COVER SHEET

Yang, Jay and Lie, Heng and Skitmore, Martin R. (1996) Expert Systems in
Construction Management: Is the Hype Over? In Proceedings INCIT 96,
International Construction Information Technology Conference, pages
pp. 131-136, The Institution of Engineers Australia, 11 National Circuit,
Barton, ACT 2600.
Accessed from: http://eprints.qut.edu.au/archive/00004546
 EXPERT SYSTEMS IN CONSTRUCTION
MANAGEMENT: IS THE HYPE OVER?
Jay Yang, Heng Li and Martin Skitmore

ABSTRACT

Because of its characteristic attributes of experience based problem-solving and qualitative data processing,
construction management has been a target for expert system development since the late seventies. The eighties
saw a hype of development of expert systems. The trend continued until recent few years when development
problems accumulated and new techniques emerged to assist decision making. This paper presents a review of
expert system applications development in the construction management area and in which four major types are
identified comprising selection systems, advisory systems, monitoring and control systems, and analysis and
evaluation systems. The reasons why expert systems seem to be out of favour are examined and typical problems
experienced by earlier systems are analysed. As a result, four major types problem areas are identified comprising
knowledge elicitation, knowledge representation, limitation of development tools, and ES development processes.

Finally, the trend of development of knowledge based applications for construction management tasks is
identified, possible areas of improvement are explored, and future concentration areas are suggested for new
expert system applications in the construction management domain.

Keywords: Expert systems, construction management, information processing.

INTRODUCTION

Expert systems (ES) research has been one of the most exciting, and potentially significant, themes in
construction management for more than twenty years. In fact, few other computer based techniques have
attracted the same level of excitement and expectation. Over a decade ago, knowledge based processing,
especially ES, was seemed to be applicable in just about every facet of construction management tasks - ranging
from site selections to risk analysis, and from construction scheduling to contractor pre-qualification. However as
with any other emerging technology, the expectations surrounding ES have been slow to come to fruition. Also,
and perhaps as a result, there has been a considerable drop in the number of new ES applications in recent years.

Have expectations been too great? Maybe. In an article titled "Expert Systems - After the Hype is Over",
Brandon (1990a) discussed the hype of ES development in the late eighties and predicted the `cool-off' period
now being experienced. In many research areas, ES seem to have quietly given way to the new breed of computer
applications, such as integrated information processing and neural networks. On the other hand, the construction
practice, under current pressure of stiff competition and low profit margins, constantly requires more efficient use
of all alternative information technology. So the questions arise: Is the hype really over? Have ES finished their
lifecycle in construction management, or is their yet more to come?

In this paper, the nature of ES is examined and their past applications for problem solving and information
processing in construction management. The reasons are investigated why ES, at their previous state of the art,
have not met practical expectations. The latest technology development in ES methodologies is summarised and
possible improvement areas are suggested together with research concentration and exploration areas for future
ES development. Some of the authors' past experience of knowledge based processing and perceptions on ES
applications are also presented as case studies.

OVERVIEW OF TYPICAL EXPERT SYSTEMS FOR CONSTRUCTION MANAGEMENT

All ES are concerned exclusively with decision making. As a result, construction management, which is
concerned with both decision making and implementation, is a very likely looking customer for ES applications.
It is therefore not surprising that, although the earlier ES seldom progressed beyond the prototype stage, the
amount of work in the field and range of application areas is impressive.

Over the last two decades, there has been ES research and development on all of the better defined construction
management tasks. The following overview covers most of the documented ES building attempts in construction
management. For convenience, these are classified into four categories, comprising:

· selection systems
· advisory systems
· monitoring and control systems, and
· analysis and evaluation systems.

Selection systems

ES are particularly useful in the decision making process where the various decision options are clearly defined but
the method of choosing among the options is, at best, of heuristic nature. Prequalification provides a classic
example of this as a great deal of objective data is collected on potential contractors but its application is very
much intuitive. QUALIFIER 1 and QUALIFIER 2 (Russel et al, 1990) are ES developed specifically for the
prequalification of the construction contractors. They are capable of taking account a range of evaluation criteria
such as a contractor's reputation, past performance, operation capacity, current workload, financial stability and
technical expertise both in qualitative or subjective format. The two systems were developed in a commercial ES
shell KES, using a backward-chaining inference mechanism.

Because of a similarly strong heuristic and reasoning approach, the selection process of construction equipment
and facilities has been an area particularly favoured for ES development. Sauce and Mommessin (1995)
developed an ES for the selection and organisation of construction sites while ESSEX is a knowledge based
system which facilitates the selection of an appropriate estimation software package (Arditi and Suh, 1991). Other
examples of these types of system include formwork selection (Hanna et al, 1992), project procurement (Mohsini,
1993) and equipment selection (Amirkhanian and Baker, 1992).

Advisory systems

With the consultation format widely adopted, the objective of advisory ES is to provide pre-determined
recommendations in a problems area and to cover as many scenarios as the construction expert can envisage.

One of the better known example in this category is Brandon's (1990b) ES ELSIE, for advisory service in the
quantity surveying domain. It contains four individual consultation modules for problem-solving in financial
budget, procurement, project time and development appraisal. On-going development allows the system to
interact with other ES in the problem domain.

ESCHEDULER is an ES for planning and scheduling construction projects (Mosehi 1990). Developed in
GURU, an ES shell, it retrieves information from duration of all activities to be performed in a building and
generates an as-possible schedule taking account a range of project planning parameters. The strength of the
system is its ability to interlink with several commercial packages and has the facility to use both heuristic
methodology and algorithmic methods during its problem solving process.

Other systems have been developed to assist decision making for public sector design-build projects (Songer et al,
1992), decision support on concrete mix design (Bai and Amirkhanian, 1994), and for the determination of
feasibility of project modularization (Murtaza et al, 1993).

Monitoring and control systems

ES for monitoring and control of earthmoving scraper operation have been developed at Stanford University
(Paulsen and Khoo, 1987). These are designed to specify a fleet of earthmoving equipment of a given project, aid
operators to understand optimum loading times for each machine, and optimise fleet production by
communicating between machines in real time.

GRAFFITI is another example system of this type. Employing a reflective architecture between the knowledge
based system and the geometrical modules, it allows a permanent graphical visualisation of the construction work
in progress (Paul et al, 1992).

Analysis and evaluation systems

SEA is an ES for contract claims analysis (Diekmann and Gjertsen, 1992). It helps determine which of several
claim types could flow from some unusual events occurring at a construction site. The characteristics of events
and nature of potential disputes are analysed prior to the system suggesting the type of claims likely to arise.

Yates (1993) has developed a knowledge based system for decision support in the process of construction delay
analysis. Named DAS (Delay Analysis System), the system is capable of determining possible causes of project
delays and suggests alternative causes of action to prevent further delays. A main feature is the system's inter-
linkage with several project management and support packages.

Other examples of ES in this category include: SIPE (Kartam et al, 1991), a system which generates project
network information; SUPERCHANGE (Diekmann and Kim, 1992), a system for the analysis of change order
claims; and a prototype system for construction planning and productivity analysis (Boussabaine, 1995).

A STUDY OF ES PROBLEMS AND LIMITATIONS

Although the research and development of ES has been among the most frequently reported of computer based
topics in construction management, little study has been made of application problems and limitations. In the
'hypegiest', ES evaluations have concentrated solely on documenting available systems and their application status
and potential. Critical work in the field has been virtually non-existent (Kangari, 1988; Mohan, 1990; Irrgang et al,
1993). The literature on specific ES applications, always biased in favour of ES, gives no count of the 'grey' areas
of the development. Only in observations of earlier ES applications, and follow-up on their final outcome by
interested researchers, are some of the problem areas revealed.

The following discussion on ES problems is based on available literature, observations, interviews with industry
professionals and lessons learned from direct experience of the authors over the last ten years. The common
problems exist in four specific areas, namely:

· knowledge elicitation
· knowledge representation
· limitation of development tools, and
· ES development processes.

Knowledge elicitation

The knowledge embodied in an ES is the most important element since the system is only as good and capable as
the knowledge contained within it. This knowledge is obtained through a process of elicitation, where
construction experts' knowledge is collected, analysed and represented in a form suitable for inclusion in a
computer format.

For much ES development, a major challenge has been to ensure the availability and authenticity of construction
management knowledge. Identifying, recruiting and retaining genuine construction experts has proved to be most
difficult, and often ignored, task (Adeli, 1988; De La Garza and Ibbs, 1990). Individuals, or 'knowledge tzar', with
expertise in the whole problem domain are extremely rare. More usually, a group of experts with partially
overlapping expertise have to be used, creating further problems where those in the group disagree.

Even when experts have been available, many ES developments have suffered from poor knowledge elicitation
resulting from:

· an inappropriate problem domain - too narrow or too wide domains have made the problem solving
process rather difficult to represent.

· a confusion of knowledge with beliefs - knowledge is assumed to be factually correct while beliefs may or
may not be so. Builders of belief based systems therefore run the risk of perpetuating the inherent
problems of human based systems, albeit in a much faster and more efficient way!

· limited sources of knowledge - sixty percent of ES applications rely on knowledge extracted mainly from
literature (Mohan, 1990).

· lack of sufficient analytical skills - the knowledge extracted from construction experts have not been
rigorously analysed, evaluated, checked and tested for completeness before being formulated into
production rules or cases. This has caused some of the systems to operate with restricted subset of all of
the possible decision-tree paths.

· use of preferential knowledge - the construction knowledge obtained has not been refereed to different
experts and wider user groups for consensus and feedback, resulting in the limited application and
acceptance of developed ES by industry as a whole.

Knowledge representation

The representation of extracted knowledge has involved many different research efforts. Over the years, there
have been `rule-based representation', `framework-based presentation', and more recently `object-oriented or
case-based presentation' (Li, 1995). It is noted that most of previous applications adopted the 'rule based'
approach (Allen 1992). While this technique offers ease of development and good explanation capacity, it suffers
from two serious setbacks. The first is an over-dependence on knowledge extraction process which is subjective,
tedious and labour intensive as discussed above. The second is the lack of flexibility for knowledge upgrade and
information updating. This situation, in turn, has confined many systems to operate only within the bounds of
originally encode knowledge and which can cope with only a limited number of scenarios.

Development tools

Over the past decade, the tools adopted for developing ES in the construction management area have evolved
through three major types. The literature suggests that applications prior to the mid eighties mostly employed the
programming languages ranging from the generic FORTRAN and PASCAL to the more AI oriented LISP and
PROLOG. However, the deep knowledge required for using these languages and their cumbersome
programming procedures resulted in a limited the number of applications (Allwood, 1989).

The late eighties ES boom was partly prompted by the wide availability of ES shells, with preformatted ES
features and production structures. A much reduced learning curve and ease of operation were among the many
advantages as most construction management ES adopted this type of tools (Mohan, 1990; Arockiasamy, 1993).
Their dominance continued until two or three years ago. While there are still occasional development efforts
using this type of tools, the complex nature of construction management tasks, and the amount of information
and knowledge to be incorporated, surpass the handling ability of most available ES shells.

Object-oriented relational database systems, with their excellent data handling and presentation capacity began to
emerge in ES development after 1989. With ample power to program complex ES features, they specifically offer
the advantage of knowledge representation in database format and the wide inter-connectivity to other software
packages (Adams, 1993; Darwood 1994). There are downsides, however, including the lack of explanation
facilities and difficulties in assembling information to cope with difference knowledge representation modes.

Development processes

In addition to the inadequacy of tools as discussed above, many previous developments involved construction
management researchers who not only worked to extract and represent construction management knowledge, but
also attempted to learn to use a development tool (such as a shell) and code ES programs. It is obvious that this
exercise is time consuming, inefficient and results in the lack of a common standard of finished products. System
quality and user-friendly features have often been sacrificed.

Another area which attracts criticism is the development approach itself. Most of early ES development in
construction management were centred around their technology advantages. In other words, system developers,
generally consisting of construction experts and knowledge engineers, focus on the extraction of construction
knowledge, knowledge representation techniques, tools used and inference processes adopted while paying little
attention to what potential users would like to see in the development. Many researchers have consequently
realised that ES applications, subject to the quality and acceptability of their embodied knowledge, are of `client
centred' nature (Basden, 1989; Brandon and Watson, 1994). The missing link between the researchers and their
clients has caused serious problems for ES to be fully recognised and utilised even within the organisation for
which they are developed (De La Garza and Panagiotis, 1991; Oloufa, 1994).

POSSIBLE IMPROVEMENT AREAS OF EXPERT SYSTEM

The on-going development of construction knowledge

It is commonly agreed that the representation of construction knowledge in ES is a difficult and cumbersome
process. To a large degree, this is caused by the fact that the construction knowledge at its present state, has not
been accumulated, documented and processed in a way suitable for common access and efficient use generally.
Recognition by a wider range of people (not just system's developers and specific users) will also be a major
advantage (Brandon, 1992). There is an unofficial claim that the name of `expert' has been dropped from many
ES applications because of the failure for most systems to acquire sufficient knowledge from construction experts
and simulate their problem solving process efficiently.

Construction management knowledge also requires consistent re-definition and development as new materials and
construction methods are brought into the industry. A fully developed knowledge body within a domain-specific
area of construction management process will become an essential requirement for future ES development. For a
task to be efficiently modelled after the ES structure and methodology, it must be 'sufficiently narrow, self-
contained and should not require knowledge from a large number of areas' (De La Garza and Mitropoulos, 1992).
`Project-specific' and `area-specific' knowledge are expected to receive increasing attention within the context of
knowledge based presentations.

Hardware and Software

ES building tools play a major role in ES development. Refined ES shells, with more dynamic user interface and
abilities to exchange information with commercial software packages, are likely to be in high demand. Various
object-oriented programming languages with relational databases or database compatibility (such as Visual Foxpro
and Visual Basic) are being used and continue to add to the list of available development tools. This is aided by
the increasingly powerful microcomputers and workstations supported with extensive networking facilities.
Construction management, with particular requirement of information processing and team work, may better
receive ES developed within such an environment.

Methodology
Recently there have been an increasing number of new methods and techniques being adopted for new system
development. The usefulness of the object-oriented approach in construction management areas has been
demonstrated (Bridgewater and Atkin, 1994; Kim et al, 1994; Elzarka and Bell 1995). The use of hypertext, a
method of text retrieval using nodes and links, seems to be a valid and efficient alternative to ES methodologies
(Williams, 1991; Abourizk and Chehayeb, 1995). Another major methodological advance is in the use of neural
networks, which employ interconnected network of many simple processing units capable of self-organisation and
learning - a possible paradigm for ES (Williams, 1994; Chao and Skibniewski, 1995; Crespo and Mora, 1995).
Therefore, the indications are that a new technological generation will not simply replace ES or be isolated from
them. Rather, an intuitive adaptation and appropriate use of these technologies may find solutions to problems
which have so far constrained earlier ES applications.

POTENTIAL CONCENTRATIONS OF ES DEVELOPMENT

For progress to be made, future ES applications need to reflect the lessons learned from previous development
while embracing new technologies as they become available. From such a perspective, the following three areas
of concentration for future ES development are envisaged:

ES development in `niche' areas

New ES development may concentrate in refined and somewhat 'niche' areas of construction management.
Researchers and practitioners attempting to develop ES applications are advised to use a `gazetting' process to
eliminate those areas which, though experience based, have an incomplete knowledge set, or lack of problem
solving procedures to a common standard. New systems are more likely to model those topics where there is
sufficient knowledge to be embodied and the domain of problem is narrow enough to be imitated. Possible
focuses include materials management, asset management, quality control, company finance (cash flow), and
constructability evaluation. Construction process modelling and simulation techniques coupled with ES as
decision support are also likely to be an area of interest, provided the problem domain is appropriate in size to
model.

Flexible knowledge presentation

It is argued earlier that a major setback for many previous ES developments was caused by an inflexible
knowledge presentation. Unless the knowledge within an ES is upgradable, the system is unlikely to be adopted
by wider user groups or will soon become obsolete.

Some pilot research work has been done in the area of flexible knowledge presentation (Scott and Yang, 1991;
Fenves, 1991). As a result of this, it now seems likely that the knowledge base of future generation of ES will
consist of two components: (a) a core knowledge base containing the standard construction management
knowledge and (b) several custom knowledge bases containing individual work routines and heuristic which can
be modified and updated. These systems, therefore, are adaptable to different environment by various users, each
of which may have individual work requirements and follow individual company operational procedures.
PACES, an ES for construction project analysis and control, was developed to initiate and test this technology
using a relational database. Along the consultation path of project feasibility analysis and planning, the system
links to project scheduling packages and spreadsheets to maintain the flow of information and maximise user-
friendly features (Yang, 1991). In addition to its upgradable knowledge base, the production rules can be
relocated and regrouped should the inference process require a difference approach.

Future development are also likely to concentrate on equipping ES with learning capabilities (Mikami et al, 1994).
Tools with learning features, such as neural networks, can be utilised in conjunction with other knowledge
representation techniques for this type of development.

Information processing in ES
Unlike other application areas, such as the detection of welding defects, construction management procedures are
dynamic, consistent and integrated. ES in this area do not work well as stand-alone systems. Moreover, they need
to be ready to take on data processed from other areas in different formats. They also need to be able to
communicate and make use of a variety of computer packages such as databases, spreadsheets, cost estimation
and project scheduling packages and other supportive systems (Teicholz, 1994). While the new breeds of ES will
hopefully be more adaptable and`friendly' to other systems, there is a need for a knowledge based software which
acts as a central control system to other systems for databases, calculations, graphics, and area-specific ES, and
which handles all communications between these systems and coordinates the action sequence and changes in
construction management processes (Yang, 1994).

Another major development area is in the representation of 3-D graphical information within ES. Previous ES
developments for project planning and site operation have suffered because of the lack of such facilities
(Echeverry et al, 1991; Mahoney and Tatum, 1994). New representation techniques are needed to enable users to
pass graphical information, from a CAD system for example, to the ES. This will help to provide answers to
questions such as, `what is the lay-out of the project site?' and 'what are the physical constraints to the movement
of a tower crane on site?'. Other potential areas of integrating CAD with ES include formwork design, project
planning and construction simulation and constructability studies.

ES and Computer Integrated Construction

Computer Integrated Construction (CIC) is an important research application of considerable potential in the
construction industry. The CIC concept relies on the various computer based systems working on individual
problems in different areas, while maintaining comprehensive links, feedbacks and coordination between these
systems. Various CIC models have been suggested (Kunz et al, 1994; Miyataki and Kangri, 1993). In future, ES
are very likely to become one of the components of CIC models and handle qualitative aspects of specialised
procedures (eg, procurement analysis and contract administration). CIC oriented technology - combining ES,
construction robots, integrated project management software and office automation - is likely to form the basis of
future construction engineering and management systems.

CONCLUSION

Expert systems for construction management have experienced a hype of development and attention across the
industry as no other computer based application. While they have made some impact on current construction
practices, the expectations are slow to be realised as a result of the limited capacity of development tools,
inadequate methodologies and absence of construction knowledge in problem domains.

The potential of ES lies within their ability to capture, incorporate and automate judgements, intuition, rules of
thumb and other forms of human expertise that are difficult to model using conventional techniques. New areas
of exploration, such as the learning abilities and flexible knowledge representation, are emerging. By their very
nature, both now and in the future, ES are likely to continue to help construction organisations make efficient use
of human resources, corporate knowledge and management procedures. Stand-alone ES, with a fixed knowledge
presentation and `rule based' inference structure, no longer seem to be appropriate to suit changing construction
management tasks. Future expectations are based on a new breed of knowledge based systems which offer an
cooperative, multi-tasking and integrated decision making environment.

ES and decision support systems have made a significant impact on the practice and thinking of construction
management and have a specific role in an area where knowledge, experience and qualitative approaches still
dominate. The hype may be over, but ES are here to stay.

REFERENCES

Abourizk, S. and Chehayeb, N. (1995) A Hyper Text Decision Support Model for Contractor Prequalification.
Microcomputers in Civil Engineering, 10(2), 111-121.

Adams, T. (1993) Knowledge representation and processing in relational databases. Journal of Computing in Civil
Engineering, 7(2), 238-256.

Adeli, H. (1988) Expert systems in construction and structural engineering, Chapman and Hall, New York.

Allen, R. Eds (1992) Expert systems for civil engineers: knowledge representation. Published by ASCE, New
York.

Allwood, R. (1989) Techniques and Applications of Expert Systems in the Construction Industry, Ellis Horwood
Limited, Chichester, UK.

Amirkhania, S and Baker, N. (1992) Expert system for equipment selection for earthmoving operations. Journal of
Construction Engineering and Management, 118(2), 318-331.

Arditi, D. and Suh, K. (1991) Estimating system selection expert, Construction Management and Economics, 9(4), 369-
381.

Arockiasamy, M. (1993) Expert systems applications for structural, transportation and environmental engineering,
CRC Press, Inc. Florida.

Bai, Y. and Amirkhian, S. (1994) Knowledge based expert system for concrete mix design, Journal of Construction
Engineering and Management, 120(2), 357-373.

Basden, A. (1989) A client centred technology for building expert systems, People and Computers, VA Sutcliff
and L MaCaulay, Eds. Cambridge University Press, UK.

Boussabaine, A. (1995) An expert system prototype for construction planning and productivity analysis,
International Journal of Construction Information Technology, 3(2), 49-64.

Brandon, P. and Watson, I. (1994) Professional involvement in development of expert systems for construction
industry. Journal of Computing in Civil Engineering, 8(4), 331-342.

Brandon, P. (1990a) Expert systems - After the hype is over. Volume 2, Design Economics and Expert Systems,
Building Economics and Construction Management - Proceedings of CIB90, UTS, Sydney.

Brandon, P. (1990b) The development of an expert system for the strategic planning of construction projects,
Journal of Construction Engineering and Management, 120(3), 561-578.

Brandon, P. (1992) Research and development of expert systems for the construction industry, Building Research
and Information, 20(3), 162-165.

Bridgewater, C. and Atkin, B. (1994) Applications of artificial intelligence techniques to component based
modular building design, Journal of Computing in Civil Engineering, 8(4), 469-488.

Chao, L. and Skibnewski, M. (1995), Neural network method of estimating construction technology acceptability,
Journal of Construction Engineering and Management, 121(1), 130-142.

Crespo, J. and Mora, E. (1995) Neural networks as inference and learning engines, Microcomputers in Civil
Engineering, 10(2), 89-98.

Dawood, N. (1994) Developing an integrated bidding management expert system for the readymix concrete
industry, Building Research and Information, 22(2), 95-102.

De La Garza, J. and Ibbs, C. (1990) Knowledge elicitation study in construction scheduling domain, Journal of
Computing in Civil Engineering, 4(2), 125-154.

De La Garza, J. and Mitropoulos, P. (1992) Flavours and mixins of expert systems technology: transfer model
for AEC industry. Journal of Construction Engineering and Management, 118(2), 435-453.

De La Garza, J. and Panagiotis, M. (1991) Technology-transfer model for expert systems, Journal of Construction
Engineering and Management, 117(4) 736-755.

Diekmann, J and Gjertsen, K. (1992) Site event adviser: expert system for contract claims. Journal of Computing in
Civil Engineering, 6(4), 472-477.

Diekmann, J. and Kim, M. (1992) Superchange: expert system for analysis and changes claims. Journal of
Construction Engineering and Management, 118(2), 399-413.

Echeverry, D., et al (1991) Knowledge based approach to support the generation of construction schedules,
Computer and Structures, 40(1), 59-66.

Elzarka, H. and Bell, C. (1995) Object-oriented methodology for materials-management systems, Journal of
Construction Engineering and Management, 121(4), 438-446.

Fenves, S. and Ibarra-Anaya, E. (1991) A framework for customizable knowledge-based expert systems,
Computing in Civil Engineering, Proceedings of the 7th Conference, ASCE, Washington, 298-307.

Hanna, A. et al (1992) Knowledge acquisition and development for formwork selection system, Journal of
Construction Engineering and Management, 118(1), 179-199.

Irrgang, R. et al (1993) Intelligent systems in AAustralian industry: a study of current status and future need,
Division of Information Technology, CSIRO, NSW.

Kangari, R. (1988) Expert systems in construction management: microcomputer applications, Proceedings of the 5th
Conference on Computing in Civil Engineering, ASCE, USA, 669-678.

Kartem, N. et al (1991) Extending artificial intelligence techniques for hierarchical planning, Journal of Computing in
Civil Engineering, 5(4), 464-477.

Kim, K, et al (1994) An object-oriented scheduling data acquisition model using hybrid expert system, Computing in
Civil Engineering, Proceedings of the first congress held in conjunction with A/E/C Systems 94, ASCE, 356-367.

Kunz, J. et al (1994) Circle integration, Computing in Civil Engineering, Proceedings of the first congress held in conjunction
with A/E/C Systems 94, ASCE, 55-62.

Li, H. (1995) Selecting KBES development techniques for applications in the construction industry. Construction
Management and Economics, in press.

Mahoney, J, and Tatum, C. (1994) Construction site applications of CAD, Journal of Construction Engineering and
Management, 120(3), 617-632.

Mikami, I. et al (1994) Expert systems with learning ability for Retro-fitting Steel Bridges. Journal of Computing in
Civil Engineering, 8(1), 88-102.

Minkarah, I. and Ahmad, I. (1989) Expert systems as construction management tools. Journal of Management in
Engineering, 5(2), 179-199.

Miyatake, Y. and Kangari, R. (1993) Experiencing computer integrated construction, Journal of Construction
Engineering and Management, 119(2), 307-322.
Mohan, S. (1990) Expert systems applications in construction management and engineering, Journal of Construction
Engineering and Management, 116(1), 87-99.

Moselhi, O. (1990) Hybrid expert system for construction planning and scheduling. Journal of Construction
Engineering and Management, 116(2), 221-239.

Mohsini, R. (1993) knowledge based design of project procurement process, Journal of Computing in Civil
Engineering, 7(1), 107-122.

Murtaza, et al (1993) Knowledge based approach to modular construction decision support. Jounral of Construction
Engineering and Management, 119(1), 115-130.

Oloufa, A. (1994) A user oriented approach to construction simulation of building, Microcomputers in Civil
Engineering, 9(6), 425-433.

Paul, J. et al (1992) Graffitti: A computer aided construction progress knowledge based system, Microcomputers in
Civil Engineering, 7(1), 181-188.

Paulsen, B. and Khoo, S. (1987) Managing construction worldwide, Vol.1 Systems for Managing Construction,
Lansley, P and Harlow, P. Eds, E & FN Spon, New York.

Russel, J. et al (1990) Qualifier-2: Knowledge based system for contractor prequalification. Journal of Construction
Engineering and Management, 116(1), 157-171.

Sauce, G. and Mommessin, D. (1995) A Knowledge based system for construction site organisation,
Microcomputers in Civil Engineering, 10(3), 187-198.

Scott, D. and Yang, J. (1991) Setting the pace with PACES: a new concept of using an expert system for project
analysisand control, Construction Management and Economics, 9(3), 529-541.

Songer, A, et al (1992) Knowledge based advisory system for public sector design build, Journal of Computing in Civil
Engineering, 6(4), 456-471.

Teicholz, P. (1994) Strategy for computer integrated construction technology. Journal of Construction Engineering and
Management, 120(1), 117-131.

Williams, T. (1991) Hypertext database applications in construction. Journal of Construction Engineering and
Management, 117(3), 460-467.

Williams, T. (1994) Predicting changes in construction cost indexes using neural networks, Journal of Construction
Engineering and Management, 120(2), 306-320.

Yang, J. (1991) An expert systems for project analysis and control, PhD Thesis, Queensland University of
Technology, Australia.

Yang, J. (1994) Integrated project information processing in Australian construction industry, Australian Institute of
Building Papers, 5(1), 121-134.

Yates, J. (1993) Construction decision support system for delay analysis. Journal of Construction Engineering and
Management, 119(2), 226-244.