Mostafa A. Abo-Hashema
Department of Civil Engineering,
P.O. Box 16569, Al Ain, UAE
Alaa Marouf Abdel Samad
Al Ain Municipality, Abu Dhabi, UAE,
Yousef A. Al-Zaroni
Al Ain Municipality, Abu Dhabi, UAE,
Mostafa M.S. El-Hawwary
IT Expert, WS Atkins and Partner Overseas,
P.O. Box 16569, Al Ain, UAE
Recently, the concern of highway agencies has shifted from focusing on methods and techniques of performing maintenance to managing maintenance activities. Studies indicated that 60% of a highway network would reach the stage of functional failure unless Pavement Maintenance Management Systems (PMMS) are implemented. For that reason, many highway agencies developed their own PMMS to improve the efficiency of decision making, provide feedback on the consequences of decisions, control the rate of deterioration, and limit maintenance costs. Adding a technological element to PMMS has been a vital step in supporting and improving decision-making. Geographic Information Systems (GIS) were found to be powerful tool in providing such exertion. The implicit goal of using GIS is presenting new techniques for proper collection, archiving, and analyzing of pavement maintenance data. Accordingly, Al Ain Municipality, UAE, decided to implement a GIS-Based PMMS through the current project “Maintenance of Roads and Bridges in Al Ain Region”. Two Levels of PMMS data are collected in this project: Network-Level and Project-Level. The network-level is an overview of the condition of a network of roads. Minimum amount of data was collected at the network-level that is needed to support the analysis of overall condition of the network and determine which streets require repairs. In project-level, more detailed information is needed to finalize the list of projects and to provide a detailed scope-of-work for each individual project. The framework of the project includes a systematic, consistent approach to gathering and analyzing data and generating recommendations and reports. This paper presents the effort made so far in integrating the technological concept of GIS and PMMS activities through the current maintenance project of Al Ain City.
The rapid growth in road construction that took place in the past two decades brought about considerable expansion of road infrastructure, which subsequently fell into disrepair through lack of maintenance. The damage is often so severe that ordinary maintenance will no longer suffice and if roads are to be fully restored, rehabilitation or even reconstruction work is necessary, at a life cycle cost three to seven times higher than that of preventive maintenance strategies. To apply such strategies, information concerning the condition of the network, its rates of deterioration, and the impact of different maintenance and rehabilitation treatments on pavement serviceability level, are required. Such components are typically integrated through the application of Pavement Maintenance Management Systems (PMMS) [TCB/AET, 2001; HPMA Manual, 2001; Abo-Hashema, 2004]
United Arab Emirates roads network played an important function in activating the development of new areas. Huge achievements in the road infrastructure have been done during the past 20 years in UAE. AL Ain is a vital city in Emirate of Abu Dhabi, the UAE capital. It possesses more than 600 kilometers of paved main roads (Centerline-Dual Carriageway), and 3000 kilometers of paved secondary roads (Single Carriageway). To preserve these investments and due to the clear understanding of Al Ain Municipality for the importance of this distinct network, it has been decided to perform maintenance project for Al Ain roads network since 2002. The project has aimed at implementing maintenance and rehabilitation activities to the deprived streets to keep the roads condition to an acceptable level. Initially, maintenance needs were defined arbitrarily. As well as, the maintenance activities were applied randomly without studying or management. In conclusion, there was no defined system in implementing preventive maintenance strategies.
Therefore, it has been decided in 2004 to develop a comprehensive PMMS based on Geographic Information System (GIS) that suits Al Ain specifications. The main objectives of using GIS are archiving maintenance activities, supporting maintenance decisions, facilitating different on-field activities, and follow-up.
GIS-Based Pavement Management Application (GPMA) is a tailored computer application that has been developed in the project. GPMA provides technology and tools to assist in network programming of highways maintenance and rehabilitation as well as project level design. In general, GPMA supports the following:
- Interactive and batch data entry and update.
- Querying, reporting, and spatial displaying.
- Thematic representation of information.
- Maintenance Decision Support
- Road maintenance needs and analysis.
This paper presents the effort made so far in implementing the technological concept of GIS into PMMS activities through the current maintenance project of Al Ain City.
2. Overview of PMMS
A Pavement Maintenance Management System should not be confused with a Pavement Management System (PMS). A PMMS is a part of a PMS program, i.e. they overlay rather than replace one to another. Figure 1 shows PMMS versus PMS and the concept of the overlay between them [Abo-Hashema, 2004; Sharaf et. al, 2003].
Figure 1 Pavement Maintenance Management System (PMMS) Versus Pavement Management System (PMS)
Pavement maintenance is the preserving and keeping of the pavement structure as possible as in its original condition as constructed or as subsequently improved, and such additional work as is necessary to keep traffic moving safely. This involves routine maintenance works (patching, filling ruts, removing surface corrugation, pouring cracks, bleeding surfaces, etc.) and major maintenance activities such as resurfacing, rehabilitation, overlays, etc. [AASHTO, 1986]
Pavement maintenance is not an exact science. The same type of road in different locations requires different maintenance operations. Consequently, repair methods, which give good results at one location, may not be the proper methods at another location. Thus it can be quickly seen that experienced personnel with good judgment are the key to proper maintenance [Haas and Hudson, 1978].
A PMMS provides the framework for decision making in pavement maintenance based upon an objective approach. The complexity of highway maintenance cannot be reduced to a series of mathematical expressions; its management should be subjected to a rigorous systems approach to ensure that policies are developed on a basis of need that performance is monitored and proper financial control is exercised. The overall systems concept of maintenance management system is shown in Figure 2 [Sharaf et. al, 2003; Pinard, 1987].
Figure 2 Maintenance Management System, an Overview
3. Overview of GIS
As a conceptual definition, GIS can be considered as a tool for creating, editing, analyzing, visualizing, and managing of geographic information. While going deeper into specialized application, GIS is introduced as a tool for managing resources technologically with a geographic prospective. Thus, GIS plays an important and efficient role in civil engineering projects and supports all projects activities from planning towards analysis and all through supervision and follow-up.
The information stored in a GIS is represented by a series of spatially referenced datasets. These datasets, which can be vector-based (representation of Geographic Features, GF, as one continuous object in digital environment) or raster-based (representation of GF as a pixels form object in digital environment), are used to organize and manage both the geometry and attribute descriptions associated with various GF [ArcGIS, 2004]. The demonstration of a GF from life in GIS is simulated into one of the polygons, lines, and points shape. According to the GF nature and scope of work, it is replicated into this shape, as shown in Figure 3.
Figure 3 Replication of GF into Digital Format and shapes
GIS datasets include tabular attributes that describe geographic objects. Multiple tables can be linked to the geographic objects by a common thread of fields (often called keys). These tabular information sets and their relationships give GIS users the data management power typically associated with relational database applications [ArcGIS, 2004]. Roads, utilities, and linear GFs are often represented as lines in a network with multiple connections, and intersections. With the interactive intelligence introduced by GIS; Engineers can handle tasks on both network and GF scopes easily, and timely, as shown in Figure 4. Intelligence (Topology) is the method used for defining different GF boundaries and interconnected relations.
Figure 4 Roads Network Representation in Intelligent Environment
4. Integrating PMMS and GIS
Figure 5 shows a schematic representation of the integration between PMMS and GIS. PMMS starts with network identification and go through data collection, data analysis, maintenance priorities, maintenance decisions; and ends with supervision and follow-up. GIS starts with data entry and manipulation, thematic mapping, and ends with tailoring and decision support. GIS acts as the core for the integration process, through managing the different corresponding activities interaction. The integration interface is a transition level that exists to ensure adequate compatibility of integration requirements.
Figure 5 Integrating PMMS and GIS
5. General Framework of the Project
The general framework of the project can be outlined by the PMMS activities workflow, as shown in Figure 6. However, it can be noticed the chained link to the GIS workflow, which is presented in Figure 7. The framework starts from the network identification and ends with work orders and investigation reports with a comprehensive connection to the corresponding GIS activities. The following titles will describe some items in the framework.
6. Network Identification (Referencing)
The first step in the development of a PMMS is to identify the roads network. The development of the referencing system should start with a review of the current practices of the agency. Each of the divisions responsible for collecting pavement data, or associated data should be identified and reviewed.
Figure 6 General Workflow of the Project for PMMS
Figure 7 General Framework of the Project for GIS
For pavement management, one of the existing methods may be selected or a new method may be defined. In either case, plan should be made for coordinating the different referencing methods to permit a free interchange of data [Haas et. al, 1994]. There are three basic methods of referencing pavement sections:
- Route milepost
- Node – Link
- Branch – Section
Once the referencing method is established, the specific pavement sections must be defined for use in the database. The methods used vary widely between agencies. Section can be defined to have either uniform characteristics or fixed length [Haas et. al, 1994].
6.2 Al Ain Network Classes
6.2.1 Main Roads
Main roads are defined as dual carriageways. There are 600 kilometers (Centerline) of paved main roads in Al Ain region. The Town Planning Department (TPD) in Al Ain city has assigned all the main roads with a unique number. It is noteworthy that ninety percent of the intersection points between the main roads are roundabouts and the others are signalized intersections. These intersection points are also defined by the TPD with a unique number.
6.2.2 Secondary Roads
Secondary roads are defined as single carriageways. All the secondary roads are located inside communities, which are located in Districts. There are 3000 kilometers of paved secondary roads located in more than 95 communities. Similar to the coding system of the main roads, secondary roads have a unique community internal coding assigned by the TPD.
6.3 Al Ain Network Encoding System
Although, there is a unique coding system used by the TPD, but there is no clear referencing system that can be used to apply the PMMS activities. Therefore, it is much wiser to extend this unique coding to formulate proper network identification.
6.3.1 Main Roads
The referencing or identification system that has been used in the main roads is Node-Link method. In the node-link method, key points in the network are defined as nodes and the sections between these nodes in each direction are defined as the links. Nodes are usually defined at intersections, boundaries, and point of change in the pavement characteristics, such as a change in surface type.
In this project, a defined section (link) in each direction consists of three traffic lanes excluding the node area. It starts at 50 meters away from each node area or intersection point. The total number of sections in Al Ain region fairly exceeds 500 sections with different lengths. Each inspected section was divided into sample units of 100 meters length per traffic lane through which data can be collected. The section code consists of 12 digits. The first four digits in the code starting from the left side represent the road code (TPD unique number), the next four digits represent starting node code (TPD unique number), and the last four digits represent ending node code. Lane code is represented by adding one more digit to the 12-digits of the section code. Furthermore, adding another three digits to the 13-digits of lane code represents 16-digits of sample code. Figure 8 and Figure 9 show an example of the coding system for Shakhboot Ibn Sultan Road between nodes 101 and 121.
On the other hand, nodes are defined at intersection points separately evaluated as a section. The node-section includes 50 meters towards each street or leg. The intersection points-TPD unique coding was used as the node code.
Figure 8 Network Identification for Main Roads
Figure 9 Coding System of Sections for Main Roads
6.3.2 Secondary Roads
The TPD code for districts, communities, and internal roads was extended to facilitate a unique code on a network scope. The referencing or identification system that has been used in the secondary roads is branch-section method.
A defined branch (District) consists of communities. A defined section (Community) consists of internal roads, which represent samples. The branch code consists of 3 digits (TPD-District unique number). Section code is represented by adding two more digits (TPD- Community unique number) to the 3-digits of the section code. Furthermore, adding another three digits (TPD- internal roads unique number) to the 5-digits of section code represents 8-digits of sample (road) code. Figure 10 shows an example of the network identification for secondary roads in Al Dhaher community.
Figure 10 Network Identification for Secondary Roads
6.3.3 GIS Integration
The selection of the Node-Link and branch-section methods as referencing systems was implemented onto the base map of GPMA in the same order defined in previous sections. While the road GF was represented in a form of vector-Lines, round about was represented in a form of vector-point shape, and districts/communities were represented in a form of vector-polygon shape forming the system spatial data. This spatial data is considered the foundation for generating GIS abstract level “Archive”.
The formulation of the secondary roads labels printed in Figure 10 consists of 8-digits code, as mentioned previously with the combination of district-community-internal road code.
7. Data Collection
A data bank is considered the main tool in which the results and the whole system will completely depend on. The data collection stage is considered one of the stages that need organization and good identification to achieve the objective of the program. Field data collection was performed through two levels:
The network-level is an overview of the condition of a network of roads. Minimum amount of data was collected at the network level that is needed to support the analysis of overall condition of the network and determine which streets require repairs. In project-level, more detailed information is needed to finalize the list of projects and to provide a detailed scope-of-work for each individual project.
It is noteworthy that the database designed for GPMA relayed on the scopes portrayed previously and with respect to the needed analysis discussed later. Collected data was manipulated and properly edited forming GFs attributes, which was linked to the base map forming the “Archive”.
7.2 Condition Survey for Network-Level
An overview of the condition of the network has been carried out covering the defined sections of the main and secondary roads. The purpose of this network condition survey is to come up with maintenance plan for the main and secondary roads. The collected data consists of sections definition, nodes definition, sections length, and general condition of sections.
7.3 Condition Survey for Project-Level
The data collection procedures have been done using special forms for the project-level purpose. Field data collection was performed through three types of data collection such as:
- Inventory Data
- Visual inspection of the Road Surface Distresses
- Materials Investigation
The following sections describe in brief the three types of the field data collection.
7.3.1 Inventory Data
The inventory data, which was carried out in the project-level, includes the following:
- Number of lanes
- Lane Width
- Shoulders (Type, Width, Condition)
- Side Walk (Type, Width, Condition)
- Curb Stones (Type, Condition)
- Median (Type and Width)
- Traffic Data
- Number of Signs
- Length of Guardrail
- Storm water, Humps and Manholes
7.3.2 Visual Inspection of the Surface Distresses
There are 19 distresses listed in PAVER system [Shahin and Kohn, 1981] for computing the Pavement Condition Index (PCI) value, as shown in Table 1. The PCI procedure is well known and approved by many highway agencies as evaluating system.
|Code||Distress Name (Units)|
|1||Alligator Cracking (Sq. m)|
|2||Bleeding (Sq. m)|
|3||Block Cracking (Sq. m)|
|4||Bumps & Sags (L.m)|
|5||Corrugation (Sq. m)|
|6||Depression (Sq. m)|
|7||Edge Cracking (L.m)|
|8||Reflection Cracking (L.m)|
|9||Lane/Shoulder Drop Off (L.m)|
|10||Long & Trans Cracking (L.m)|
|11||Patching (Sq. m)|
|12||Polished Aggregate (Sq. m)|
|14||Railroad Crossing (Sq. m)|
|15||Rutting (Sq. m)|
|16||Shoving (Sq. m)|
|17||Slippage Cracking (Sq. m)|
|18||Swell (Sq. m)|
|19||Weathering & Raveling (Sq. m)|
Table 1 Distresses Encoding Based on PAVER System
A comprehensive visual distress survey has been carried out for the defined sections. All distress types are recorded in a special form to be entered in a computer program for computing the PCI values. One of the most important steps in the distress survey is to create a “Distress Map” for each sample. The distress map gives a complete picture of the existing pavement surface distresses and their locations. This distress map has been digitized into GPMA as spatial data. Nodes have also been inspected.
7.3.3 Materials Investigation
The purpose of the material investigation is to evaluate the condition of the existing pavement layers based on visual observation and scrutinizing the cores and test results of cut-out sample obtained from the area in order to help the decision maker in finding the final maintenance requirement. After material investigation, the final maintenance decision might be changed than the pre-defined recommended maintenance depending on the condition of pavement materials. The material investigation includes the following:
- Scope of Works
- Laboratory Exercise: for the following properties:
- In-situ density
- Sieve Analysis
- Moisture Density Relation
- Sand Equivalent
- Plasticity Index
- California Bearing Ratio
- Test Results
- Analysis & Comments
8. Data Analysis
8.1 Data Analysis for Network-Level
One of the most important benefits of the PMMS program is the ability to relate between the network condition and its maintenance needs. In this regard, it is important to differentiate between two main outputs, which are the needs and the program or plan. Normally, the needs come close to the program or the plan whenever the budget is exceeded until they match together as in case of open budget. This situation is rare to happen; therefore the identification of the maintenance plan or program under specific budget necessitates complicated mathematical procedures to achieve the balance as possible between the network condition and the available budget, i.e., ideal exploitation of the available budget.
Therefore, in this project it has been decided to generate a maintenance program through three maintenance stages. The budget available for a 2-year plan was 100-million Dirham, which was allocated to fulfill the maintenance needs of the first maintenance stage. Second and third maintenance stages were deferred to be considered in the next 2-year plan.
Maintenance priority setting techniques, using a simplified worst first scheme (ranking), has been applied to all surveyed sections to come up with the three stages of maintenance program. Generating thematic maps using GPMA has been introduced in this task to assist the phasing process and support maintenance priorities setting. Figure 11 and Figure 12, GPMA output, show the final setting of the maintenance program through three stages for the main and secondary roads, respectively. The black color represents the first stage, the dark-gray color represents the second stage, and the gray color represents the third stage.
Figure 11 Maintenance Program Stages for main roads
Figure 12 Maintenance Stages for Secondary roads
8.2 Data Analysis for Project-Level
Regarding data collection, 100% of the first stage and 65% of the second stage have been completed, so far. After data collection, editing, and saving, the data analysis operation started as follow:
- PCI Calculations
- Common Distresses
- Distress Distribution
- Structural and Non- Structural Failure
8.2.1 PCI Calculations
The reason for computing the PCI values is to take a general picture about the pavement surface for rating. Maintenance decisions DO NOT rely on the PCI values, since in some cases; it does not represent the pavement condition [Abo-Hashema, 2000]. All collected data have been entered in the GPMA database. Figure 13 shows an example of the PCI distribution per sample for an inspected section of a main road located in Al Ain Roads Network.
Figure 13 PCI distribution per Sample for a Main Road
8.2.2 Common Distresses
Figure 14 shows an example of distress frequency for a right lane in an inspected section of a main road located in Al Ain Roads Network.
Figure 14 Distress Frequency for a Right Lane of a Main Road
8.2.3 Distress Distribution
Figure 15 shows an example of distress distribution in Khalifa Ibn Zayed Al Awwal Main Road. Notice that the shown figure is based on the “distress map” generated through the data collection phase and after manual digitizing into GPMA. Furthermore, using the advanced tools for interactive visual zoom this level of details is displayed. On the other hand, Figure 16 introduces enhanced representation of distress distribution in the form of symbols (Pinpointed) with descriptive data.
The distribution of each distress can be also presented in a bar-chart graph form, as shown in Figure 17.
Figure 15 Distress Distribution (Digitized Distress Map)
Figure 16 Distress Distribution (Symbolic Distress Map)
Figure 17 Rutting Distribution for a Right Lane of a Main Road
8.2.4 Structural and Non-Structural Failure
Structural Failure is account for load-associated distresses such as Alligator Cracking, Corrugation, Depression, Edge Cracking, Potholes, and Rutting. Figure 18 presents an example of the distribution of structural failure and non-structural failure per sample unit for an inspected section of a main road located in Al Ain Roads Network.
Figure 18 Distribution of Structural and Non-Structural Failures per Sample for a Main Road
9. Maintenance Decisions
In this project, a maintenance decision criterion is based on the density of required localized maintenance (patching, crack sealing, etc.) not on the basis of the distress density (density is equal to the area of the distress divided by the total area of an inspected section). “The main concept behind the applied criterion is to significantly reduce the number of associations between distress information (type, severity and density) and required M&R alternatives. This way, the applied decision system addresses the complex combinations between the distress types and maintenance alternatives through an intermediate level, the localized repair requirements” [Abo-Hashema, 2004].
Table 2 shows the maintenance needs per sample for an inspected section of a main road located in Al Ain Roads Network depending on the density of localized maintenance. These maintenance needs are considered the first approach in obtaining the final maintenance decision. This preliminary maintenance needs is obtained from the Data Analyzer, which is a costumed program developed on the basis of the density of localized maintenance.
The final recommended maintenance might be changed than these maintenance needs based on the material investigation and/or road geometric assessment. Surface treatment could be either slurry seal or sand seal or fog seal or milling & overlay, as an example. For any type of surface treatment, existing distresses must be maintained, i.e. Distress-By-Distress repair. If there is milling & overlay, surface preparation should be done.
GPMA provides an interface for receiving analysis results from the Data Analyzer. GPMA supports the final maintenance decision using the excellent displaying and querying tools it offers, and by overlaying received data with the material investigation and/or road geometric assessment results. In a final step, GPMA archive the maintenance decisions, maintenance reports, final work-orders, distress maps, multimedia files (distresses pictures, material cores pictures, etc.), and all relevant data with a spatial link for future historical review.
Table 2 Maintenance Needs (Decisions) per Sample Unit
The objective of this research is to present the effort made so far in integrating pavement maintenance management activities and Geographic Information Systems (GIS) for Al Ain roads network through the current project “Maintenance of Roads and Bridges in Al Ain Region”. The implicit goal of using GIS is presenting new techniques for archiving and analyzing of pavement maintenance data as well as supporting and improving decision-making. Network identification was the first step then data collection procedures have been done through two levels of study: Network and Project Levels. The purpose of the network-level analysis is to come up with maintenance plan for main and secondary roads. In project-level analysis, more detailed information is needed to finalize the list of projects, to provide a detailed scope-of-work for each individual project, and to issue work orders. The effort framework was governed by the GIS-Based Pavement Management Application (GPMA) that has been developed in this project. Looking ahead includes adding the emergency maintenance activities, automating distress survey, and automating data entry.
- AASHTO, 1986, “AASHTO Guide for Design of Pavement Structures”, American Association of State Highway and Transportation Officials, Washington, D.C.
- Abo-Hashema, M.A. and Sharaf, E.A., 2000, “A simplified Pavement Maintenance Decision Tree for Use in Developing Countries”, Proceedings of Al-Azhar Engineering Sixth International Conference (AEIC), Vol. (4), pp 308-339, 1-4 September, Faculty of Engineering, Al Azhar University, Cairo, Egypt
- Abo-Hashema, M.A. and Sharaf, E.A., 2004, “MU Method: A Simplified Maintenance Decision System for Selecting the Maintenance Strategies in Flexible Pavements”, 10th World Conference on Transport Research (WCTR04), July 4-8, Istanbul, TurkeyArcGIS, 2004, “ArcGIS® Solutions for Civil Engineering”, ESRI Publications, www.esri.com
- Haas, R. and Hudson, W.R., 1978, “Pavement Management Systems” McGraw-Hill
- Haas, R., Hudson, W.R., and Zaniewski, J., 1994, “Modern Pavement Management”, Krieger Publishing Company, Malabar, Florida, USA
- HPMA Manual, 2001, “Highway Pavement Management Application (HPMA) User Documentation”, V4.50, GARBLT Research Project, Technical Consultations Bureau & Applied Engineering Technologies (TCB/AET), Infrastructure Management Center of Technical Excellence (Stantec Consulting)-Cambridge, ON, Canada, Egypt.
- Pinard, M.I., 1987, “Factors Affecting the Development and Implementation of Pavement Management Systems in Developing Countries”, Second North American Conference on Managing Pavements, Proceedings, Volume 1 Toronto, Ontario, Canada, November 2-6, pp. 300-310.
- Shahin M.Y., and Kohn S.D., 1981, “Pavement Maintenance Management for Roads and Parking Lots”, United States Army Corps of Engineers, Technical Report M-294.
- Sharaf, E.A., Fahmy, M. M., Abo-Hashema, M.A., 2003, “Pavement Maintenance Priorities Using Ranking and Optimization Models”, The 2nd International Operation & Maintenance Conference in the Arab Countries (OMAINTEC), 30 June – 3 July 2003, Phoenicia Inter-Continental, Beirut, Lebanon
- Sharaf, E.A., Abdel-Ghany, M., Abo-Hashema, M.A., 2003, “Development of a Pavement Management System for Egyptian Roads Network”, Proceedings of Al-Azhar Engineering 7th International Conference (AEIC), 7-10 April 2003, Faculty of Engineering, Al Azhar University, Cairo, Egypt
- TCB/AET, 2001, “Development of Pavement Maintenance Management System using Geographic Information Systems (GIS) in Egypt”, Final Report of GARBLT Research Project, Technical Consultations Bureau & Applied Engineering Technologies (TCB/AET), Stantec Consulting-Canada, 2001, Egypt
Authors are grateful to Al-Ain Municipality’s staff who contributed in this research. Special appreciation is to Professor Dr. Essam Sharaf, who was working in Al-Ain Municipality as Pavement Maintenance Expert, and Engineer Nasser Al-Ariani, who is working as Head of Roads Design Section. This research was developed at WS ATKINS & Partners Overseas (Consultant) in conjunction with Al-Ain Municipality, UAE (Clint).