Survey and Evaluation of Ice/Snow Detection Technologies Final Report

Survey and Evaluation of Ice/Snow Detection Technologies Final Report
NORTHLAND ADVANCED TRANSPORTATION SYSTEMS
RESEARCH LABORATORIES
Final Report
Research Project Title:
Survey and Evaluation of Ice/Snow Detection
Technologies
Principal Investigators Information:
Name: Fernando Rios-Gutiérrez
Department: ECE
Position Title: Assistant Professor
Address: 271 MWAH
Phone : 726-8606
Fax : (218) 726-7267
E-Mail : friosgut@d.umn.edu
Name: Mohammed A. Hasan
Department: ECE
Position Title: Associate Professor
Address: 271 MWAH
Phone: 726-6150
Fax: (218) 726-7267
E-Mail: mhasan@d.umn.edu
Index
INDEX ............................................................................................................................................................ 2
FIGURE INDEX ............................................................................................................................................ 2
ACKNOWLEDGEMENTS .......................................................................................................................... 3
SURVEY AND EVALUATION OF ICE/SNOW DETECTION TECHNOLOGIES ............................ 4
ABSTRACT ........................................................................................... ERROR! BOOKMARK NOT DEFINED.
BENEFITS...................................................................................................................................................... 4
INTRODUCTION ............................................................................................................................................ 5
IRID LOCATION ................................................................................... ERROR! BOOKMARK NOT DEFINED.
RESEARCH GOALS ....................................................................................................................................... 6
IRID SENSOR'S HARDWARE AND SOFTWARE .............................................................................................. 6
TASKS DESCRIPTION ............................................................................................................................... 8
INITIAL SYSTEM CHARACTERIZATION ......................................................................................................... 8
Pan and Tilt Angle Measurements and Conversions ............................................................................. 8
TESTING POINTS IDENTIFICATION.............................................................................................................. 12
TESTING METHODOLOGY .......................................................................................................................... 14
Baseline Identification. ........................................................................................................................ 15
TESTING RESULTS .................................................................................................................................. 16
Initial Testing ....................................................................................................................................... 16
PROBLEMS ENCOUNTERED ........................................................................................................................ 20
CONCLUSIONS............................................................................................................................................ 21
FUTURE WORK .......................................................................................................................................... 21
Figure Index
FIG. 1. - IRID ORIGINAL LOCATION IN DOWNTOWN DULUTH .......................................................................... 5
FIG. 2. - IRID NEW LOCATION IN AN OFF RAMP NEAR CLOQUET, MN .............................................................. 6
FIG. 3 .- IRID SENSOR NEW LOCATION ............................................................................................................. 7
FIG. 4. - IRID HEIGHT AND BEAM MAXIMUM RANGE ........................................................................................ 9
FIG. 5. - SENSOR’S PAN AND TILT ANGLES REFERENCE .................................................................................... 9
FIG. 6. - RELATIONSHIP BETWEEN TILT AND PAN ANGLES AND ROAD POINT COORDINATES ........................... 10
FIG. 7 .- IRID POSITIONING WINDOW ............................................................................................................. 12
FIG. 8. - TOOLSET PROGRAM MAIN WINDOW .................................................................................................. 13
FIG. 9. - TESTING POINTS IDENTIFICATION ..................................................................................................... 14
FIG. 10 .- SUMMER INITIAL MEASUREMENTS ................................................................................................. 17
FIG. 11 .- ICE ON ROAD MEASUREMENT DETAIL ............................................................................................ 18
FIG. 12 .- SENSOR’S RESPONSE TO DEICING CHEMICALS ............................................................................... 19
FIG. 13 .- SENSOR RESPONSE FOR CEMENT SURFACE .................................................................................... 19
FIG. 14 .- IRID ICE/SNOW CHARACTERIZATION ............................................................................................. 20
FIG. 15 .- WEATHER CONDITIONS DISPLAYED BY THE IRID’S CAMERA ......................................................... 22
2
ACKNOWLEDGEMENTS
We would like to thank a number of people who have assisted in performing the research
presented in this report. Particularly, we thank Ed Fleedge for his help and advice in the
initial setting of the IRID system, Dave Keranen for his technical support during the
development of the project, and Carol Wolosz for her patient and support. Finally, we
thank NATSRL for the financial support to conduct this research.
3
Survey and Evaluation of Ice/Snow Detection Technologies
Summary
Weather is a principal factor that contributes to traffic accidents. Thus Road and Weather
Information Systems (RWIS) has been deployed by MNDOT to proactively detect
adverse weather and road conditions so as to provide motorists with advanced warning of
hazardous conditions. Ice on the roadway is one of the leading contributors to winter
weather accidents. There are many ice detection technologies, however it is not clear
whether these sensors are accurate for detection of ice on the roadway surface. The
development of a reliable ice detection sensor would provide MNDOT engineers and
maintenance personnel the tools they need to warn drivers of potentially hazardous road
conditions due to ice formation on the road surface and mobilize MNDOT's maintenance
fleet with anti-icing treatments to the road surface.
The usefulness of any weather sensor is determined by the accuracy of the parameter(s) it
sensed. An accurate ice detection sensor could provide the tools necessary for engineers
to make informed decisions on proper use and sensor specifications. Weather sensors
accuracy is affected by temperature, light availability, visibility, pavement’s conditions
and wind. Often, many vendors do not provide detailed information regarding sensor
specifications and proper application. In this research, a thorough evaluation of the
Infrared Road Ice Detection System IRID is being conducted. IRID, which is an active IR
remote ice sensor, offers distinct advantages over embedded road sensors. It has lower
installation costs, lower cost of ownership, improved safety, and gives better results.
The objective of this research has been to investigate the IRID sensor in terms of
accuracy and sensitivity to distance and different deicing materials. Different
measurements have been collected in different weather conditions, and on concrete and
asphalt pavements. Data analysis indicates that this sensor is sensitive to weather
conditions and the presence of two contaminants salt brine (NaCl) and Magic
(Magnesium Chloride). Thus the ultimate goal after a successful evaluation is to mount
this sensor on a bridge on a busy highway and use it to monitor the weather conditions
remotely. In addition to the infrared ice sensor, the IRID comes with a camera that can
be used to show pictures of different locations near the pavement using pan/tilt capability.
Benefits
Automating pavement condition sensing in locations that experience severe winter
weather conditions is important for both the public and highway maintenance personnel.
This research could significantly improve highway safety by providing a better
understanding of IRID sensing capabilities. The development of an accurate ice
detection sensor could provide the tools necessary for engineers to make informed
decisions on sensor specifications and proper application.
4
Fig. 1. - IRID Original Location in Downtown Duluth
Introduction
Remote measurement of amount and type of precipitation that can be accessed remotely
is very important in many transportation applications. This information is crucial for the
maintenance manager in deploying available resources most effectively [1]. There are
many ice detection technologies, however it is not clear whether these sensors are
accurate for detection of ice on the roadway surface [2]. An active IR remote ice sensor
offers distinct advantages over embedded road sensors [3, 4]. It has lower installation
costs, lower cost of ownership, improved safety, and gives better results. Additionally,
IRID sensors can be mounted on roads or bridges. Information regarding the pavement
surface condition at the sensor location can be accessed through a server based Internet
Web page.
This project will help to improve the understanding on the use of a non-intrusive sensor
to detect the presence of ice, snow, water, etc, on the road’s surface and to evaluate its
accuracy. Based on these parameters we will evaluate its ability to detect development of
pavement icing and the present of chemical agents on the roadway, and as a final result to
improve the level of confidence on both the road’s maintenance crews and the traveling
public.
Currently there is no comprehensive spectral guide for deicing salts and freezing point
depressants [5,6] that can be used in remote sensing applications. In this research we used
the IRID sensor (Figure 1), to test if is possible to use, identify, and differentiate among
these chemicals.
5
The Innovative Dynamics Inc.’s (IDI) Active Infrared Ice Detection Sensor (IRID) was
relocated from downtown Duluth, to the University of Minnesota Duluth’s Advanced
Sensor Research Laboratory (ASRL), which is located outside Duluth on I-35 (Figure 2).
During the spring, 2003 the infrared sensor was re-installed after the manufacturer fixed
some hardware problems. The IRID sensor was designed originally to detect the presence
of water, snow and ice on the roadway [7]. We are presently investigating the
performance and applications of IRID.
Fig. 2. - IRID new Location in an off ramp near Cloquet, MN
Research Goals
The main goals of this research are:
1) To examine the performance of the refurbished IRID Active Infrared sensor in
terms of detection and accuracy.
2) To measure the sensitivity of this sensor to different surface conditions of the
pavement and to investigate whether it can be used to measure the thickness of
ice, water, or snow layers
3) To evaluate IRID to determine whether the presence and concentration of freezing
point depressant chemicals can be detected and measured on the roadway surface.
IRID Sensor's Hardware and Software
In May 2002, the IRID sensor was removed from its original location in downtown
Duluth, and relocated on top of a bridge, on a side road along Interstate 35, nearly 20
miles south from UMD (Figure 3).
6
Fig. 3 .- IRID sensor new location
The IRID sensor is designed to detect the presence of water, snow and ice on the
roadway. It is a non-contact and non-intrusive technology that does not require embedded
or contact sensors. Although this kind of sensor provides information only at a limited
area of the road, its pan/tilt capability could provide reliable information about the
pavement surface in multiple points.
The sensor uses a three-program software package to control its operation. These
programs are Main, Dispatcher and Toolset. The Main program is used to provide the
number of testing points and to specify their location on the road and the order in which
they will be tested. This set of points is called the Road Map Canvas.
The Main program is used to provide the number of testing points and to specify their
location on the road and the order in which they will be tested. This program main use is
to create the road map canvas.
The Toolset is the principal program used to control the infrared sensor. This program
can be used to set or modify the sensor parameters such as gain, position, base line, etc.
In addition, this program is used to receive the raw information measured by the sensor
for the points specified in the road map canvas.
7
The Dispatcher program, as its name implies, is used by the main operator of the system
to display the weather conditions retrieved from the sensor, using a user friendly display
window that specifies the present status (dry, wet, snow, ice, etc) for each point in the
road map canvas.
Tasks Description
During the period of this project, we achieved the following tasks:
1) Testing and performance evaluation of the IRID system.
2) To investigate the effectiveness of this system in detecting dry, wet, water, snow
and ice covered road surface conditions
3) Evaluation of ranging and accuracy of detection of water, snow and ice layers.
4) To analyze the performance using different situations as to detect ice through a
thin layer of snow, or to detect ice or snow in the presence of deicing chemicals,
etc.
5) Characterize road contaminants and environmental factors
6) Explore the detection of deicing chemicals such as Sodium Chloride and MAGIC.
7) Define if the overall IRID system can be used as a non-invasive, remote pavement
ice detector to be used as an Intelligent Road Weather System.
Initial System Characterization
Water and ice have been characterized as having different reflectance spectra in the near
and middle infrared region [7, 8]. Because of this, we wanted to evaluate the response of
the IRID sensor to different weather conditions on the road. In order to accomplish this
we identified a set of testing points on the pavement to be used later through the sensor
pan/tilt operation. This process is described next.
Pan and Tilt Angle Measurements and Conversions
As mentioned before, the sensor is now located on top of a metallic bridge (18 feet
above the road level), pointing downwards to the road, as shown in Figures 3 & 4.
8
Fig. 4. - IRID height and beam maximum range
From its original location the sensor can be rotated horizontally in a 360-degree angle
(pan angle), and rotated vertically in a 90-degree angle (tilt angle), as shown in Figure 5.
With this, the sensor can be positioned to point to any location dawn the road in a half
sphere, with a maximum radius of about 300 feet (as shown in Figure 4).
Fig. 5. - Sensor’s Pan and Tilt angles reference
9
Consequently, one of the first activities that had been performed was to develop a
practical procedure to determine the spherical coordinates (tilt and pan angles) that
correspond to any point on the road, i.e., we needed to determine the relationship between
the spherical coordinates and the latitude and longitude angles or (tilt/pan). This
relationship can be derived as follows. Assuming that the sensor is the origin of a
spherical system in which the z-axis is the line passing through the sensor and
perpendicular to the road. Using the diagram shown in Figure 6, the following equations
are obtained:
Sensor
Area of IRID beam
Fig. 6. - Relationship Between Tilt and Pan Angles and Testing Point Coordinates
( 1)
( 2)
( 3)
( 4)
Clearly,
( 5)
Hence:
( 6)
10
( 7)
( 8)
From these equations, it follows:
( 9)
( 10)
( 11)
( 12)
Thus, with the measured height h = z = 18 feet, we compute the latitude and longitude
angles using the above equations. In the control programs (Main, Toolset and Dispatcher)
the pan and tilt angles are given using angular displacements of the mechanisms, and
they have a range of 0-5607 and 5400-10300 respectively. Given these ranges, we needed
to find a relationship between these units and the respective angles. We verified through
measurements that each unit of the pan angle corresponds to approximately 6 degrees. So
that we can write the following equations:
( 13)
Moreover, the relationship between θ and tilt angle is given by:
( 14)
Using Equations 1 through 14, the correct pan and tilt angles are entered as inputs to the
Toolset program (see Figure 7). This triggers IRID to illuminate the location of the
testing point with infrared energy. A red laser pointer also indicates this location
visually.
11
Thus using a measuring tape, the Cartesian coordinates are converted to spherical as tilt
and pan angles for any point on the road. This task would not have been necessary if
IRID software is fully operational.
Fig. 7.- IRID Positioning Window
Next, we proceeded to identify test points on the road. We selected two types of surface
for testing: cement and asphalt to verify if the sensor detects any possible differences
between these surfaces. At this time, there are no indications of any differences for the
tested points under dry conditions.
Testing Points Identification
During the summer 2003, the infrared sensor was re-installed after the manufacturer fixed
some hardware problems. Several measurements for the road under different weather
conditions were taken using the only functioning software Toolset. The main window for
this program is shown in Figure 8.
Using the toolset program the user can specify different parameters to initialize and
control the operation of the infrared sensor. Some of the parameters that can be specified
by the user are Gains, Thresholds, Baselines values, etc. Also on the bottom of this
window the user can observe the sensor’s actual measurements (circled values), notice
that these four values represent the response of the sensor to the road conditions as
measured using the two wavelengths (1550 and 1430 nm), and two polarizations (S and
P). These values are the ones that we used in our measurement and characterization
process for the road’s surface weather conditions.
12
Fig. 8. - Toolset program main Window
The next task that we accomplished was to define an appropriate set of testing points to
be used in our measurements (Figure 9). The locations for the testing points were selected
so that we could test the sensitivity of the sensor to both location and road type (asphalt
and cement). For this research, we identified a total of six testing points. Four of the
testing points are on the asphalt side, two are located far from the sensor; two more are
located closer to the sensor. The other two points are located on the concrete side of the
road, one far and the other nearer to the sensor. The points were marked so that they
could be easily identified. The coordinates of the testing points are given in Table 1.
The coordinates of these points were found by measuring the perpendicular distance d
and height z (see Figure 6). Then, by using Equations 1-14 we devised a mathematical
method to find the pan and tilt angle for the sensor in terms of the actual coordinates of
the testing points.
13
Fig. 9. - Testing Points Identification
The Pan and Tilt angles found by applying the set of equations given above are shown in
Table 1.
Point
1
2
3
4
5
6
Pan Angle Tilt Angle Road Type
0
8920
Asphalt
200
8060
Asphalt
1160
6910
Asphalt
2340
8110
Asphalt
5407
8060
Cement
5407
7032
Cement
Table 1. - Testing Point Coordinates
Testing Methodology
During the summer 2003, we proceeded to test the sensor by simulating wet and winter
weather conditions on the road’s surface. Thus we devised a method to artificially create
wet, icy and frosty conditions on the road surface in the region of interest. Water
detection tests were performed by spraying water on the area of interest to create damp
and wet situations. To get thicker layers of water, we simply poured water on the testing
spot to simulate heavy rain weather conditions. To create ice, we used blocks of dry ice
(CO2) to cool the pavement. It takes about 30 minutes for the pavement to cool bellow
the freezing point of water. After the blocks of dry ice are removed or evaporated, we
14
used a mist sprayer to create uniform layers of ice and frost. Using the dry ice, we were
able to create uniform layers of frost and ice of different thickness.
To measure the water and ice thickness, we used a feeler gauge and a micrometer to
estimate their value. It should be noticed that by using this method, we couldn’t get a
precise thickness measurement, because of the natural imperfections on the road’s surface
which generates pooling of the water and hence the ice layers. In the measuring of the ice
thickness we scratched the ice surface down to the pavement. Although the thickness
measurement is not precise, we still could get a good estimation of the water and ice
layers thickness. In addition, we tested the sensor to check if it was able to detect the
presence of freezing point depressant chemicals and their concentration.
Snow measurements for naturally produced icy and snowy conditions have been
collected starting November 2003. These include measurements involving thin and thick
ice layers, and unpacked and packed snow layers on the road surface. Additionally, we
sprayed freezing point depressant chemicals, to observe whether there is a distinguishable
pattern or signature that can be detected. The results that we obtained will be shown later.
Several other data sets using real weather conditions were collected in December 2003,
and during January and February 2004. We also tested the system for different
concentration of freezing point depressant chemicals, namely salt brine and MAGIC, to
observe whether there is a distinguishable pattern or signature that can be detected.
Preliminary data analysis was conducted to compare the new results with those of the
data collected using dry ice during the summer 2003.
Baseline Identification.
In order to recognize the road’s weather condition, the identification process requires an
infrared (IR) baseline measurement be obtained to set the detector gain parameters based
on the user’s range and incidence angle requirements.
The baseline is also used, as a dry reference threshold needed by the general decision
algorithm in obtaining ice, water, and snow calibration. The sensor discriminates ice from
water based on laser energy absorption ratios, substantial change of the baseline
condition will decrease the identifying algorithms overall reliability.
Baseline stability measurements were taken during the testing. It was discovered that for
surface measurements taken at lower incidence angles, the baseline was less stable than
for those obtained at higher incidence angles, where the laser is more normal to the
surface. Also, seasonal changes in the IR reflectivity/absorption of the road surface have
been shown to causing a shift in the baseline, due to variation in surface roughness and
road contaminates. Though the amount of seasonal change should not affect the ice
detection capability, it can affect sensitivity. It is therefore recommended that the user
periodically update the system baseline readings, being most desirable to calibrate just
prior to each experiment. The unit can update the baseline either manually or
15
automatically. In the automatic mode, the program uses a feature that ensures dry surface
conditions exist when the new baseline measurement is taken.
Testing Results
Initial Testing
We start with initial measurements performed with the IRID sensor during the summer
2003. For every road condition, a set 12 of measurements were taken. These
measurements are plotted in Figure 10. This graph particularly shows the data collected
by the sensor for wet and icy conditions at the test point 2 (see Table 1). In this figure,
each of the four colors represents the response of the sensor for the combination of the
two frequencies and two S and P polarizations. The blue dots represent the response to
the road’s conditions for the 1840 nm frequency with S polarization, the Magenta dots
represent the 2875 nm frequency with S polarization, the Yellow dots indicate the
response to the 1840 nm frequency with P polarization, and the Teal dots is the response
to the 2875 nm frequency with P polarization, respectively.
In all graphs that follow, the vertical axis represents the signal level reflected by the road
(in Volts), while the horizontal axis represents time. On top of this graph, we specify the
actual road’s surface conditions. Notice that the sensor’s response changes drastically
depending on the particular situation. It is interesting to note that, as expected, the signal
level of the sensor for dry road is practically constant, indicated by the flat response in
this part of the graph. On the other hand, the response for the wet and icy road conditions
was not constant, perhaps due to evaporation and to ice melting.
The positive slope of the graph corresponding to the wet condition was perhaps due to the
natural evaporation of the water on the road. Thus the road becomes drier due to high
temperature, while the negative slope in the icy road response is due to the deicing
process of the road, i.e., ice is converting to water creating a wet road’s surface.
While gathering these initial measurements, the IRID laser pointer (see Fig 15) was
turned on and it seems it affected the sensor’s response when it was left on during the
testing. Thus all data obtained afterwards are taken while the laser pointer is turned off.
16
Dry, Wet, Icy Road Measurements
Dry
Damp
Wet
Water
Ice
Fig. 10 .- Summer Initial Measurements
The response of the IRID sensor to different ice thickness is shown in Figure 11. Again,
this graph shows the rapid deicing effect due the hot environmental conditions during the
summer. In addition, the sensor’s 1840 frequency, in both polarizations S & P (blue and
Yellow points) is more sensitive to the ice layers than the 2875 frequency (Magenta and
Teal points).
From these initial observations, we can say that the sensitivity of the sensor to changes on
the road surface is very good. However, this high sensitivity is affected by many factors.
For example, on windy days the movements generated on the bridge structure by the
wind currents affected the sensor’s measurements.
17
Icy Road Measurements
Fig. 11 .- Ice on Road Measurement Detail
In Figure 12, the response of the sensor is displayed for a case where we created two
layers of different thickness (approximately of 2 and 4 mm respectively) of Water, and
the two freezing point depressants Salt Brine, and MAGIC. As can be seen from the
graph, the sensor is able to differentiate between simple water on the road, and the two
different depressant types. Moreover it is able to detect different thickness of the ice
layers.
The above graphs displayed the behavior of IRID responses at a testing point on asphalt.
In Figure 13, we show a graph of the measurements we performed with the IRID sensor
for the cement road surface during the summer. In particular, this graph shows the data
collected by the sensor when for wet and icy conditions around test point 5 (see Table 1).
The data measurements shown in Fig. 13 corresponds to the response of the sensor to test
point 5, which is on the cement side of the road. Notice that the basic response of the
sensor does not changes for this kind of surface. These measurements were collected
during the fall, when the ambient temperature was not as hot as during the summer.
Because of this, the slopes in these curves are not as steep as the ones showed on Figures
10 & 11.
18
Response to Chemicals
For Different
Thickness
Water
Salt Brine
Magic
Fig. 12 .- Sensor’s Response to Deicing Chemicals
Fig. 13 .- Sensor’s Response for Cement Surface
19
The effect of deicing chemicals on the IRID response is also examined. In Figure 14, we
show the response of the sensor for naturally formed layers of snow and ice that were
collected on the month of November. Notice that the ambient temperature was not a
factor in this case, as it was during the fall and summer experiments; therefore we obtain
flat responses for snow and ice regions.
Ice
Snow
Snow
Fig. 14 .- IRID Ice/Snow Characterization
Problems Encountered
We encountered many hardware and software problems in this research effort. Soon after
the initial testing process, the whole IRID unit stopped functioning due to some electronic
component failure perhaps caused by lighting or an over-voltage from the power supply.
After contacting the manufacturer (IDI Company), they presented two options, repairing
the board/unit at the IDI location, or we could exchange the original ice detector with a
smaller unit that they recently developed, which does not require any software, using an
RS232 link. They also suggested providing a wireless link on the unit to send the serial
data directly to their computer provided the user is within 500 feet or so. They also
suggested signing a $5000 per year Maintenance agreement, in order for them to perform
the necessary changes and upgrades to the system, which at the present time has not been
signed. We decided to take the first option of repairing the unit, so it was shipped to IDI
for its repair. Repairing the IRID lasted the whole winter and Spring Semesters, and
consequently we did not have the full opportunity to work on this project until it was
repaired and installed again in May 2003. According to IDI, most of the electronics was
damaged, the computer board (where a burnt communication chip was found) had to be
replaced. The board was re-programmed and tested to check that it was fully functional.
20
However, testing the unit after we received it from IDI revealed that it still has software
problems. It turned out that among the three programs, only the Toolset is operational.
We tried unsuccessfully to contact IDI many times to fix these problems. Thus we have
been using IRID with much less than its full potential. Without full functionality, it
took more work to calibrate and locate the points on the pavement. Nonetheless, the
results obtained so far are promising.
Conclusions
Extracting pavement conditions is a difficult task due to many contaminants that exist on
the road. It is shown in this research that the IRID sensor is capable of distinguishing
various cases of rainy, snowy or icy road conditions. We also evaluated IRID sensitivity
for the presence and concentration of freezing point depressant chemicals on the roadway
surface. Most of the evaluations are accomplished using controlled conditions, while
some others used real weather conditions. Although IRID has shown to be effective
under controlled conditions, more testing of IRID is needed in real world environment. It
is our hope that this research will eventually result in significant improvement in
providing meaningful pavement condition information to motorists and transportation
professionals [9].
Future Work
We have evaluated IRID sensitivity for the presence and concentration of freezing point
depressant chemicals on the roadway surface. However, many other questions about
IRID remained unresolved. These include:
1. Sensitivity of IRID to distance, i.e., how far can IRID detect pavement conditions
reliably?
2. As most of the tasks in Phase I are accomplished using very controlled conditions,
it would be natural in Phase II to expand and test its effectiveness in real world
conditions.
3. The IRID is equipped with a camera (Fig 15), with pan/tilt capabilities that can be
used to show pictures of the environmental conditions of the different locations
near the pavement. We would like to use image-processing algorithms to enhance
the overall IRID data so that reliable information about weather conditions can be
obtained.
4. The remote access of IRID data is crucial for real time weather information, thus
automatic remote access of IRID is needed.
21
Fig. 15 .- Weather Conditions displayed by the IRID’s Camera
References
1. Scharching, Helmut: "Results of a Field Testing of Six Different Ice Warning
Systems", Proceedings of the International Workshop on Winter Road
Management, January 26-29, 1993 Sapporo, Japan, pp. 185 -199.
2. AURORA Group's "Standardized Testing Methodologies for Pavement Sensors"
dated December 7, 1999.
3. Man-Li C. Wu, "Determination of Cloud Ice Water Content and Geometrical
Thickness Using Microwave and Infrared Radiometric Measurements," Journal of
Applied Meteorology: Vol. 26, No.8, pp. 878884, 1988.
4. Tapkan, Baskin, " Active Microwave Remote Sensing of Road Surface
Conditions”, Proceeding of the Fourth International Symposium at Reno, Nevada,
August 11-16, 1996, pp. 73- 80.
5. 2002 Pacific Northwest Snowfighters, Snow and Ice Control Chemical Products
Specifications and Test Protocols For The PNS Association of British Columbia,
Idaho, Montana, Oregon and Washington, “Chemical Deicing Specifications”,
Washington State DOT, 2002.
6. Howari, Fares, “Reflectance Spectra of the Common Salts Prevalent in the Arid
Soils (320-2500 nm): Applications of Remote Sensing”, University of Texas El
Paso, Ph D Dissertation Proposal.
7. Innovative Dynamics, Inc, “Infrared Sensor for Pavement Ice Detection”, IDI
Technical Report, Ithaca New York, December 2001.
22
8. Sakk, Eric, et. al, “Remote Sensors for Pavement Ice Detection Phase II SBIR
Option”, Final Report, Innovative Dynamics, Ithaca New York, October 2001.
9. US Department of Transportation, FHWA, "Manual of Practice for an Effective
Anti-icing Program: A guide for Highway Winter Maintenance Personnel",
Publication No. FHWA-RD-95-202, June 1996.
23
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertising