Citation
Curb and gutter data management system for the city and county of Denver, Colorado

Material Information

Title:
Curb and gutter data management system for the city and county of Denver, Colorado
Creator:
Usagani, Robinson
Publication Date:
Language:
English
Physical Description:
xviii, 136 leaves : illustrations ; 28 cm

Subjects

Subjects / Keywords:
Streets -- Inventories -- Colorado -- Denver ( lcsh )
Street gutters -- Inventories -- Colorado -- Denver ( lcsh )
Curbs -- Inventories -- Colorado -- Denver ( lcsh )
Sidewalks -- Inventories -- Colorado -- Denver ( lcsh )
Database management -- Colorado -- Denver ( lcsh )
Curbs ( fast )
Database management ( fast )
Sidewalks ( fast )
Street gutters ( fast )
Streets ( fast )
Colorado -- Denver ( fast )
Genre:
Inventories. ( fast )
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )
Inventories ( fast )

Notes

Bibliography:
Includes bibliographical references (leaves 135-136).
General Note:
Department of Civil Engineering
Statement of Responsibility:
by Robinson Usagani.

Record Information

Source Institution:
|University of Colorado Denver
Holding Location:
|Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
57543313 ( OCLC )
ocm57543313
Classification:
LD1190.E53 2004m U72 ( lcc )

Full Text
CURB AND GUTTER DATA MANAGEMENT SYSTEM FOR
THE CITY AND COUNTY OF DENVER, COLORADO
by
Robinson Usagani
B.S., Colorado School of Mines, 2001
A thesis submitted to the
University of Colorado at Denver
in partial fulfillment
of the requirements for degree of
Master of Science
Civil Engineering
2004


This thesis for the Master of Science
degree by
Robinson Usagani
has been approved
by
CS Bruce Janson
0^ 2-00*1
Date


Usagani, Robinson Aloysius (M.S., Structural Engineering)
Curb and Gutter Data Management System for the City and County of Denver,
Colorado
Thesis directed by Associate Professor Kevin L. Rens
ABSTRACT
This thesis analyzes the results of a street subsystem inventory and condition
assessment for the City and County of Denver (CCD). Street subsystems include the
minor infrastructure associated with a pavement street such as inlets, cross-pans,
sidewalks, and curbs and gutters. While this research covers all street subsystem
components, the condition rating procedure presented includes only the CCD curb
and gutter network. While infrastructure management systems for bridges and
pavements are well established, it has only been recently that minor infrastructure
associated with street subsystems has become a priority. The main reason for this lies
in a new government regulation from the Governmental Accounting Standards Board
and in particular, statement called GASB34, which required an inventory and
assessment report for all minor infrastructures.
m


In early October 2001, University of Colorado at Denver Civil Engineering
Department was contracted by the CCD to inventory and assess curb and gutter, and
other associated street subsystems in the CCD. The data is managed by ArcGIS
software, which allows queries of the data. The main deliverable of this tudy is a
proposed curb and gutter condition rating that will help to establish future
deterioration predictions. This in turn will help administer financial resources for the
maintenance and rehabilitation of the CCD curb and gutter network.
This abstract accurately represents the content of the candidates thesis. I recommend
its publication.
Signed
IV


DEDICATION
I dedicate this thesis especially to my wife, Staci McComb, for her understanding and
constant support while I was writing this thesis.


ACKNOWLEDGEMENT
I would like to express my appreciation to:
Professor Kevin L. Rens, chair of my thesis committee, for his guidance
during these past two years.
City and County of Denver Public Works infrastructure management team for
their financial support and expertise in helping finish this thesis.
W. Patrick Kennedy, Senior Engineer and Benjamin Allen, Staff engineer for
the City and County of Denver for providing information and computer files.
Dan R. Roberts, Director of Street Maintenance for the City and County of
Denver for providing his expertise in pavement management systems.
Irena Kahanek, for her resourceful report that helped me describes all of the
features in the data collection of which I expanded.
Dr. Kevin Rens and Samuel Brown for spending tremendous amount of time
proof reading this thesis.
My parents, Ben and Johanna Usagani for their spiritual and financial
supporting in finishing my degree.


Several people who worked on the curb and gutter inspection along with my
self such as: Samuel Brown, Aaron Erfinan, Michael Doyle, Cade Caldwell,
Paul Boutry, and others.
Professor Bruce Janson and Professor Lynn Johnson, for being thesis
committee members.


CONTENTS
Figures.....................................................................xi
Tables....................................................................xvii
Chapter
1. Introduction..........................................................1
2. Objective.............................................................7
2.1 Governmental Accounting Standards Board...............................7
2.2 Rating System.........................................................8
2.3 Pavement Management System............................................9
2.4 Latest Technology Supporting the Inventory...........................11
2.4.1 Global Positioning System............................................12
2.4.2 Geographic Information System........................................13
2.4.3 Portable Computers...................................................14
2.4.4 Data Collecting Methods..............................................18
2.5 Street Subsystem Data Collecting Process.............................19
2.6 Description of Street Subsystems.....................................22
2.6.1 Point Features.......................................................23
2.6.1.1 Curb Ramps..........................................................23
vm


2.6.1.2 Inlets.............................................................. 27
2.6.1.3 Driveways and Alley Approaches...................................... 31
2.6.1.4 Curb and Gutter Distresses............................................32
2.6.2 Linear Features...................................................... 33
2.6.2.1 Sidewalks.............................................................33
2.6.2.2 Cross-pans.......................................................... 35
2.6.2.3 Curb and Gutters......................................................36
2.7 Preparing the Data for Rating System...................................45
3. Condition Rating System.............................................. 53
3.1 Condition Index........................................................55
3.1.1 Xmax and Weighting Factor..............................................59
3.1.2 Cl Adjustments for Curb and Gutter................................... 66
3.2 Simplified Condition Index......................................... 72
3.3 Distress Density Map...................................................73
3.4 Error Analysis of Data.................................................75
3.4.1 February 2002 (5th month analysis).....................................75
3.4.2 March 2004 (30th month analysis).......................................77
3.4.3 Error Analysis Summary.................................................79
3.5 Summary of Curb and Gutter Data Collection............................84
4. Deterioration Model of Curb and Gutter................................95
IX


4.1 Rehabilitation Action..............................................101
4.2 Cost Analysis......................................................103
5. Future Data Collection and Rating System...........................106
5.1 Proposed Changes for Future Data Collection........................106
5.1.1 Distress Priority..................................................106
5.1.2 Curbstone Length...................................................107
5.1.3 Gross Length and Net Length...................................... 108
5.1.4 Distress Collection................................................108
5.1.5 Distress Severity..................................................109
5.2 Future Data Format.................................................110
6. Summary, Conclusions, and Recommendation for
Further Work.......................................................113
6.1 Summary............................................................113
6.2 Conclusions........................................................115
6.3 Recommendation for Further Work....................................116
Appendix
A. Standard Drawings and Details......................................118
Bibliography.............................................................135
x


FIGURES
Figure
1.1 Overview of data collection in CCD.....................................3
1.2 Typical curb and gutter section in City and County of Denver. This example
shows a 6 vertical curb that has no overlay..........................4
1.3 UCD inspector performing the inventory of curb and gutter in CCD.......4
1.4 Miles of curb and gutter inventoried in CCD each month.................5
1.5 Cumulative miles of curb and gutter inventoried in CCD.................6
2.1 Variable lengths of curb and gutter sections located in the CCD........9
2.2 Illustration of GIS layers showing a distress layer such as cracks and spalls,
a curb, street and aerial photography layer...........................14
2.3 Leica GS20 using keypads for data entry...............................15
2.4 Compaq Ipaq uses touch screen for data entry..........................16
2.5 Vocarta System with Fujitsu Stylistic used by UCD with voice recognition
data entry software...................................................16
2.6 Laser Range Finder used in weak GPS signal area.......................17
2.7 Inspector on site.....................................................20
xi


2.8 Vocarta systems components used in CCD curb and gutter data collection....21
2.9 Data collection process flowchart............................................21
2.10 Illustration of point on feature to construct a curved inventory component
such as curb.................................................................22
2.11 An example of data collected using the backpack system on ArcMap............23
2.12 Ramp type comer with ramp color attribute red................................25
2.13 Ramp type trough with ramp color attribute gray..............................25
2.14 Ramp type walkover with ramp color attribute red.............................26
2.15 Ramp type stub out with ramp color attribute gray............................26
2.16 Ramp type none needs to be collected to ensure that indeed no ramp exists..27
2.17 Inlet type 14 that is 5 feet wide and condition good.......................28
2.18 Inlet type 16................................................................29
2.19 Inlet type stone top that is 3 feet wide.....................................30
2.20 Inlet type chase that is 2 feet wide.........................................30
2.21 Typical driveways in CCD with drainage condition.............................31
2.22 Illustrated definition of a curbstone in CCD.................................33
2.23 Concrete sidewalk and flagstone sidewalk in CCD..............................34
2.24 Typical cross-pans in CCD....................................................35
2.25 Six-inch vertical curb in CCD................................................37
2.26 Nine-inch CCD vertical curb .................................................38
2.27 Combo curb in CCD............................................................39
xii


2.28 Flagstone curb in the CCD.............................................39
2.29 Valley pan curb in CCD................................................40
2.30 Curb condition major crack.......................................... 41
2.31 Curb condition spalls.................................................41
2.32 Club condition curb gone..............................................42
2.33 Curb condition heaved.................................................42
2.34 Curb condition settled................................................43
2.35 Curb condition drainage...............................................43
2.36 Curb condition undermined.............................................44
2.37 Multiple columns that can be matched for a rating system..............47
3.1 Aerial view of a portion of the CCD with curb and gutter data showing up
as occurrences.......................................................53
3.2 Distresses on curb with a rating system that can be interpreted by user
much easier when compared to Figure 3.1..............................55
3.3 Condition index as related to X (a measurement or distress occurrence) /
Xmax (limiting value of the distress)................................61
3.4 Adjustment factor for weighting.......................................65
3.5 Multiple distresses on a curbstone....................................67
3.6 Flow chart of overall Cl (OCI) calculation............................71
3.7 Sample of Cl ArcMap output............................................72
3.8 Sample of SCI ArcMap output...........................................73
xm


3.9 Typical area in CCD that can trigger false distress density outputs due to a
sparse population of curb and gutter inventory............................74
3.10 Example of distress density map for curb and gutter in CCD.................75
3.11 Number of distresses result in error analysis study........................76
3.12 Number of distresses result in error analysis study in March 2004..........78
3.13 Major crack that appears to look like a curbstone line.....................80
3.14 Distress that is covered with vegetation................................. 81
3.15 Minor distress that can cause inconsistency in error analysis..............82
3.16 Deposits on a gutter pan that may result in data inconsistency.............83
3.17 Length of different type of curbs..........................................86
3.18 Percentage of different type of curbs......................................86
3.19 Different type curbs in CCD................................................87
3.20 View of Cl in CCD..........................................................88
3.21 View of SCI in CCD.........................................................89
3.22 View of Cl in CCD with Jenks optimization..................................91
3.23 View of SCI in CCD with Jenks optimization.................................92
3.24 Sub-division distress density with Jenks optimization......................93
4.1 OCI and SCI deterioration curve for combined curbs.........................98
4.2 OCI and SCI deterioration curve for 6 vertical curb.......................99
4.3 OCI and SCI deterioration curve for combo curb............................100
xiv


4.4 Rehabilitation alternatives effects to the remaining life of curb and gutter.. .102
4.5 Repair cost in 2004 USD for repairing entire damaged curbstones........106
5.1 Distress that may be joined by Arc Map into incorrect curb due to GPS
error.................................................................109
5.2 Future data collection flowchart.......................................112
A.l Typical street sections................................................119
A.2 Typical curb sections........;........................................120
A.3 Typical curb, gutter pan and sidewalk..................................121
A.4 Standard residential driveway..........................................122
A.5 Standard commercial and multi family driveway..........................123
A.6 Curb ramp notes and typical section....................................124
A.7 Typical curb ramp type trough (type 1) and type comer (type 2).........125
A.8 Different type of ramp type trough.....................................126
A.9 Curb ramp type trough in the downtown area.............................127
A. 10 Curb ramp type walkover in the downtown area...........................128
A. 11 Cross pan..............................................................129
A. 12 Inlet type chase.......................................................130
A. 13 Typical alley layout...................................................131
A. 14 Alley approach with attached sidewalk..................................132
A. 15 Alley approach with detached sidewalk..................................133
xv


A. 16 Typical alley cross-section


TABLES
Table
2.1 Advantages and disadvantages in different input systems...................18
2.2 Curb ramps attributes in the data collection..............................24
2.3 Inlet attributes in the data collection...................................28
2.4 Driveway and alley attributes in the data collection......................32
2.5 Sidewalk attributes in the data collection................................34
2.6 Cross-pans attributes in the data collection..............................35
2.7 Curb and gutter attributes in the data collection.........................36
2.8 Distress data after spatial join..........................................49
2.9 Adding the number of distresses...........................................51
3.1 Condition index zones.....................................................58
3.2 Example of overall Cl.....................................................59
3.3 Xmax data as obtained from 6 experts.................................... 60
3.4 Weighting factors as obtained from 6 experts..............................62
3.5 Average weighting factor and used for any type of curb....................64
3.6 Example of weighting factor adjustments...................................66
3.7 Adjustment factor of 0 modification when a distress does not exist........68
xvii


3.8 Second overall Cl calculation by treating all distress if they were the
same type..........................................................69
3.9 Example of weighting factor adjustments combined....................70
3.10 Data comparison result in March 2004 error analysis.................78
3.11 Curb type population used in the research and extrapolated population
in CCD........................................................... 85
3.12 Mileage of curb and gutter condition in CCD.........................94
5.1 Priority level for future data collection..........................107
5.2 Example of future data format......................................Ill
xvm


1. Introduction
During October of 2001, the City and County of Denver (CCD) contracted the
University of Colorado at Denver (UCD) to develop an inventory and condition
assessment of their street subsystem assets. Included in these assets are curb and
gutter pans, sidewalks, ramps, cross-pans, and inlets. These data are managed by the
CCDs Geographic Information Systems (GIS) database network. While this thesis
will present the half of the inventory, only curb and gutter is focused on for condition
prediction and deterioration modeling. Figure 1.1 shows the subsystem data collected
in CCD such as curbs, sidewalks, inlets, curb ramps and cross-pans.
A curb is defined as an edging usually made of concrete built along a street to form
part of a gutter pan. A gutter is defined as a low area such as at the edge of a street to
carry off surface runoff water to locations such as storm sewers. Figure 1.2 shows a
typical 6 vertical curb and gutter section located in the CCD. Curbs and gutters are
minor infrastructure that is part of the major street infrastructure system and are an
integral part of the transportation system. Curb and gutter pans are used to channel
storm drainage, stabilize the roadways, represent a demarcation point for where the
street ends, and give the roadways an aesthetic appearance.
It is estimated that approximately 3300 linear miles of curb and gutter exist in the
CCD. Although the data collection process is somewhat tedious, relatively good
1


progress has been made since the projects inception in 2001. For example, the
average amount of data collected every month continually averages around 100 miles
per month. The cost to the CCD in 2004 US dollars is around $100 per mile. The
UCD Department of Civil Engineering research team physically walks these curbs
and gutters and collects data as shown in Figure 1.3. Detail of the data collection
process and equipment will be detailed in later chapters. Figure 1.4 shows the
amount of monthly mileage collected by UCD. Mileage variances each month are
due to weather and equipment related problems. The cumulative mileage of the data
collected is shown in Figure 1.5. At the time of this thesis publication approximately
50% of the street subsystem inventory had been completed. Using GIS technology in
combination with the data collected, the CCD will be able to query the condition of
curb and gutter in CCDs jurisdiction. This thesis addresses the following goals:
Chapter Two discusses some of the latest technology suitable for curb and
gutter inventory and condition assessment collection.
Chapter Three contains the condition rating system and thorough summaries
of the curb and gutter inventory and assessment data collected by the UCD
research team.
Chapter Four contains a proposed deterioration model of curb and gutter
Chapter Five contains a proposal for future data collection
Chapter Six presents conclusions and recommendations for future studies.
2


The CCD curb and gutter data collected only shows the curb lines and locations of the
distresses. A goal of this thesis is to improve this curb and gutter data and to study
deterioration trends by developing a rating system. Such a system will be a powerful
tool for the CCD infrastructure management office to augment decision-making
regarding the curb and gutter maintenance.
3


Figure 1.2 Typical curb and gutter section in City and County of Denver. This
example shows a 6 vertical curb that has no overlay.
Figure 1.3 UCD inspector performing the inventory of curb and gutter in CCD
4


Curb and Gutter Inventory
Monthly Milage Totals
# # # # # ^ ^ ^ ^ ^ ^ rf ^ ^ ^ rf rf ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ,
o^yv^A^V ///VV off o*VV /y ^ ^ ^ ^ ^ #
Month
Figure 1.4 Miles of curb and gutter inventoried in CCD each month.


Curb and Gutter Inventory
Cumulative Miles Inventoried
2500
2000 -
1500 -
1000 -
500
2,177
///////////////// # # / / / / / / / f fjf
////// ^vvvvv^vv ,yy //// / ^vy ^
Month
Figure 1.5 Cumulative miles of curb and gutter inventoried in CCD,


2. Objective
The basic need for this research lies in the recently formed Governmental Accounting
Standards Board. In addition, the need to fully develop rating schemes for both major
and minor CCD infrastructure was a priority. Each of these objectives is discussed in
the following sections.
2.1 Governmental Accounting Standards Board
Major infrastructure such as bridges, dams, and pavements have been inventoried and
assessed for years. In 1967, Silver Bridge connecting Point Pleasant, West Virginia,
and Gallipolis, Ohio, failed killing 46 people. The federal government required the
Secretary of Transportation to apply National Bridge Inspection Standards to all
bridges over 20 feet. This action was a major benchmark in the development of
Bridge Management Systems (BMS). [Transportation Research Circular, 1999]. A
large amount of research has been completed in BMS. For instance, Carnot L.
Nogueria researched on application of nondestructive evaluation techniques in bridge
inspections. [Nogueria, 1998]. Other past research mainly involved different types
of inspection and data collection schemes, in addition to deterioration models and life
cycle cost analysis, etc. [Transportation Research Circular, 1999]. However,
recently, minor infrastructure such as curbs, sidewalks, cross-pans, curb ramps, and
inlets have become a priority to be assessed and inventoried. One of the main reasons
7


for the assessment and inventory of minor infrastructure such as curb and gutter is to
meet the new Governmental Accounting Standards Board (GASB) regulation. GASB
was organized by the Financial Accounting Foundation (FAF) in 1984 to set
standards of financial accounting and reporting for state and local governments. In
June 1999, GASB issued Statement No. 34 for basic financial statements and
management discussion and analysis for State and Local Governments. The objective
of this statement is to improve the financial reports of state and local governments to
the public, legislative, investors, and creditors. [Governmental Accounting Standards
Series, 1999]. Therefore starting in 1999, minor infrastructure inventory and
assessment have become popular.
2.2 Rating System
Curbs and gutters vary in length and condition. Any system developed must be able
to rate these curbs in a scale of 1-10 consistently using an inspectors subjective
judgment. Figure 2.1 shows the variable lengths of curb and gutter sections
encountered in CCD. By collecting only the number of distresses, a rating system can
determine the condition of these curbs and gutters more consistently on a scale of
100. To achieve consistency among inspectors, a detailed curb and gutter data
collection training must be Completed along with an easily understood user manual.
This thesis is intended to deliver these rating system essentials.
8


Figure 2.1 Variable lengths of curb and gutter sections located in the CCD. Note that
any rating system must take into account the long, medium, and short street sections.
2.3 Pavement Management System
The dynamic growth of City and County of Denver has formed a vital need to
objectively manage the roadway inventory. The high number of new streets being
constructed adds to the traditional challenges of maintaining the serviceability of
existing roads and streets within their design life in addition to reconstructing older,
failing streets.
A Pavement Management System can be defined as a set of all activities and tools
that provide cost effective and optimum management strategy plans for decision
makers at all levels by using all available and reliable information. [Sathantip, 2002]
9


The citys Pavement Management System helps maintain an accurate inventory of
streets, tracks pavement deterioration, diagnoses the cause of the deterioration and
evaluates design solutions. This system allows objective determination of pavement
maintenance and rehabilitation strategies to be used to maintain and, in some
instances, even improve and extend the performance life of roadways. By using
effective maintenance rehabilitation methods, the City and County of Denvers
roadways can provide higher levels of service for longer periods of time, resulting in
direct, immediate savings to both the city and the motoring public.
The main goal of the Pavement Management System is preventive maintenance as
opposed to complaint driven maintenance. Complaint driven maintenance require the
public to report any defected infrastructure. Most of the maintenance of curb and
gutter in CCD has been driven by complaint driven maintenance. The city receives
several complaints regularly. The city sends their staff to review the complaint and
decide whether repair is needed or not. With the data inventory and assessment of
curb and gutter, the city is trying to use more preventative maintenance than
complaint driven maintenance. Preventative maintenance treatments in pavement
include crack sealants, repair failed curb and gutter curbstones, patch failed sections
of asphalt such as pot holes, rutting, etc, and to halt oxidation aging which results in
hardening of the asphalt layer. By inspecting each street, making timely repairs and
10


resurfacing the pavement surface while repairing damaged curb and gutter sections,
one can lengthen the life of the street infrastructure.
The strategy of preventative maintenance work is to recognize and carry out the
suitable work at the proper time and thus preventing the asset from falling into the
next more expensive level of repair. In addition all new construction, reconstruction,
rehabilitation and maintenance projects should employ some level of economic
evaluation to determine the most cost effective method and timing. Preventing the
next level of repair minimizes major interruption to the traffic when the repairs are
eventually performed. Most essential to preventive maintenance is that it is fairly
inexpensive, improves the pavement condition, reduces the deterioration rate and
extends the longevity of the streets and curb and gutter. [The City of Las Vegas,
2004].
2.4 Latest Technology Supporting the Inventory
Inventory of minor infrastructure is fairly complex considering the size of the
population. However, advancement of different technologies significantly aids the
inventory process. Each of these technologies is discussed in the following sections.
11


2.4.1 Global Positioning System
The U.S. Department of Defense and Ivan Getting invented Global Positioning
System (GPS), at the cost of twelve billion dollars. [Beyond Discovery, 2003]. GPS is
a navigational system that uses satellites as reference points to calculate geographical
positions, accurate to a matter of meters and in some cases, centimeters. GPS has
been used to pinpoint friendly or hostile ships and submarines in the oceans. It was
originally invented for military purposes only [Beyond Discovery, 2003]. In fact,
GPS was not available for civilian usage because the U.S. was concerned that GPS
could be used to aid smugglers, terrorist and hostile forces. GPS receivers have been
miniaturized to just a few integrated circuits, becoming very small and economical.
Today, GPS can be found in cars, boats, planes, construction equipment, and even
laptop computers, personal data assistants, and cell phones.
GPS plays an important role in spatial data collection in Geographic Information
Systems. It uses satellite technology to quickly determine precise latitude, longitude,
and sea level information at any time and in most weather and terrain conditions.
However, GPS does occasionally have errors due to several factors. The most
common source of error is caused by atmospheric and ionospheric delays. The range
distance between the satellite and the GPS unit is determined by multiplying the
speed of the GPS signal (speed of light) by the time the signal takes to travel from
satellite to the receiver. It is assumed that the speed of light is constant. However,
12


just like sound, a GPS signal travels at different rates in different mediums.
Atmospheric density variation between the satellite and the GPS unit is highly
variable, causing errors in positional information. There is however ways to increase
the accuracy of GPS called differential correction. Differential correction is
necessary to get accuracies within 1-5 meters, or even better, with advanced
equipment. Differential correction requires a second GPS receiver, a base station,
collecting data at a stationary position at a precisely known point. Because the
physical location of the base station is known, a correction factor can be computed by
comparing the known location with the GPS location determined by using the
satellites [Bolstad, 2002].
2.4.2 Geographic Information System
For centuries humans have studied the world by using models such as globes and
maps. When computers became more commonly used, people started to put these
models into computers. These computer models plus the computers capability in
manipulating data make up a Geographic Information System. With the right data,
GIS can create a variety of maps. Using ArcMap, common GIS software, maps can
be simplified without losing all of the important features like on regular paper maps.
The GIS software can add multiple layers such as the: sidewalk layer, inlet layer, or
ramp layer to the curb and gutter layer and these can be turned on and off as each
layer as needed as shown in Figure 2.2.
13


Figure 2.2 Illustration of GIS layers showing a distress layer such as cracks and
spalls, a curb, street, and aerial photography layer.
2.4.3 Portable Computers
Prior to this research, the CCD used paper inspection forms to collect field data (hard
copy). For example in the year of2000, the UCD research team collected condition
inventory data on its 5,000 alleys using inspection forms [Sathantip, 2002]. These
data are later manually inputted into the database at the office. This method can create
errors caused by ineligible writing and data inputting error when the data have to be
stored into the database. On the other hand, double checking the data by manual
entry does provide a measure of checks and balances as inspectors were required to
enter data the same day as it was obtained. The computer and GIS software can work
14


more easily with GPS in collecting and processing the data. On the other hand, this
equipment has a relatively expensive first cost ($30,000) and can be relatively
expensive to maintain should hardware or software problems arise. Hardware failures
have occasionally resulted in loss or error of data. Therefore, the data has had to be
collected again. Instead of collecting the data from the GPS and manually entering
the data into the computer, a computer integrated GPS system can bypass the data
entry process. By integrating the computer to the GPS, the computer will record the
position of the data automatically in a digital format, ready to be downloaded into a
database. Figures 2.3 through 2.5 illustrate various types of GIS data collection
hardware. Each of these computers uses a different type of data entry methodology.
Figure 2.3 uses a keypad while Figure 2.4 uses a touch screen. Figure 2.5 uses an
integrated voice recognition system called Vocarta. The advantages and
disadvantages of these different data entry methods will be discussed in the next
section.
Figure 2.3 Leica GS20 using keypads for data entry.
15


Figure 2.4 Compaq Ipaq uses touch screen for data entry
Figure 2.5 Vocarta System with Fujitsu Stylistic used by UCD with voice recognition
data entry software.
In some situations, it is not possible to use the conventional GPS in areas where
adequate satellite vision is compromised. Such areas include downtown areas where
16


large buildings exist or in areas of leafy vegetation where a canopy of trees exist.
Therefore, most off the shelf GPS receivers shown in Figures 2.3 through 2.5 cannot
receive the required number of satellite signals. In 2003, a pilot study was completed
in assessment and inventory of minor infrastructure in Areas with Weak GPS Signals
[Kahanek, 2003]. To solve this problem, another device has to be used along a GPS
data collector unit called the Impulse 200LR- Laser Rangefinder shown in Figure 2.6.
A fairly accurate GPS location can be determined in these problem areas by shooting
the laser from an area where GPS signal is not weak, into a prism located at the spot
of the desired location. The maximum distance from the device to the prism is
approximately 1886 ft. The device added error of .1 ft in addition to the existing error
of the GPS data collector. [Kahanek, 2003].
Figure 2.6 Laser Range Finder used in weak GPS signal area
17


2.4.4 Data Collecting Methods
There are several ways spatial data can be collected. The simplest way is by using
keypads similar to that shown in Figure 2.3. The most common feature in digital data
collection used today is the touch screen as pictured in Figure 2.4. The equipment
used by the UCD research team does have a touch screen, but it also has voice
activated data collection options. UCD uses a voice recognition program in the data
collection process. The advantages and disadvantages of each method are listed in
Table 2.1.
Method Advantages Disadvantages
Key Pad Quicker computer response Gets too complicated for more data entry choices
Stylus Pen Similar to mouse Easy to use Slower data entry
Voice Recognition Quick Operator can pay attention more to the surrounding rather then the computer Takes a lot of practice to speak to the computer properly Very hard to use in a noisy/windy area
Pen and Paper Very simple Uses too many pages of paper for high number of inspections Error in ineligible writings Time consuming
Table 2.1 Advantages and disadvantages in different input systems.
18


2.5 Street Subsystem Data Collecting Process
One to three data collector GPS units (shown in Figure 2.5) along with two to six
inspectors are dispatched regularly to collect CCD street subsystem assets. For
safety reasons, it is preferable to use 2 inspectors for each backpack unit. Using the
integrated backpack equipment an inspector physically walks a curb section and
collects data by speaking to the computer as shown in Figure 2.7. The equipment
consists of a backpack, portable laptop computer, GPS receiver, and headset as
detailed in Figure 2.8.
Using software developed by Datria System Inc. an inspector can speak commands to
record the citys street subsystems asset locations and condition. The voice
recognition software enables the inspector to operate the computer without having the
need to use a keyboard or mouse. Every time the inspector speaks a command, to
verify that the computer acknowledged the correct command, the system will repeat it
to the user for self-verification. If in error, the user has an opportunity for correction.
The data recorded by the equipment can then be downloaded into the CCD main GIS
database. This process still requires manual data entry for the location attribute. If
there is an error or incomplete data, most of the time the error can be manually fixed
during the download process without having to recollect the data.
19


CCD GIS database contains other GIS data such as other infrastructure, aerial
photographs, quarter mile map, subdivision map, bus routes, pipe network, etc. These
data will be useful to the city financial report and research. All these processes in this
section are shown in the flowchart in Figure 2.9.
Figure 2.7 Inspector on site
20


Datria Systems: Setting Up a Backpack
Equipmonl:
01 Vocarta Backpack with Frames for equipment
02 GPS power cable
03 Fujitsu 3400 with touch screen
04 Trimble GPS receiver
05 GPS batteries
06 Trimble GPS antenna
07 Trimble GPS coaxial cable
08 Headset
Figure 2.8 Vocarta systems components used in CCD curb and gutter data collection.
Figure 2.9 Data collection process flowchart
21


2.6 Description of Street Subsystems
The street subsystems inventory data are divided into two types, linear features and
point features. Linear features include such infrastructure as curbs, sidewalk and
cross-pans, while point features include distress locations on linear features in
addition to inlets, curb ramps, alley approaches, and driveways. Linear features can
be collected by locating the two ends of the subsystem. If there is a radius of
curvature on a curb, several curb points need to be recorded to create an arch because
the system basically connects dots, or point locations, to form line features, as shown
in Figure 2.10. The red dots shown in Figure 2.10 indicate point on features used in
data collection. Point features are used to locate curb ramps, inlets, driveways, alley
approaches and also curb distresses. A sample of data recorded is shown in Figure
2.11.
Figure 2.10 Illustration of point on feature to construct a curved inventory
component such as curb
22


Figure 2.11 An example of data collected using the backpack system on ArcMap
(blue lines represent curb linear features while the dots represent distresses such as
major cracks, spalls, settled, heaved, curb gone, drainage, and undermined)
2.6.1 Point Features
The following sections explain the street subsystem components that are considered
as point features in this research. A point feature shows unique location of certain
street subsystem and does not define length. These point features include curb ramps,
inlets, driveways, alley approaches, and curb distresses.
2.6.1.1 Curb Ramps
To comply with American with Disabilities Acts, when streets and roads are newly
constructed, altered, or repaired they must have ramps wherever there are curbs or
other barriers to entry from a pedestrian crossing. Also, when new sidewalk or
walkways are being built or altered, they also must have ramps when they intersect a
23


street. With this curb ramp data collection, the city can plan on ramp installations
when street resurfacing or curb repairs are completed. [ADA, 2004]. All attributes of
curb ramps in the data collection process are shown in Table 2.2. When a curb ramp
contains certain types of anomalies or distresses listed in the table, they need to be
spoken to the computer.
VOICE CURB RAMP } %. 'l, v COLOR CONDITION
CORNER GRAY GOOD
THROUGH RED DRAINAGE
WALKOVER OTHER MAJOR CRACKS
STUB OUT HEAVED
NONE SETTLED
SPALLS
UNDERMINED
Table 2.2 Curb ramps attributes in the data collection, bold-faced commands are
default attributes that do not need to be spoken during the data collection process.
There are four types of curb ramp in the CCD. A set of voice commands is included
in the description of each curb ramp type.
24


Comer: this type of curb ramp is usually installed when two sidewalks that
both are attached to the curb meet together in the comer of a block as is shown
in Figure 2.12. An example of a typical data collection command procedure
shown in Figure 2.12 includes Ramp type comer, Ramp color red. Due to
the condition of the ramp is good, the condition command is not necessary
as it defaults to good.
Figure 2.12 Ramp type comer with ramp color attribute red
Through: this type of curb is also called directional curb ramp. It is
commonly used for a detached sidewalk shown in Figure 2.13. An example
of a typical data collection command procedure shown in Figure 2.13 includes
Ramp type trough. Due to the color of the ramp is gray, the color
command is not necessary as it defaults to gray.
Figure 2.13 Ramp type trough with ramp color attribute gray.
25


Walkover: Walkovers exist when the surface of sidewalk and curb flushed
with the street surface in the comer shown in Figure 2.14. An example of a
typical data collection command procedure shown in Figure 2.14 includes
Ramp type walkover, Ramp color red.

Figure 2.14 Ramp type walkover with ramp color attribute red.
Stub out: a ramp that is constructed before the sidewalk is constructed. A
sidewalk may or may not be built in the future like shown in Figure 2.15. An
example of a typical data collection command procedure shown in Figure 2.15
includes Ramp type stub out.
Figure 2.15 Ramp type stub out with ramp color attribute gray.
26


None: when no ramps exist in the comer of a block as shown in Figure 2.16,
ramp type none has to be spoken to make it a positive command that indeed
no ramp exists. An example of a typical data collection command procedure
shown in Figure 2.16 includes only Ramp type none.
Figure 2.16 Ramp type none needs to he collected to ensure that indeed no ramp
exists.
2.6.1.2 Inlets
Inlet structures collect water from the street surface to the storm system network.
Inlet attributes include inlet type, inlet size, and inlet condition. All attributes in the
inlet data collection are shown in Table 2.3. When an inlet is fairly damaged, the
inlet condition command of drainage, cracking, heaved, settled, spalls, or undermined
needs to be spoken.
27



14 1-25 FEET GOOD
16 DRAINAGE
STONE TOP MAJOR CRACKS
CHASE HEAVED
OTHER SETTLED
SPALLS
UNDERMINED
Table 2.3 Inlet attributes in the data collection. Commands in bold are default
attributes that do not need to be spoken.
There are four types of inlets in the CCD street subsystem network. A set of voice
commands is included in the description of each of the inlet types in the following
sections:
14: this type of inlet has an opening along the vertical side of the curb (throat)
as is shown in Figure 2.17. An example of a typical data collection command
procedure shown in Figure 2.17 includes Inlet type 14 and Inlet size 5
feet.
Figure 2.17 Inlet typel4 that is 5 feet wide and condition good.
28


16: This type of inlet contains metal vents. In some cases it also has throat
similar to inlet type 14 as shown in Figure 2.18. An example of a typical data
collection command procedure shown in Figure 2.18 includes Inlet type 16
and Inlet size three feet.
Figure 2.18 Inlet type 16 (left: with a throat, right: without a throat)
Stone top: This type of inlet can usually be found in the older neighborhood
areas surrounding the downtown Denver area. It is essentially similar to type
14, but it is covered with natural flagstone instead of concrete as shown in
Figure 2.19. These inlets are very durable and have been in service for long
period of time. An example of a typical data collection command procedure
shown in Figure 2.19 includes Inlet type stone top and Inlet size three
foot. The inlet size describes the opening of the throat as opposed to the
width of the stone.
29


Figure 2.19 Inlet type stone top that is 3 feet wide.
Chase: This type of inlet usually transfers rainwater from a roof to the gutter
pan as is shown in Figure 2.20. An example of a typical data collection
command procedure shown in Figure 2.20 includes Inlet type chase and
Inlet width 2 feet.
Figure 2.20 Inlet type chase that is 2 feet wide.
Other: This feature is to cover all other type of inlets that are not covered in
the inlet type list above.
30


2.6.1.3 Driveways and Alley Approaches
Driveways and alley approaches are similar to ramps. They connect the street with
residential driveways as is shown in Figure 2.21, as well as business parking lots and
alleys. The drive ways and alley approaches data are very essential to calculate the
net length of the curb and gutter for the curb and gutter management system. The list
of attributes for driveways and alley approaches are as shown in Table 2.4. An
example of a typical data collection command procedure shown in Figure 2.21
includes: Driveway, Driveway size twelve foot and Driveway condition
drainage.
31


pipIP^IS wsitismBsm ipMipgpii CONDITION
DRIVEWAY 1-199 FEET GOOD
ALLEY DRAINAGE
MAJOR CRACKS
HEAVED
SETTLED
SPALLS
UNDERMINED
Table 2.4 Driveway and alley attributes in the data collection where boldfaced
commands are default attributes that do not need to be spoken.
2.6.1.4 Curb and Gutter Distresses
The last type of point feature is the curb and gutter distress list. Because curbs and
gutters are fairly long linear features, having only one condition attribute entry is not
sufficient enough. The research team decided that only one distress could be
recorded on each curbstone shown in Figure 2.22, therefore in a case where there are
two types of distresses on a curbstone, the worst type of the two is recorded. By
doing this, the city can estimate the number of damaged curbstones. Although curb
and gutter distress is a point feature and the data is separated from the curb and gutter
data, the explanation of the distresses will be discussed in section 2.6.2.3 along with a
detailed curb and gutter explanation.
32


2.6.2 Linear Features
The following sections explain the street subsystems commands that are considered
as linear features. A linear feature includes a unique starting and ending location of
certain street subsystem infrastructure such as curbs, gutter pans, sidewalks, and
cross-pans. ArcMap GIS will automatically calculate the length of these features.
2.6.2.1 Sidewalks
Sidewalks encourage pedestrians to walk safely separated from the street traffic and
encourage no walking on grassed areas. There are four types of sidewalks of which
depend on the type of construction material:
Concrete
Flagstone
33


Asphalt
None
A positive command is utilized when no sidewalk exists at a particular location. The
none sidewalk command needs to be spoken when sidewalk does not exist.
Sidewalk conditions are not collected. Examples of concrete and flagstone type
sidewalks are shown in Figure 2.23. The list of attributes for sidewalk is listed in
Table 2.5.
Figure 2.23 Concrete sidewalk and flagstone sidewalk in CCD.
f /^lDEV^LK, TyiPE'A - >:STDEWALK "CBETAqtlElSifSi v'V 'applicable^;. : .
CONCRETE 2-30 FEET 0-20 FEET
FLAGSTONE 5 ft. is default 0 is default
ASPHALT
NONE
Table 2.5 Sidewalk attributes in the data collection
34


1.6.2.2 Cross-pans
Cross-pans transfer storm water from one gutter to another gutter across a street as
shown in Figure 2.24. Cross-pans are divided into 4 types depending on its size. The
list of attributes for cross-pans is shown in Table 2.6.

4 FOOT GOOD
6 FOOT DRAINAGE
8 FOOT HEAVED
10 FOOT SETTLED
SPALLS
UNDERMINED
MAJOR CRACKS
Table 2.6 Cross-pans attributes in the data collection.


Figure 2.24 Typical cross-pans in CCD. The list of commands to collect the cross-
pan in this figure is as follows: Begin 6 foot cross-pan (at A), Cross-pan condition
major crack, Finish 6 foot cross-pan (at B).
35


2.6.23 Curb and Gutters
Curbs and gutter pans direct water from the street surface into downstream storm
water inlets. Unlike other subsystems, the condition of curbs and gutter at distinct
points can be spoken more than once for each curb length. Each condition represents
only one 8 ft to 12 ft curbstone. Therefore, as was discussed in chapter 1, only one
condition can be assigned to each curbstone. All attributes corresponding to curb and
gutter in the data collection process are shown in Table 2.7.
!iipKPSi iftisiasffn BII11S1IIIISS |WSB||8iSl #ipsspiii SpPlIi
6 VERTICAL GOOD GOOD 1-8 INCHES
9 VERTICAL DRAINAGE EROSION
FLAGSTONE HEAVED
COMBO SETTLED
MOUNTABLE SPALLS
NONE UNDERMINED
VALLEY PAN CURB GONE
ASPHALT MAJOR CRACKS
Table 2.7 Curb and gutter attributes in the data collection.
Curb and gutter data are classified into several different types. These types consist
of: six- inch vertical, nine- inch vertical, combo, valley pan, flagstone, mountable,
36


and asphalt types of curb. When no curb exists, the inspector is required to positively
speak the command of none curb. Illustrations of each curb and gutter types are as
follows:
A six-inch vertical curb has approximately 2 feet of horizontal concrete gutter
pan with a six-inch high curb face. An example is shown in Figure 2.25. The
list of commands to collect the curb and gutter in this figure is as follows:
Begin six-inch vertical curb (at one end of the curb), Finish six-inch
vertical curb (at the other end of the curb). The condition of each curbstone
needs to be collected between the two ends of the curb.
Figure 2.25 Six-inch vertical curb in CCD.
A nine-inch vertical curb is similar to a six-inch vertical curb, but it has a
higher curb face and the gutter pan is usually paved over with asphalt. The
condition of the overlaid asphalt and the curb exposure need to be collected.
An example is shown in Figure 2.26. The list of commands to collect the curb
37


and gutter in this figure is as follows: Begin nine-inch vertical curb (at one
end of the curb), Overlaid erosion, Curb exposure 5 inches, Finish nine-
inch vertical curb (at the other end of the curb). The condition of each
curbstone needs to be collected between the two ends of the curb.
Figure 2.26 Nine-inch CCD vertical curb
A combo curb type is the most common curb in the CCD. Combo curbs can
mostly be found in neighborhood areas. This type of curb is usually installed
when the layout of the driveways for homes are not known. It also serves as
sidewalk with a width ranging from 3 to 5 feet. An example is shown in
Figure 2.27. The list of commands to collect die curb and gutter in this figure
is as follows: Begin combo curb (at one end of the curb), Finish combo
curb (at the other end of the curb). The condition of each curbstone needs to
be collected between the two ends of the curb.
38


Figure 2.27 Combo curb in CCD
Flagstone curb is fairly rare and can be found in the older Denver downtown
neighborhood areas. These curbs were mostly installed in the late 1800s and
early 1900s. An example is shown in Figure 2.28. The list of commands to
collect the curb and gutter in this figure is as follows: Begin flagstone curb
(at one end of the curb), Overlaid erosion, Curb exposure 5 inches,
Finish flagstone curb (at the other end of the curb). The condition of each
curbstone needs to be collected between the two ends of the curb.
Figure 2.28 Flagstone curb in the CCD
39


Valley pan curb types can mostly be found in the industrial areas. This type
of curb serves the purpose of a gutter and allows industrial trucks the ability to
drive over this type of curb thus limiting the damage incurred. An example is
shown in Figure 2.29. The list of commands to collect the curb and gutter in
this figure is as follows: Begin valley-pan curb (at one end of the curb),
Point on valley-pan curb (at multiple location as discussed in section 2.6),
Finish valley-pan curb (at the other end of the curb). The condition of each
curbstone needs to be collected between the two ends of the curb.
Figure 2.29 Valley pan curb in CCD
There are several types of distresses associated with curb and gutter pans, which were
determined by joint meetings between UCD and the CCD. These distresses include
major cracks, spalls, curb gone, heaved, settled, drainage and undermined. The
descriptions of these distresses are as follows.
40


Major cracks are fractures on concrete curbs excluding hairlines. An example
is shown in Figure 2.30.
Figure 2.30 Curb condition major crack
Spalls are when a portion of the curb larger then 3 inches diameter has been
chipped or fragmented. An example is shown in Figure 2.31.
Figure 2.31 Curb condition spalls
41


Curb gone is when a curb should be and once was in that location but no
longer exists. An example is shown in Figure 2.32.
Figure 2.32 Curb condition curb gone
Heaved is when the curb is uplifted more then 2 inches. An example is shown
in Figure 2.33.
Figure 2.33 Curb condition heaved.
42


Settled is when the curb sinks more then 2 inches. An example is shown in
Figure 2.34. Curb condition drainage may also be collected instead of settled
in this figure.
Figure 2.34 Curb condition settled.
Drainage is when there is standing water found in the gutter or evidence that
43


Undermined is when the base/sub grade of the curb has been washed away
and weakened by water. An example is shown in Figure 2.36.
Figure 2.36 Curb condition undermined
There are several factors causing each these distresses. These factors are listed
below.
Major Crack:
o Heavy tire load on gutter or curb
o Weak concrete strength
o Freeze thaw action
Spalls:
o Heavy tire load on the tip of the curb
o Impact load due to accident or snow plow
44


o Freeze thaw action
Heaved
o Uplift caused by tree root system
o Freeze thaw action
Settled
o Weak sub grade
o Utility backfill causes weak sub grade under the curb
o Undermine of sub-base from a drainage condition
Drainage
o Poor grade design
o Asphalt overlay erosion
o Undermined
o Water stream corroding away sub grade material trough a crack
creating an air void under the curb. This can lead to settle or heave
distress.
2.7 Preparing the Data for Rating System
In order to develop a curb and gutter condition rating system, several modifications to
the existing data were needed. The basic data as collected was only designed for a
map display purpose using ArcMap GIS as shown in Figure 2.11. Data modifications
included completing a spatial join using ArcMap, using a spreadsheet to calculate the
45


condition rating, and then bringing back the data to ArcMap for display purposes
using the Join Table Attribute Excel table can be joined back into ArcMap by using
Join Table Attribute feature. Completing this process cab be rather time consuming
as a high number of manipulations with the GIS data are needed. The basic problem
is that the curb inventory and distress data are located in two independent files. One
can view the relationship between distress occurrence and location using ArcMap
display, but it difficult to see this relationship when viewing raw data as shown in
Figure 2.37. Some of the difficulties encountered in trying to identify occurrence
and location are as follows:
o More than 1 data collector unit is used, therefore; using TIME and DATE
field for correlation will generate errors as all three units as the three data
collectors operate around at the same time.
o Curb Location is divided into three columns (FROM, TO, TRAVELLING).
These three columns must be cross-correlated with the three columns on the
distress data and the curb data in order to make a correlation between location
and occurrence.
o Some streets contain more than one curb type
46


-0
S3 Microsoft Excel curb.xls
jdj.
mm . s^tpp c fSillllitit i
m TIME DATE LENGTH TYPE FROM TO TRAVELLING
§p| 11:06:22.64 01/04/2002 287,800000 corribo N DECATUR ST N elm; ST 48TH AVENUE SOUTH DR
IP 11:11:29.89 01/04/2002 580.100000 9" vertical W47THAVE , 4BTH AVENUE SOUTH DR N ELM CT
m 11:16:14.77 01/04/2002? !594.100000 9" vertical W 46TH AVE A W47THAVE & N ELM CT |
m 11:20:51.18 01/04/2002 ^610.800000 9 vertical W45TH AVET W46THAVE 1 N ELM CT I
m 11:26:15.59 01/04/2002 594:100000 9 vertical W44THAVE 1 W45THAVE 1 N ELM CT !
ikS 09:58167^5 01/04/2002 138.000000 9 vertical N DECATUR Sf N ELM CT | W44THAVEI
j£8l 10:03:27.8 ft 01/04/2002 124,600000 9 vertical N DECATUR ST N DECATUR ST 1 |W 44TH AVE|
1 fr-iT
ms.
M1

10:11:12.11
TO: 19:26.67
M

' 1
SI


IfiArial
: lisawsaa aauwyjwtw. *
~-10
mmmm telSIS
fawfeSr-fe,^-# jjg%SWssS3S
RELATE ID
0fM4/20Q
01/&4/2Qoill
2] Microsoft Excel distress.xls

10-3949 14 mynj/pm *X- rffjf: S§ 1 . E. -'r eiiiSliSI H amwLfmu?
m 10:45:42.59 6i/04/200: TIME ^ DATE CONDITION FROM TO 1 TRAVELLING
10:57:20.73 01/04/200 m 11:13:04.20 01/04/2002 heaved W47THAVE 48TH AVENUE SOUTH DR N ELM CT
11:16:00.11 01/04/200 m 11:13:35.21 01/04/2002 heaved W 47TH AVE 48TH AVENUE SOUTH DR N ELM CT
m 11:21:17,79 01/04/200 11:14:06.68 01/04/2002 heaved W47THAVE 48TH AVENUE SOUTH DR N ELM CT
W; 11:24:31.68 01/04/200 11:14:29.98 01/04/2002 heaved W47THAVE 48TH AVENUE SOUTH DR N ELM CT
mi 11:30:07.20 01/04/200 m 11:14:59.71 01/04/2002 settled W47THAVE 48TH AVENUE SOUTH DR N ELM CT
mi 11:34:52.28 01/04/200 m 11:16:48.30 01/04/2002 heaved W46THAVE W47THAVE N ELM CT
m 10:09:50.46 01/04/200 8 11:17:29.22 01/04/2002 heaved W46THAVE W47THAVE N ELM CT
m 10:18:34.76 01/04/200 m 11:21:38.11 01/04/2002 settled W45THAVE W46THAVE N ELM CT
Figure 2.37 Multiple columns that can be matched for a rating system


The best way to correlate these two tables shown in Figure 2.37 is by using an
ArcMap GIS. Using the spatial join technique, curb data can be merged into distress
data based on location. Spatial join is a process that ArcMap uses to join two
different data sets on a map based on its geographical location. After the spatial join,
the merged data table can be exported into a database file that can he opened with
Microsoft Excel or any other spreadsheet as shown in Table 2.8. Using a spreadsheet
program, the number of each distress types on each curb can be computed as shown
in Table 2.9. This process can only accommodate a finite number of rows depending
on the limitations of the spreadsheet program used. For example, Microsoft Excel
will only allow 65,536 rows of data, therefore, only 65,536 rows of distress data can
be processed at any given time. Therefore, to be able to rate the entire curb and gutter
data, the CCD must be broken into multiple sections not exceeding 65,536 rows of
distress data.
48


4^
Condition Length Type Relate ID Curb Gone Spalls M^or Crack Heaved/ Settled Drainage Undermined
Mqor cracks 177.3 6 743852 0 0 1 0 0 0
Spalling 177.3 6 743852 0 1 0 0 0
Spalling 177.3 6 743852 0 1 0 0 0 0
Major Cracks 251.5 6 757214 0 0 1 0 0 0
Spalling 251.5 6 757214 0 1 0 0 0 0
Spalling 251.5 6 757214 0 1 0 0 0 0
Spalling 251.5 6 757214 0 1 0 0 0 0
Spalling 251.5 6" 757214 0 1 0 0 0 0
Drainage 304.fi 6 766661 0 0 0 0 1 0
Major cracks 365.1 6" 887988 0 0 1 0 0 0
Spalling 365.1 6 887988 0 1 0 0 0 0
Settled 365.1 6 887988 0 0 0 1 0 0
1 Curb gone 443.4 9" 1099757 1 0 0 0 0 0
Settled 443.4 9 1099757 0 0 9 1 0 0
Spalling 443.4 9" 1099757 0 1 0 0 0 0
Spalling 584.9 Flag stone 1175209 0 1 0 0 0 0
Settled 584.9 Flag stone 1175209 0 0 0 1 0 0
Spalling , 584.9 Flag stone 1175209 0 1 0 0 9
Table 2.8 Distress data after spatial join


Condition Length Type Relate ID Curb Gone Spalls Major Crack Heaved/ Settled Drainage Undermined
Settled 584.9 Flag stone 1175209 0 0 0 1 0 0
Major cracks 387.4 combo 1331567 0 0 1 0 0 0
Major cracks 387.4 combo 1331567 0 0 1 0 0 0
Major cracks 387.4 combo 1331567 0 0 1 0 0 0
Major cracks 387.4 combo 1331567 0 0 1 0 0 0
Table 2.8 Distress data after spatial ioin (continued),


RELATEJD LENGTH TYPE CONSTYEAR Curb Gone Snails Major cracks Heaved /Settled Drainage Undermined
743852 177.300000 6" 1895 0 2 1 0 0 0
757214 251.500000 6 1895 0 4 1 0 0 0
766661 304.600000 6 1874 0 0 0 0 1 0
803545 12.400000 6 1964 0 0 0 0 0 0
850377 64.500000 Flagstone 1868 0 0 0 0 0 0
887988 365.100000 6 1973 0 1 1 1 0 0
920210 280.300000 9 1888 0 0 0 0 0 0
921844 306.400000 9 1888 0 0 0 0 0 0
1099757 443.400000 9 1877 1 1 0 1 0 0
1175209 584.900000 flagstone 1868 0 2 0 2 0 0
1331567 387.400000 combo 1966 0 0 4 0 0 0
Table 2.9 Adding the number of distresses


After the individual number of each distress type has been determined for each curb
section, the process of rating can begin. Two relatively complex tables can be merged
into a single simple spreadsheet table, whereby mathematical operations can occur
using the spatial join process described above. This process is necessary regardless of
which type of condition ration algorithm is chosen. The rating number of each curb
and gutter section, along with a unique number generated by ArcMap can be saved as
a spreadsheet database extension. This table can then be joined back into ArcMap by
using the command join attribute from a table feature. It is possible to add these
data processing steps to the main data collection algorithm and condition rating could
be completed simultaneously along side data collection. A proposed data-
collecting/rating algorithm will be discussed later in chapter five.
52


3. Condition Rating System
A rating system for curbs and gutters will assist the CCD in curb and gutter
maintenance and repair strategies. The city would like to repair curb and gutter in
conjunction with street maintenance thus closing street access only once. In order to
have a good quality street and street subsystem product, the curb and gutter
installation and repair should occur prior to the asphalt overlay process. Therefore,
by studying curb and gutter inventory and assessment data, the CCD can dispatch
asphalt and concrete repairs concurrently. However, as shown in Figure 3.1, the
condition of the curbs cannot be easily determined by only viewing the map with
distress data showing.
Figure 3.1 Aerial view of a portion of the CCD with curb and gutter data showing up
as occurrences. Each dot on the map represents major cracks, spalls, settled,
heaved, curb gone, drainage, and undermined.
53


Using a rating system, which will be discussed in the next several sections, these
clustered distresses that are very hard to interpret can be transformed into a color-
coded distress map as shown in Figure 3.2. With this map, the CCD would be able to
more easily determine which areas need more maintenance attention. This rating
system breaks down the curbs into 3 colors, red for poor condition, yellow for
medium condition, and green for good condition. With further post processing the
condition rating can be broken down into as many rating categories as necessary.
The street subsystem data needs to be processed prior to assigning condition ratings.
Information needed for the rating system presented in this thesis is dependent on
knowing the length of the curbstone inspected. Unfortunately this data was not
collected. In this study it is assumed that all curbstones are 8 feet long, which is a
reasonable assumption given that curbstones in the city tend to vary from 5 feet to 12
feet. The 8-foot stone is most common.
54


Figure 3.2 Distresses on curb with a rating system that can be interpreted by user
much easier when compared to Figure 3.1. This rating system breaks down the curbs
into 3 color coded conditions, red for poor condition, yellow for medium condition,
and green for good condition.
3.1 Condition Index
The condition index (Cl) is a rating system used to evaluate the condition of a
structure based on a relative scale of 0-100. The Cl was developed by the U.S. Army
Corps of Engineers in the 1990s to have uniform condition assessment procedures
for its civil works structures such as locks, dams, and the components associated with
each. [Rens, 1989]. Although curbs and gutters are not U.S Army Corps of
Engineers structures, the idea of the Cl rating system can be can be utilized and
applied. The goal of such a rating system is to capture the ideas and knowledge of an
expert person and apply it to a computer algorithm. In order to get an unbiased
55


rating system, questionnaires were sent to several experts in pavement management
systems. Several meetings were conducted to explain how the Cl works. In so much
as possible, the ideas of these experts have been incorporated into the rating system
presented in this research.
The main equation of the Condition Index is given by Equation 3.1:
Cl 100 ( 0.4 (3.1)
X max is a limiting value of X. When applied to curb and gutter, X is the number of
distress occurrences, where X max is the maximum number of distresses when the curb
is considered in poor condition and immediate repair action is required. Therefore, if
a curb does not have any distress, the Cl will be 100. On the other hand, when X is
equal to X max, the Cl will be equal 40. Figure 3.3 illustrates graphically the equation
while Table 3.1 explains each condition index zone.
56


Condition Index .
Condition Index
Figure 3.3 Condition index as related to X (a measurement or distress occurrence)
/ Xmax (limiting value of the distress)


Zone Condition CI Recommended Action
Zone 1 Excellent 90-100 Curb is most likely newly constructed and have few or no distresses
Good 70-89 Normal tear and wear is noticeable. A minor repair is recommended.
Zone 2 Fair 40-69 Curb is deteriorated and looks unattractive. Repair is recommended.
Zone 3 Poor 0-39 Immediate repairs or replacement is needed (whole curb section)
Table 3.1 Condition index zones
Because the lengths of street vary, some adjustments are needed in normalizing the
value of the CI. X max is determined by an expert panel questionnaire based on a 400
ft (approximately 50 stones) length of curb and gutter. Each curb and gutter CI is
broken into six individual Cls, one for each distress type. Since the heaved and
settled distress are somewhat similar, these distresses were combined into one CI.
The overall Condition Index (OCI) is then calculated by summing the Cl of each
distress multiplied by its weighting factor. The weighting factor is the percentage of
how much each distress controls the overall CI. The overall CI equation is:
OCI Wmajor cracks Clmajor cracks spalls CI spalls undermined CI undermined (3-2)
The expert panel also determined the weighting factors for each distress. An
example of this process is shown in Table 3.2.
58


Distress Cl Weighting Factor Cl* WF
Major cracks 75 14.4 1080
Spalls 83 6.5 539.5
Curb Gone 92 20.1 1849.2
Heaved/Settled 79 24.4 1927.6
Drainage 100 20.4 2040
Undermined 100 14.3 1430
Summation 100.0 8866.3
Overall Cl = E(CI*WF)/EWF= 89
Table 3.2 Example of overall Cl
3.1.1 X max and Weighting Factor
In order to come up with a fair and unbiased rating system, a panel of experts was
chosen. Six CCD employees who have experience in infrastructure management,
particularly curb and gutter, were included. This group was asked to determine Xmax
and the weighting factor of each type of distress on different curb types. The Xmax for
this rating system is the maximum number of stones (out of 50 stones) when the curb
is considered poor and immediate action is required. This immediate action was
determined to be a complete repair or replacement of curb and gutter section. The
59


weighting factor for each distress is related to how significant each type of distress
impacts the overall condition index. The weighting factor varies for the each
different type of curb. The results are as shown in Table 3.3 through 3.5. Table 3.3
shows different X max obtained from the expert for different distresses and curb type.
Table 3.4 shows the raw unedited data for different weighting factors for each distress
as a function of curb type.
X Max Curb Gone
Expert 1 Expert 2 Expert 3 Expert 4 Expert 5 Expert 6 Average
6 inch 20 25 50 25 25 40 29.0
9 inch 20 35 50 25 25 40 31.0
combo 13 25 50 25 35 35 29.6
mountable 17 25 50 25 35 35 30.4
Overall Average = 30.0
X Max Major Cracks
Expert 1 Expert 2 Expert 3 Expert 4 Expert 5 Expert 6 Average
6 inch 40 30 40 40 38 40 37.6
9 inch 40 35 40 40 38 40 38.6
combo 40 30 40 40 38 40 37.6
mountable 40 25 40 40 38 40 36.6
Overall Average = 37.6
X Max Spalls
Expert 1 Expert 2 Expert 3 Expert 4 Expert 5 Expert 6 Average
6 inch 40 40 50 25 40 45 39.0
9 inch 40 40 50 25 40 45 39.0
combo 40 35 50 25 30 45 36.0
mountable 40 40 50 25 30 45 37.0
Overall Average = 37.8
Table 3.3 X max data as obtained from 6 experts.
60


X Max Settled/Heaved
Expert 1 Expert 2 Expert 3 Expert 4 Expert 5 Expert 6 Average
6 inch 12 25 40 45 40 25 32.40
9 inch 12 30 40 45 40 25 33.40
combo 10 25 40 45 35 25 31.00
mountable 12 25 40 45 35 25 31.40
Overall Average = 32.1
X Max Drainage
Expert 1 Expert 2 Expert 3 Expert 4 Expert 5 Expert 6 Average
6 inch 30 25 50 25 25 25 31.0
9 inch 30 35 50 25 25 25 33.0
combo 30 25 50 25 25 25 31.0
mountable 30 20 50 25 25 25 30.0
Overall Average = 31.3
X Max Undermined
Expert 1 Expert 2 Expert 3 Expert 4 Expert 5 Expert 6 Average
6 inch 20 20 40 35 30 30 29.0
9 inch 20 30 40 35 30 30 31.0
combo 13 25 40 35 30 30 28.6
mountable 17 25 40 35 30 30 29.4
Overall Average = 29.5
Table 3.3 X max data as obtained from 6 experts (continued).
61


Six Inch Vertical Curb Weighting Factor
Distress Expert 1 Expert 2 Expert 3 Expert 4 Expert 5 Expert 6 Average
Curb Gone 10 30 13.33 13 51 20 22.89
Major Cracks 20 15 13.33 25 4 5 13.72
Spalls 5 5 6.67 10 4 5 5.95
Settled/Heaved 30 20 26.67 30 30 10 24.45
Drainage 10 15 20 20 10 40 19.17
Undermined 25 15 20 2 1 20 13.83
Sum 100 100 100 100 100 100 100
Nine Inch Vertical Curb Weighting Factor
Distress Expert 1 Expert 2 Expert 3 Expert 4 Expert 5 Expert 6 Average
Curb Gone 10 25 13.33 20 51 20 23.22
Major Cracks 20 10 13.33 30 4 5 13.72
Spalls 5 5 6.67 8 4 5 5.61
Settled/Heaved 30 30 26.67 15 30 10 23.61
Drainage 10 15 20 25 10 40 20
Undermined 25 15 20 2 1 20 13.83
Sum 100 100 100 100 100 100 100
Combo Curb Weighting Factor
Distress Expert 1 Expert 2 Expert 3 Expert 4 Expert 5 Expert 6 Average
Curb Gone 25 30 11.76 2 5 30 17.29
Major Cracks 20 10 17.65 25 15 4 15.28
Spalls 5 5 11.76 12 10 1 7.46
Settled/Heaved 25 20 23.53 31 30 10 23.26
Drainage 5 15 17.65 22 30 40 21.61
Undermined 20 20 17.65 8 10 15 15.11
Sum 100 100 100 100 100 100 100
Table 3.4 Weighting factors as obtained from the 6 experts
62


Mountable Weighting Factor
Distress Expert 1 Expert 2 Expert 3 Expert 4 Expert 5 Expert 6 Average
Curb Gone 15 25 13.33 13 5 30 16.89
Major Cracks 20 10 13.33 25 15 5 14.72
Spalls 5 5 6.67 10 10 5 6.95
Settled/Heaved 30 30 26.67 30 30 10 26.11
Drainage 10 15 20 20 30 30 20.83
Undermined 20 15 20 2 10 20 14.5
Sum 100 100 100 100 100 100 100
Table 3.4 Weighting factors as obtained from the 6 experts (continued).
From the results of the raw data gathered from the expert panel, one can determine
that the X max for distresses for different type of curbs are very similar. Originally the
plan was to have to a different rating formula for each individual curb type.
Therefore, to simplify the condition algorithm, the overall average X max from each
distress type was used for all types of curbs. The same argument was used for the
weighting factor. Because curb and gutter is not a critical structure like a bridge,
building or interstate, when the condition is poor, the X max gathered from the panel of
judges was appropriately high. The X max and Weighting Factor used for the Cl are
shown in Table 3.5.
63


Distress WF av. Xmax
Curb Gone 20.1 30
Major Cracks 14.4 38
Spalls 6.5 38
Settled/Heaved 24.4 32
Drainage 20.4 32
Undermined 14.3 30
Average 33
Table 3.5 Average weighting factor and Xmax used for any type of curb
In some cases, when most of the distresses are of one type, the individual distress for
that particular type of distress will be extremely low. Depending on the weighting
factor which varies between 10 and 20, the Overall Condition Index may still be
relatively high and on the order of 70 or 80. Therefore, an adjustment factor was
developed to account for this scenario. [Rens, 1989]. When a Cl of an individual
distress type is lower then 40, then the individual weighting factor is amplified by a
factor of 8. When the Cl is between 40 and 70, the amplification factor varies
linearly between 8 and 1 as shown in Figure 3.4. An example of this process is
detailed in Table 3.6.
64


Adjustment Factor for Cl
Figure 3.4 Adjustment factor for weighting


Distress Cl WF AF WF* AF New WF Unadjusted Cl Adjusted Cl
Major cracks m 14.4 8 115.2 55.2 561.6 2152.8
Spalls 65 6.5 2.2 14.3 6.9 422.5 445.4
Curb Gone 98 20.1 1 20.1 9.6 1969.8 943.8
Heaved/Settled 97 24.4 1 24.4 11.7 2366.8 1134.1
Drainage 100 20.4 1 20.4 9.8 2040.0 977.5
Undermined 100 14.3 1 14.3 6.9 1430.0 685.2
SUM 100 208.7 100 87.9
Table 3.6 Example of weighting factor adjustments. Note that the individual
distresses of cracks and spall would have resulted in the overall Cl to be 87.9 (good),
however in using the adjustment factor results in the overall Cl to be 63.4 (Fair).
Since the Cl of major cracks is less then 40, the weighting factor is amplified by the
factor of 8. In addition, the Cl of spalls is between 40 and 70, therefore, the
adjustment factor for spalls is amplified by the factor of 8- 7(CI-40)/30 = 2.2.
3.1.2 Cl Adjustments for Curb and Gutter
The most powerful aspect of the Condition Index is the ability to trigger a low rating
number when a vital distress is collected. In 1989 the Cl, was introduced to Miter
gates by the U.S. Army Corps of Engineers. [Rens, 1989]. In that work, the Cl was
divided into 10 different distresses, which included anchorage movement, elevation
change, miter offset, gaps, downstream movement, cracks, leaks/boils, dents,
noise/jumping/vibration, and corrosion. Each distress represented a portion of the
overall condition of the entire miter gate structure. However, for curb and gutter,
66


only one distress is utilized per curbstone, even when a stone contains multiple
distresses. Therefore, the Cl rating system used by the Corps had to be modified. For
example, on a curbstone that is shown in Figure 3.5 four individual distresses exist
consisting of major cracks, drainage, and undermined, only the distress of drainage
would be recorded due to drainage has the highest weighting factor compared to the
other 2 distress types.
Figure 3.5 Multiple distresses on a curbstone.
These modifications are explained below:
Originally, distresses that do not occur on a curb can cause the Cl to be
artificially inflated. For example, when a distress does not occur on a curb (Cl
67


= 100), the adjustment factor for that weighting factor is 1. For curb and
gutter, it was decided that when a distress does not occur on a curb section,
this individual distress should not artificially inflate the Cl calculation and
should not be a part of the overall CI. With this new adjustment, only
distresses that occur on a cur section control the OCO. Example of this
modification is shown in Table 3.7. Notice how the OCI of the modified
adjustment in Table 3.7 is 15 points lower thus moving the OCI from fair to
poor which is more reflective of the actual condition of the section.
Distress CI WF AF WF AF Adjusted WF CI
Major cracks 40 14.4 8 115.2 57.3 2293.7
Spalls 98 6.5 1 6.5 3.2 317.1
Curb Gone 97 20.1 iMPlIlPlD rwt- fry s -i 20.1 10.0 970.5
Heaved/Settled 100 24.4 24.4 12.1 1214.5
Drainage 100 20.4 20.4 10.2 1015.4
Undermined 100 14.3 isms 14.3 7.1 711.8
SUM 100 200.9 100 65.2
Modified
Distress CI WF AF WF AF Adjusted WF CI
Major cracks 40 14.4 8 115.2 82.5 3300.9
Spalls 100 6.5 ItttlSIlll 0 0.0 0.0
Curb Gone 100 20.1 0 0.0 0.0
Heaved/Settled 98 24.4 24.4 17.5 1712.9
Drainage 100 20.4 mmi 0 0.0 0.0
Undermined 100 14.3 S§lil 0 0.0 0.0
SUM 100 139.6 100 Sill "
Table 3.7 Adjustment factor of 0 modification when a distress does not exist.
When all the distress CIs are almost equal, the overall CI can also be
artificially inflated. Calculating an alternative CI by treating all types distress
68


as if they were the same can solve this problem. The X max for this Cl
calculation is 33 taken from the average of X max in Table 3.3. Example of this
new adjustment is shown in Table 3.8. Notice how the Cl for this new
adjustment is 22 points lower bringing the condition from excellent to good.
Distress # Distress Cl Weighting Factor Adjustment Factor WF* AF Cl WF Cl WF AF
Major cracks 2 95 14.4 14.4 1372.2 1372.2
Spalls 2 95 6.5 1.0 6.5 619.4 619.4
Curb Gone 2 94 20.1 to 20.1 1890.9 1890.9
Heaved/ Settled 2 94 24.4 1.0 24.4 2304.2 2304.2
Drainage 2 94 20.4 20.4 1926.5 1926.5
Undermined 2 94 14.3 1.0 14.3 1345.3 1345.3
SUM 12 100 100.1 9458.4 9458.4
#STONE = 50
OCI = X (CI*WF*AF) / X WF*AF= 94
OCI2 = 100 (0.4)A(12/33)= 72
Table 3.8 Second overall Cl calculation by treating all distresses as if they were the
same type.
To better understand the changes of the effect of these modifications, refer to the flow
chart shown in Figure 3.6 and the example shown in Table 3.9. Applying the logic
explained in the previous paragraphs to a portion of the CCD is shown in Figure 3.7.
In this figure, different colors are shown to represent different conditions.
69


Distress # Distress Cl Weighting Factor Adjustment Factor WF AF Cl WF
Major cracks 21 mimm 14.4 3.3 47.1 2838.6
Spalls 1 98 6.5 1.0 6.5 634.5
Curb Gone 1 97 20.1 1.0 20.1 1949.5
Heaved/Settled 2 94 24.4 1.0 24.4 2304.2
Drainage 0 100 20.4 0.0 0.0 0.0
Undermined 0 o 14.3 0.0 0.0 0.0
SUM 25 100 98.1 7726.9
Total # Stones = 50
OCI = £ (CI*WF) / 2 WF*AF= 79
OCI2= 100*0.4 A(25/33) = 50 50
OCI min= 50
Table 3.9 Example of weighting factor adjustments combined
70


Figure 3.6 Flow chart of Overall Cl (OCI) calculation
71


Figure 3.7 Sample of Cl ArcMap output
3.2 Simplified Condition Index
Another way to present the relative condition of curb sections is to calculate a straight
percentage of the stones in good condition. Unlike the Condition Index previously
discussed, this Simplified Condition Index (SCI) treats every type of distress the same
way and gives a linear output as opposed to exponential (Equation 3.3). The SCI can
be calculated by dividing the number of undamaged stones by the number of total
stones in a given section of curb length. The number of total stones can be calculated
by dividing the length of the curb by the length of the stone. SCI numbers are usually
slightly higher than the OCI numbers, however SCI is extremely easier to calculate.
An inspector can easily calculate SCI without the need of a computer. SCI can be
used as long as the weighting factors obtained from the experts are fairly equal to one
72


another (Table 3.5). A sample of SCI ArcMap display is in Figure 3.8. This display
is almost identical to that as shown in Figure 3.7.
SCI = 100 (#undamaged stone / # stone) (3.3)
3.3 Distress Density Map
The distress Density Map is similar to a population map of the planet earth. Each
country has different color depending on the size of its population. To see how
densely populated each country is, the number of individuals are divided by its
surface area. Using GIS, the relative health of the curbs and gutter network can be
similarly classified by using the ratio of distress count to surface area. The area can
be a neighborhood subdivision, a quarter mile section, or any other convenient area to
be examined. To produce this map, a similar process is required like was used when
calculating the Cl. First the distress data is joined with any area (polygon data shape)
73


GIS data. Then using a spreadsheet, the number of distresses for each area can be
calculated. Then this table can be exported back to ArcMap and the areas symbolized
by different colors depending on how many distresses the area has, normalized by the
size of the area. Distress density does not use any weighting factors and all types of
distresses are treated equally. Also, some areas are more densely populated with curb
and gutter than other. In a neighborhood subdivision shown in Figure 3.9, the curb
and gutter population is not very dense, this area will be shown in distress density
map as in good condition although the curb and gutter in that area may be in poor
condition. Therefore, this map output does not rank the areas fairly, however it can
aid decision-making process. An example of the output of this map can be seen in
Figure 3.10.
Figure 3.9 Typical area in CCD that can trigger false distress density outputs due to a
sparse population of curb and gutter inventory.
74


Figure 3.10 Example of distress density map for curb and gutter in the CCD (Green =
condition 1, Yellow = condition 2, Orange = condition 3, Red = condition 4) where
condition 1 would be relatively more healthy than condition 4.
3.4 Error Analysis of Data
In order to quality control data acquisition, two studies during the course of this
research were completed to help identify data collection inconsistencies. These
studies were completed at the 5th and 30th month milestones and the results are
discussed in the sections that follow.
3.4.1 February 2002 (5th month analysis)
In 2002, a study was completed by the UCD research team in which comparisons of
data collected by different inspectors were analyzed for similarities and differences.
In order to eliminate bias, each inspector completed the target area independently and
without knowledge of the study. Four teams of two people were utilized to inventory
75


and assess the same target area. In addition to the teams, a supervisor also
inventoried the same area. The results of the data comparison indicated that some
inconsistencies were occurring in the data acquisition at a relatively early point in the
study. A partial result of this study is shown in Figure 3.11 where only the raw
number of distresses collected by each of the inspection teams is presented.
Distresses
w
o
(A
(0
0)
.2
'5
75
o
i-
A B C D Supervisor
Data collector
Figure 3.11 Number of distresses result in error analysis study
Although the data was inconsistent for the 5-month study, it resulted in positive
outcomes. Inspectors were notified about this study and the results were presented.
The main reason why inspector B and C have approximately twice as many distresses
as everyone else was because B and C collected more than 1 distress per curbstone.
Upon closer evaluation of all the inspectors, it was discovered that communication
76


problems during initial training sessions led to this problem and were subsequently
corrected. As a result, open forum discussions were initiated which resulted in a
better overall training scheme to help address ongoing questions. The supervisor also
occasionally comes along with the inspection teams to ensure quality control and
uniformity of the data collection. Finally, written examinations are also periodically
completed to facilitate the open forum discussion. These examinations cover data
collecting errors that the supervisor notices either in the field or during the data
download process.
3.4.2 March 2004 (30th month analysis)
In March 2004, a follow up error analysis was completed. Six inspectors collected
data in the target study area, four of which collected the data without the knowledge
of the study. The other two sets of data were collected independently by the author of
xl.
this thesis and by die supervisor. Unlike in 5 month analysis, all inspectors knew
that only 1 type of distress could be collected per curbstone. The partial result of this
study is shown in Figure 3.12. Table 3.10 divided the number of distresses into
different curb sections. Similar table format for 5th month study is not possible due to
lack of data. Table 3.10 shows that there is still some inconsistency in the data
collection, but the results are better. The sources of inconsistency, which affected
both 5 and 30 month studies, will be discussed in the next section.
77


The total number of distresses in Table 3.10 varies quite a bit, however the numbers
of distress of each individual curb section are fairly close to each other. The
maximum overshot or undershot is no more than 5 distresses. Especially after the
SCI calculation, the overall SCI only ranges between 79.4 and 88.6.
Figure 3.12 Number of distresses result in error analysis study in March 2004.
i i i i i i i i
# Distress SCI
Structure Length A B c D CS1 CS2 Average SD A B C D CS1 CS2 Average SD
Curb A 267.3 4 3 7 4 6 7 5.2 1.7 88.0 91.0 79.0 68.0 82.0 79.0 84.5 5.2
Curb B 598.5 10 8 16 10 14 14 12.0 3.1 86.6 89.3 78.6 (£.6 81.3 81.3 84.0 4.1
Curb C 274.8 5 5 8 6 6 7 6.2 1.2 85.4 B5.4 76.7 82.5 82.5 79.6 82.0 3.4
CurbD 587.6 11 13 19 13 19 19 15.7 3.7 65.0 82.3 74.1 82.3 74.1 74.1 78.7 5.1
CurbE 265.2 3 10 11 6 7 6 7.2 2.9 91.0 69.8 66.8 81.9 78.9 81.9 78.4 8.8 -
Curb F 599.4 3 1 9 2 3 4 3.7 2.8 96.0 98.7 88.0 97.3 96.0 94.7 95.1 3.7
Curb G 600.2 13 10 19 16 21 18 16.2 4.1 82.7 86.7 74.7 78.7 72.0 76.0 78.5 5.4
Curb H 258.5 0 0 0 0 0 0 0.0 0.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 0.0
SUM 3451.5 49 50 89 57 76 75 66.0 88.6 88.4 79.4 86.8 82.4 62.6 84.7
Table 3.10 Data comparison result in March 2004 error analysis
78


3.4.3 Error Analysis Summary
The results of the data comparison show that there has been an increase in quality
compared to the results obtained in 2002 as shown in Figure 3.12 and Table 3.10. The
average, maximum, and minimum of number of distress in the 30 month analysis are
66, 89, and 49. Therefore, the maximum is approximately 35% higher than the
average, and the minimum is approximately 26% lower than the average. In
iL
comparison with the 5 month analysis, the average, maximum, and minimum
numbers of distress are 84.6, 123, and 49. Therefore, the maximum is about 45 %
higher than the average, and the minimum is about 42% lower than the average. The
overshot and undershot has decreased by 10% and 16%. In conclusion, the main
difference of the study in 5th and 30th month is that in the 30th month study all
inspectors knew that only one distress may be collected every curbstone, where in the
5th month study not all inspectors were aware.
The sources of inconsistency of both studies are summarized in the 6 points below.
1. Confusing a crack with a curbstone joint
The curbstone joint is the cut line between curbstones. Some distresses can
appear to be similar to a curbstone joint or vice versa. Therefore, occasionally
an inspector would call or miss distress conditions. An example of this
confusion is shown in Figure 3.13.
79


Figure 3.13 Major crack that appears to look like a curbstone line
2. No priority level was established
One inspector may observe different distress for each curbstone. On a
curbstone that has multiple distresses like shown in Figure 3.5, the inspectors
may collect four different results. One inspector may call out major crack,
drainage, or undermined when only drainage, which has the highest weight
factor, was to be recorded.
3. Human error
Inspectors occasionally missed some of the distress due obstacles such as
vegetation, cars, snow, or leaves, as shown in Figure 3.14.
80


, ,x r. t y

* V ^ * ftp-- -i .,V- f. ffciF v r * v {> V" ^ 'hv: v i -A
> * . , v/ :l l :
' W * ' t> t "jvf
* * ,
Figure 3.14 Distress that is covered with vegetation
4. Different opinions from different inspectors
When a distress is not significant enough, it does not need to be
collected. However, different inspectors have different opinion on
whether a distress is significant enough to be collected. Figure 3.15
shows an example of a spall distress that may or may not be collected
by different inspectors. Some inspectors consider this distress too
small to be collected, but some consider the spall significant enough.
In this example, the spall is probably a distress that should have been
recorded.
81


Figure 3.15 Minor distress that can cause inconsistency in error analysis
5. Drainage distress on dry gutter pan
When there is an indication of deposits from a drainage problem on a
gutter pan, the drainage distress needs to be collected whether the
gutter pan is wet or dry. The cleanliness of an area and the recent
weather in the area can cause significant inconsistency in data. In
other words, inspecting a curb while the gutter is wet may result in
higher number of distress encountered. Figure 3.16 shows deposits
on a curb pan that may be a result of drainage deposit or lack of street
sweeping. In this example, the distresses of drainage condition
probably should have been recorded.
82