped
vehicles
Greater viewing
distance in fog than
with visible
wavelength sensors
Potential degradation by
heavy rain and heavy snow
Greater viewing
distance in fog than
with visible
wavelength sensors
Potential degradation by
obscurants in atmosphere and
by inclement weather
Directly measures
vehicle speed
Potential degradation by
inclement weather
Large vehicles can
obscure smaller vehicles
Shadows, reflections
from wet pavement, and
day/night transitions can
result in missed or false
detections
Assumptions
Potential Overhead
Technology
Signalized
intersection control
True-presence microwave
radar
Passive infrared
Laser radar
Ultrasound
Video image processor
Signalized
intersection control
True-presence microwave
radar
Ultrasound
Long-wavelength
imaging infrared video
processor
Signalized
intersection control
Detection of stopped
vehicles not required
Inclement weather
True-presence microwave
radar
Doppler microwave
detector
Ultrasound
Long-wavelength
imaging infrared video
processor
Real-time adaptive
signal control (e.g.,
SCOOT)
Vehicle counting
True-presence
microwave radar
Doppler microwave
detector
Passive infrared
Laser radar
Ultrasound
Video image
processor
Vehicle speed
measurement
Vehicle classification
By length
Vehicle classification
By profile
Laser radar
vehicles within its field of view using the Doppler principle, where the difference in
frequency between the transmitted and received signals is proportional to the vehicle
speed. Thus, the detection of a frequency shift denotes the passage of a vehicle. This type
of detector cannot detect stopped vehicles and is, therefore, not suitable for applications
that require vehicle presence such as at a signal light or stop bar.
The second type of microwave radar detector transmits a sawtooth waveform, also called
a frequency-modulated continuous wave (FMCW), that varies the transmitted frequency
continuously with time. It permits stationary vehicles to be detected by measuring the
range from the detector to the vehicle and also calculates vehicle speed by measuring the
time it takes for the vehicle to travel between two internal markers (range bins) that
represent known distances from the radar. Vehicle speed is then simply calculated as the
distance between the two range bins divided by the time it takes the vehicle to travel that
distance. Since this detector can sense stopped vehicles, it is sometimes referred to as a
true-presence microwave radar.
Passive Infrared Detectors
Passive infrared detectors can supply vehicle passage and presence data, but not speed.
They use an energy sensitive photon detector located at the optical focal plane to measure
the infrared energy emitted by objects in the detectors field of view. Passive detectors do
not transmit energy of their own. When a vehicle enters the detection zone, it produces a
change in the energy normally measured from the road surface in the absence of a
vehicle. The change in energy is proportional to the absolute temperature of the vehicle
and the emissivity of the vehicles metal surface (emissivity is equal to the ratio of the
energy actually emitted by a material to the energy emitted by a perfect radiator of energy
at the same temperature). The difference in energy that reaches the detector is reduced
when there is water vapor, rain, snow, or fog in the atmosphere. For the approximately 20
ft (6.1 m) distances typical of traffic monitoring applications with this type of detector,
these atmospheric constituents may not produce significant performance degradation.
Active Infrared Detectors
Active infrared detectors function similarly to microwave radar detectors. The most
prevalent types use a laser diode to transmit energy in the near infrared spectrum
(approximately 0.9 micrometer wavelength), a portion of which is reflected back into the
receiver of the detector from a vehicle in its field of view. Laser radars can supply vehicle
passage, presence, and speed information. Speed is measured by noting the time it takes a
vehicle to cross two infrared beams that are scanned across the road surface a known
distance apart. Some laser radar models also have the ability to classify vehicles by
measuring and identifying their profiles. Other types of active infrared detectors use light
emitting diodes (LEDs) as the signal source.
Ultrasonic Detectors
Ultrasonic vehicle detectors can be designed to receive range and Doppler speed data.
However, the most prevalent and low-cost ultrasonic detectors are those that measure
range to provide vehicle passage and presence data only. The ultrasonic Doppler detector
that also measures vehicle speed is an order of magnitude more expensive than the
presence detector. Ultrasonic detectors transmit sound at 25 KHz to 50 KHz (depending
on the manufacturer). These frequencies lie above the audible region. A portion of the
transmitted energy is reflected from the road or vehicle surface into the receiver portion
of the instrument and is processed to give vehicle passage and presence. A typical
ultrasonic presence detector transmits ultrasonic energy in the form of pulses. The
measurement of the round-trip time it takes for the pulse to leave the detector, bounce off
a surface, and return to the detector is proportional to the range from the detector to the
surface. A detection gate is set to identify the range to the road surface and inhibit a
detection signal from the road itself. When a vehicle enters the field of view, the range
from the detector to the top of the vehicle is sensed, and being less than the range from
the detector to the road, causes the detector to produce a vehicle detection signal.
Passive Acoustic Detectors
Vehicular traffic produces acoustic energy or audible sound from a variety of sources
within the vehicle and from the interaction of the vehicles tires with the road surface.
Arrays of acoustic microphones are used to pickup these sounds from a focused area
within a lane on a roadway. When a vehicle passes through the detection zone, the signalprocessing algorithm detects an increase in sound energy and a vehicle presence signal is
generated. When the vehicle leaves the detection zone, the sound energy decreases below
the detection threshold and the vehicle presence signal is terminated.
Video Image Processors
Video image processors (VIPs) identify vehicles and their associated traffic flow
parameters by analyzing imagery supplied by video cameras. Using personal computertype architectures, the images are digitized and then passed through a series of algorithms
that identify changes in the image background, that is changes in the quiescent contrast
level between the pixels (picture elements) that make up the image. Information about
vehicle passage, presence, speed, length, and lane change movement can be supplied,
depending upon the type of image processing technique used. Some VIPs insert vehicle
detection zones into the field of view and detect changes in pixel contrast in these areas
caused by vehicle passage; others track vehicles through the entire field of view by
identifying and following the path produced by the changes in pixel contrast. Artifacts
such light reflected from wet pavement and shadows have historically affected the
performance of VIPs. Since the VIP processes an image that can encompass several lanes
or images from multiple cameras, it is often a cost-effective approach for monitoring
traffic flow in multiple lanes and in multiple zones within a lane.
EFFECT OF DATA OUTPUT STRUCTURE ON COMPARING DETECTOR
OUTPUTS
When comparing output data from different detectors, the effect of unique data format
structures and data integration intervals must be recognized and accounted for [3]. The
performance of detectors with RS-232 interfaces, such as the speed-measuring
microwave detectors and the video image processors, can be difficult to compare due to
the lack of standardization of their data output intervals as shown in Table 4. For
example, the Whelen Doppler devices output data on a per vehicle basis, while the RTMS
microwave radar that was evaluated had software that limited the data output to a
minimum of approximately 10 seconds. The software has been modified in newer models
to allow data from individual vehicles to be output. The IDET-100 outputs vehicle
detections and computes speeds on a per vehicle basis, while the CCATS -VIP 2 outputs
results accumulated over 5-second integration intervals. The serial interface protocol for
the AUTOSCOPE was not available for the field tests. Thus, the only data recorded
from the AUTOSCOPE were: (1) the transition of the output state of the optically
isolated transistors in the electronic interface module, and (2) the time of the event
corresponding to the passage of a vehicle. In order to compare the outputs from the
detectors under evaluation, the data were integrated (during post-processing) over an
interval equal to the least common multiple of the collection intervals used by the devices
in the comparison group.
Table 4. Data and Update Intervals in Detectors with RS-232 Interfaces
Detector
Update
Interval
Count
Lane
Occ.
Speed
Vehicle
Typeb
10
seconds
to 10
minutesa
10 s to 1
h
1 minute
5
seconds
per
vehicle
per
vehicle
1 second
To return to ITS Online's Cover Page, press the "back" button on your web browser or
click on our logo below.