30-01-2013, 04:43 PM
Internet-based database management system for project control
1Internet-based database.pdf (Size: 428.99 KB / Downloads: 57)
Introduction
Database management systems (DBMSs) are essential in supporting project tracking
and control functions. A database provides a platform to organize, store and retrieve
the planned and actual performance data of projects in a logical and efficient manner.
The DBMS queries the stored project data using SQL (structured query language) to
generate different management reports for control purposes. It follows that the design
of the database should follow a well-defined structure to support the tracking and
control of individual tasks at different levels of reporting. The data structure should
also facilitate the linkage of those individual tasks to their respective construction
trades. A work package model is commonly used to describe the data structure of a
project. The work package is a general expression that represents a well-defined scope
of work that usually terminates in a deliverable product. It includes activities and tasks
within those activities as depicted in the work breakdown structure (WBS) shown in
Figure 1.
Overview of the developed system
The developed system consists of seven main components:
(1) user interfaces;
(2) input/edit;
(3) evaluation;
(4) reasoning;
(5) forecasting;
(6) reporting; and
(7) databases.
The system is shown in Figure 2. The user interface provides a viewing/input window
that allows users to interact with the system through the worldwide web using an
internet browser; the input/edit is a data input, update and delete component that is
developed to facilitate data entry and edit; the evaluation component assesses the cost
and schedule variances using the earned-value method (Canadian General Standards
Board, 1999; Department of Defense, 1967); the reasoning component diagnoses
possible reasons behind the variances through a set of built-in indicators; the
forecasting component predicts the project’s time and cost at completion; the reporting
component generates the performance reports on a daily, weekly, monthly, and/or
yearly basis. The reports have a tabular format as well as a graphic format that can be
generated in a user-friendly manner. Two relational databases (project and historical)
are developed. The project database stores planned and actual cost and schedule data
for the project being considered. Upon completion of the project, all information
collected in the project database is transferred to the historical database. The historical
database has the same structure as the project database.
Developed database
Data structure is essential to the development of an efficient database. This is
particularly important in supporting project control functions. Project control is carried
out using a set of control objects as proposed by Moselhi et al. (2001, 2002). The control
object, which is shown in Figure 1, could represent a phase of a project, a work
package, and/or a cost account defined using cost breakdown structure (CBS). It has
the resources necessary to complete the tasks included within that control object.
Based on this, all of the project data can be treated as an aggregation of a pre-defined
set of control objects. Each control object has its resources of labor, material,
equipment, and sub-contractors. In the proposed system, sub-contractors are treated as
a type of resource. Each resource has a budgeted value and an actual consumed value.
Each control object has relation(s) to other control objects as well as its own method of
resource allocation. It also has attributes that describe its characteristics, such as
sensitivity to weather and site congestion, as well as a set of threshold values that
defines unacceptable performance. Each type of resource in a control object may have
single or multiple sub-resources. It should be noted that each control object is an
abstract representation of a physical component of a project, as stated earlier. The
budgeted resources of a control object serve as a control reference as they are actually
consumed over the project duration.
Implementation
A three-tier client-server computer architecture is used to implement the developed
system, as shown in Figure 2. It involves the presentation tier, the application
logic/middle tier, and the data tier. The user interfaces are the components of the
presentation tier, which handles the system’s communication with the user. The
databases are components of the data tier. The input/edit, evaluation, reasoning,
forecasting, and reporting components belong to the application logic/middle tier.
Adjacent tiers are connected through the internet.
The database as described in the ER diagram shown in Figure 3 is located in the
data tier. It is implemented using the Microsoft Access 2000 environment (see the
tables and relationships shown in Figure 4). In essence, these tables map the entities
and their respective relationships. The data type of the primary key in the entity tables
is “auto-number”, which avoids the redefinition of the key. The attributes of the entities
and relationships record the budgeted and actual data described below.
Summary and conclusions
An internet-based database management system has been presented for the tracking
and control of construction activities. The paper has focused primarily on the design
and implementation of the system’s relational database. The developed system can
generate earned-value based project status reports at user-specified reporting dates.
Fifteen entities and 20 relations exist in the developed database. By taking advantage
of the worldwide web, the system provides a real-time data sharing environment and
accordingly supports the generation of timely site progress reports. A numerical
example is presented to demonstrate the capabilities of the developed system. The
earned-value report generated by the developed system was identical to that of
Abudayyeh (1991; Abudayyeh and Rasdorf, 1993) when threshold values of time and
cost variances were set equal to zero. In addition, the developed system integrates the
entire project data using the object model proposed by Moselhi et al. (2002, 2004). The
developed database not only considers the sequence of the tasks performed, but also
the relationships among the control objects. It supports project-tracking function of
progress reporting, using earned-value method. It can also be applied to support
forecasting time and cost at completion as well as reasoning about unacceptable
performance.