Lawrence A. Klein
Consultant
Ten different detector technologies were recently evaluated as part of the FHWAsponsored Detection Technology for IVHS program. The two primary goals of the program were:
1.
To determine traffic parameters and their corresponding measurement accuracies for future Intelligent Transportation Systems (ITS) applications, and
2.
To perform laboratory and field tests with above-the-road mounted, surface, and subsurface detectors to determine their performance.
Detectors representative of all tested technologies were found to satisfy current traffic management requirements. However, improved accuracies and new types of information, such as queue length and vehicle turning or erratic movements, may be required from detectors for future traffic management applications. The choice of a detector for a specific application is, of course, dependent on many factors, including data required, accuracy, number of lanes monitored, number of detection zones per lane, detector purchase and maintenance costs, vendor support, and compatibility with the current and future traffic management infrastructure.
The results of this evaluation project is being presented in two parts. Part 1 introduces the theory of operation and the strengths and weaknesses of the various overhead detector technologies. Part 2 will provide field evaluation data and some general conclusions about detector performance and applications. Copies of the Final Report, a set of five compact disks containing the detector evaluation data, and other reports are available from the FHWA by writing to Mr. Pete Mills at HSR-1, 6300 Georgetown Pike,
McLean, VA 22101.
Note: The detector performance data presented in this article were obtained by Dr. Klein when he was the project’s Principal Investigator at Hughes Aircraft Company.
INTRODUCTION
Maximizing the efficiency and capacity of the existing ground transportation network is made necessary by the continued increase in traffic volume and the limited construction of new highway facilities in urban, intercity, and rural areas. Smart street systems that contain traffic monitoring detectors, real-time adaptive signal control systems, and motorist communications media are being combined with freeway and highway
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 .
Symbol Technology
U-1
Ultrasonic
Doppler
U-2
Ultrasonic
Presence
Table 1. Detectors Used During Field Tests
Manufacturer
Sumitomo
Sumitomo
Model
SDU-200 (RDU-
101)
SDU-300
U-3
M-1
Ultrasonic
Presence
Microwave
Detector,
Motion,
Medium
Beamwidth
Microwave Sensors
Microwave Sensors
TC-30C
TC-20
M-2
M-4
M-5
M-6 a
Microwave
Detector,
Motion,
Medium
Beamwidth
Microwave
Detector,
Motion,
Narrow
Beamwidth
Microwave Sensors
Whelen
TC-26
TDN-30
Microwave
Detector,
Motion,
Wide Beamwidth
Whelen
Microwave
Radar,
Narrow
Beamwidth
Electronic
Integrated Systems
TDW-10
RTMS-X1
IR-1
IR-2
Active IR,
Laser Radar
Passive IR
Presence
Schwartz
Electro-Optics
Eltec
780D1000
(Autosense I)
842
Output Data
Count, speed
Count, presence
Count, presence
Count
Count, speed binning
Count, speed
Count, speed
Count, presence speed, occupancy
Count, presence, speed
Count, 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 d
VIP-2
Video Image
Processor
Computer
Recognition
Systems
Traffic Analysis
System d
VIP-3 e
Video Image
Processor
VIP-4 b
Video Image
Processor
VIP-5 c
Video Image
Processor
A-1 f
MA-1
L-1
b
Magnetometer
Microloop
Traficon
Sumitomo
EVA
Passive Acoustic
Array
AT&T
Midian
Electronics
3M
CCATS
IDET-100
2000
-VIP 2 d d d
SmartSonic
TSS-1
Count
Self Powered
Vehicle Detector
Count, presence
701
Count, presence
T-1 b
Tube-Type
Vehicle
Counter
Timemark Delta 1 Count 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
Microwave Doppler
Compact size, ease of installation
Good in inclement weather
Directly measures vehicle speed
May be sensitive to temperature and air turbulence
Cannot detect stopped vehicles or vehicles moving less than approximately 5 mph
Microwave true presence Good in inclement weather
Detects stopped vehicles
Operates in sidelooking mode
Requires narrow beam antenna to confine footprint to single lane in forwardlooking mode
Passive (receive only) infrared
Active (transmit and receive) infrared
Visible spectrum video image processor
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
Directly measures vehicle speed
Potential degradation by obscurants in atmosphere and by inclement weather
Provides data for traffic management and imagery for incident management
Potential degradation by inclement weather
Large vehicles can obscure smaller vehicles
Single camera and processor can serve multiple lanes
Rich array of traffic data provided
Shadows, reflections from wet pavement, and day/night transitions can result in missed or false detections
Table 3 shows examples of overhead detector technology compatibility with several traffic management applications. The assumptions shown concerning the application dictate, in part, the appropriateness of the technology.
Table 3. Overhead Detector Technology Applications to Traffic Management
Application
Signalized intersection control
Signalized intersection control
Signalized intersection control
Real-time adaptive signal control (e.g.,
SCOOT)
Assumptions Potential Overhead
Technology
Detect stopped vehicles
Weather not a major factor
Detect stopped vehicles
Inclement weather
Detection of stopped vehicles not required
Inclement weather
True-presence microwave radar
Passive infrared
Laser radar
Ultrasound
Video image processor
True-presence microwave radar
Ultrasound
Long-wavelength imaging infrared video processor
True-presence microwave radar
Doppler microwave detector
Ultrasound
Long-wavelength imaging infrared video processor
Desirable for detector footprint to emulate a 6-ft by
6-ft inductive loop
Side-mounting capability
Video image processor
True-presence microwave radar
Passive infrared (with suitable aperture beamwidth)
Vehicle counting
(surface street or freeway)
Detect and count vehicles traveling at speeds greater than 2 to 3 mi/h
True-presence microwave radar
Doppler microwave detector
Passive infrared
Laser radar
Ultrasound
Video image processor
Vehicle classification
Vehicle speed measurement
Vehicle classification
Detect and count vehicles traveling at speeds greater than 2 to 3 mi/h
By length
By profile
True-presence microwave radar
Doppler microwave detector
Laser radar
Video image processor
Video image processor
Laser radar
Laser radar
THEORY OF OVERHEAD DETECTOR OPERATION
The following paragraphs give a brief explanation of the underlying operating principles for microwave, passive infrared, active infrared, ultrasonic, passive acoustic, and video image processor detectors.
Microwave Radar
Microwave radars used in the U.S. for vehicle detection transmit energy at 10.525 GHz, a frequency allocated by the FCC for this purpose. Their output power is regulated by the
FCC and certified by the manufacturer to meet FCC requirements. No further certification is required of the transportation agencies for their deployment.
Two types of microwave radar detectors are used in traffic management applications. The first transmits electromagnetic energy at a constant frequency. It measures the speed of 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 detector’s 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 vehicle’s 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 vehicle’s 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
Type b
Whelen TDN-30 & TDW-10
Doppler Detectors per vehicle x x x x x Electronic Integrated Systems
RTMS-X1 True Presence
Microwave Radar
10 seconds to 10 minutes a
Econolite AUTOSCOPE 2003 VIP c
10 s to 1 h x x x x x x x x Computer Recognition Systems
Traffic Analysis System VIP
Traficon CCATS -VIP 2
1 minute
5 seconds x x x
Sumitomo IDET-100 VIP x x x
EVA 2000 VIP per vehicle per vehicle x x x x
Grumman Infrared VIP 1 second x x a User selected in 10-s increments. Update interval set to minimum value of 10 s in field tests.
b Based on user-selected vehicle lengths. c AUTOSCOPE serial data were not made available during the field tests.
IN PART 2...
In the next installment of this article, the measured performance of many of these detectors will be presented using data gathered at the freeway and surface-street arterial evaluation sites. Summaries of the detector accuracies and a qualitative assessment of the best performing detector technologies for acquiring specific traffic flow parameters will be shown.
References
[1] Detection Technology for IVHS – Task L Final Report , Federal Highway
Administration Contract DTFH61-91-C-00076, U.S. Department of Transportation,
Wash., D.C., 1995.
[2] Detection Technology for IVHS – Final Report Addendum , Federal Highway
Administration Contract DTFH61-91-C-00076, U.S. Department of Transportation,
Wash., D.C., 1995.
[3] L.A. Klein, M.R. Kelley, and M.K. Mills, " Evaluation of Overhead and In-Ground
Vehicle Detector Technologies for Traffic Flow Measurement," Journal of Testing and
Evaluation , JTEVA, Vol. 25, No. 2, March 1997, pp. 215-224.
ABOUT THE AUTHOR
Lawrence A. Klein is a consultant developing multiple sensor system, communications, and system architecture concepts for intelligent transportation systems and defense applications. He holds a Ph.D. in electrical engineering from New York University.
While at Hughes Aircraft Company, Dr. Klein was the Principal Investigator on the
Detection Technology for IVHS program and led several ITS architecture development projects. In addition to extensive publications in the ITS arena, he has authored the books
Sensor and Data Fusion Concepts and Applications for the SPIE and Millimeter-Wave and Infrared Multisensor Design and Signal Processing for Artech House. Dr. Klein can be reached at: Phone: (714) 996-9066. E-mail: laklein@wdc.net
.
To return to ITS Online's Cover Page , press the "back" button on your web browser or click on our logo below.
© 1997 by
Communication Alchemy, Inc.
, Austin, TX.
Phone: (512)219-1966 FAX: (512)219