Farhi Marir, Dr.
OD & KE Group, School of Computer Sc. & Electronic Sys.,
University of Kingston, UK.
email: F.Marir@Kingston.ac.uk
Ghassan Aouad, Dr.
Department of Surveying, University of Salford, UK.
email: G.Aouad@surveying.salford.ac.uk
Grahame Cooper, Dr.
Information Technology Institute, University of Salford, UK.
email: G.S.Cooper@iti.salford.ac.uk
SUMMARY: This paper presents OSCONCAD, an interactive system for integrating CAD and construction related applications to address the problems of design fragmentation and the gap that exists between construction and design processes. It provides a vehicle for storing architectural design information in an integrated construction object-oriented database that can be shared by a range of computer applications. The OSCONCAD model is characterised by several new features. It uses the object-oriented modelling approach to establish standard models for architectural design that comply with Industry Foundation Classes (IFC) for common interpretation of construction design objects and with CORBA (Common Object Request Broker Architecture) for distribution of the objects amongst the construction applications. It aims to achieve independence from the display environment by providing a set of Abstract Factory and Abstract Design Classes, which provide abstractions that the design model classes can use to draw and render themselves in any display environments. More importantly, graphical and textual information about the building design components is directly saved as instances in an object-oriented database without passing through the existing CAD databases. To demonstrate the independence from the display environment, two applications using OSCONCAD models are implemented. The first is an interactive AutoCAD application, which creates instances of the OSCONCAD design model and stores them directly in the distributed object database. The second A web-based application using VRML (Virtual Reality Modelling Language) for remotely interrogating information stored within the integrated database, visualising and manipulating the design components in 3D environment. Also, to demonstrate the feasibility and practicability of the OSCON (Open Systems for Construction) object-oriented product model, three OSCON construction applications that access and share the OSCONCAD building design instances are presented.
KEYWORDS: Distributed Construction Object Database, Model-Based CAD System, CORBA, IFC, OSCONCAD, UML, and VRML.
However, the main difficulties of integration are exacerbated by the fragmentation of design information. The drawings are expected to provide the experts from different areas of interest the information they require and serve as the main medium for integration. Despite the fact that computer-aided design (CAD) systems are extremely powerful, they are not being utilised thoroughly in the AEC industry. Specialists in the AEC industry have used today's CAD systems as simply automated drafting tools, thereby automating their own narrow areas of specialisation. Each participant uses unique drafting conventions and their own CAD systems. Information is scattered about the project in an uncontrolled and uncoordinated way, on a variety of information systems and media, so that the design cannot be viewed as a complete entity. Such obstacles to the free flow of information between parties to the construction process lead to data re-entry and the consequent inaccuracies, which prejudice future design 'migration paths' and operational flexibility. These are responsible for many of the quality problems and for adding to the costs of construction projects.
This paper presents the development of OSCONCAD system, which addresses the design fragmentation problems and provide a vehicle for storing architectural design information in an integrated object-oriented database shared across a range of computer applications developed within the OSCON (Open Systems for CONstruction) project. One of the main aims of the OSCON project is the assertion of construction industry requirements for an 'integrated solution' comprising common definitions of objects and functions, supported by software applications written for dedicated use in the sector. OSCON achieves this integration by establishing standard object-oriented models that all applications can adopt and share.
From these models, the integrated object-oriented model used by OSCONCAD has been developed. The OSCONCAD models have been designed to comply with the Industry Foundation Classes (IFC), a major industry standard data model, which is developed to be adopted as the standard of real world construction objects (IFC, 1997). The use of IFCs allows a common interpretation of these objects and the construction-related applications using such objects do so in a consistent and interoperable manner. Moreover, to share objects (function and data sharing in an integrated way) distributed amongst applications, the OSCONCAD design model will be complying with the CORBA (Common Object Request Broker Architecture) standards.
The content of the paper is organised as follows. The second section reviews and analyses related research in the area of design-construction integration. The third section is devoted to the description of the features of the OSCONCAD system. The fourth and fifth sections present two front-end applications used to display instances of the OSCONCAD distributed object database. The first is interactive AutoCAD application that allows users to create and interact with instances of architectural components of a building. The second is a web-based application that uses VRML (Virtual Reality Modelling Language) for remotely interrogating information stored in the object database, and visualising and manipulating the design components in a 3D environment. The sixth section gives a brief description of three OSCON construction applications that have been developed to share and manipulate the building design instances of the OSCONCAD object database. This section includes a prototype for estimating and planning, a piece of wrapper software for CA-Superproject®, and an application to support project managers in managing communication and monitoring the progress of different phases of the construction project. The last section presents the conclusion of this project.
In the last few years, researchers and practitioners have started to develop interfaces between CAD systems and relational and object-oriented databases that support the design process. Atkin and Gill at Reading University have configured an experimental integrated system to provide information for management purposes, besides producing drawings (Atkin and Gill, 1986). This system is based on the idea of integrating computer-aided design and relational database management systems. Details of specifications, costs, duration and resources are stored in the database system, which accepts graphical data from the CAD model and relates them to their associated non-graphical attributes. It is suggested that design practices should be modified to adopt a component-based approach and that the computer technology is available now to develop new approaches to the management of construction using CAD systems (Atkin and Gill, 1986).
An interface for an integrated plant model was developed (Zabilski and Hall, 1989). In this system, the capabilities of 3-D solid computer models are combined with the efficiency of a cost and schedule database to plan and schedule the construction process. BUILDER is a prototype system, which generates construction schedules from architectural drawings using KEE running on a LISP machine (Cherneff et al., 1991). BUILDER involves the development of an object-oriented CAD program that captures building component data as the graphics are created by making use of the KEE-Pictures utility embedded in the KEE system. Data generated by KEE-Pictures become an object-oriented model within the KEE environment. The object model serves as input to knowledge based planning and scheduling system implemented in KEE. While exploring the capabilities of KEE to create a CAD system within an artificial intelligence (AI) environment is innovative and interesting from the research point of view, utilising a non-standard CAD system running on a LISP machine is both uncommon and expensive in the AEC industry (Kartam, 1994). Timberline Corporation in the US, has developed a link, known as CAD integrator, between AutoCAD and their Precision Plus estimating system (Timberline, 1990). The CAD integrator performs take-off directly from AutoCAD drawings. It then moves the information into Precision Plus, an estimating spreadsheet-based program. Once estimating has been done, information can also be tied directly to Primavera and Microsoft Scheduling products to give more precise job planning. Another system, called CADlink, has been developed as an interface for translating a CIFECAD CAD model into an object-oriented database. CIFECAD is a component-based customised AutoCAD application from which the CAD link interface transfers blocks representing components (Ito et al., 1989). Kartam has developed an intelligent and customised CAD interface known as ISICAD (Intelligent Computer-Aided Design System) that is capable of generating an object oriented model during design (Kartam, 1994). This model uses a component-based approach. The system has been implemented within the AutoCAD 12 environment. It integrates a CAD system with databases, to incorporate project data, and expert systems to evaluate the evolving design solution and resolve conflicts that may arise as a direct result of decisions made by the designer. The emphasis in this system is more on creating an intelligent design environment for the architect than on integrating structural design and construction. In the COMBINE project (Dubois et al, 1995) which relied on STEP-based data exchange integrate both Microstation and AutoCad systems. It developed its own data model of a building and integrated applications to it, without relying on CAD formats. Alshawi at Salford University developed SPACE (Simultaneous Prototyping in An integrated Construction Environment), an environment that allows the integration of industry standard software across a construction project's lifecycle (Alshawi, 1996) using similar approach to that taken in the ISICAD system. A similar approach has been taken in the DICE (Distributed and Integrated environment for Computer-aided Engineering) project (Sriram and Logcher,1993, Sriram et al., 1995, Gorti et al.,1996). DICE has addressed the co-ordination and communication problem in engineering. It is designed as a network of agents or Knowledge Modules (KMs) which communicate through a shared workspace.
The first step in using common models for computer-based applications integration is to establish a common data model or modelling language-that is, a standard paradigm for representing and communicating information in general. The OSCONCAD project is committed to a fully object-oriented approach, in which an emphasis is placed on inter-working between software objects rather than on sharing of data using common formats. This approach provides a number of advantages in terms of the ability to use abstraction to handle complexity (Ahmed et al., 1991, Cooper, 1995, Booch, 1994):
This allows the OSCONCAD implementation immediately to take advantage of the inter-working facilities provided by the CORBA.
Figure 2: A subset of IFC1.2 Objects.
The root cause of the different problems in construction projects is the information or data upon which all participants depend. Because of fragmentation within the industry, there are many different interpretations of the semantics of the data in use. The formal ways in which to describe information semantics are referred to as models, which represent a formal description of a view of a domain of interest. A model is always a representation of the real world to some acceptable degree of approximation for the viewpoint we hold (Wix and Storer, 1996). In practice, the word `modelling' gives rise to many interpretations and the solution lies in the fact that there is only one actual product to be conceived, designed, manufactured and maintained. Therefore, there is a need for only a single model of the real world product from which the other consistent representations can be derived and through which different presentations are generated as user views. In the ICON methodology, this concept is referred to as a "canonical model" (Cooper et al., 1994) and also termed a "product model" in (Wix and Storer, 1996). Modelling design information is the most debatable aspect of designing the product model. It involves dealing with graphical and non-graphical data and many complex issues such as representing form, function, topology, appearance and many other aspects. Major efforts have been undertaken by researchers in the STEP community and outside it in order to define a generic model that can describe a building. The Technical Research Centre of Finland has made multiple studies about a building product-model (RATAS) using relational databases and hypermedia (Pentilla, 1989). Several researchers have suggested models for space and space enclosure and have also developed object-oriented product models that could be used to store richer kinds of data and knowledge about a product, including knowledge about its design, manufacturing and operational parameters (Svensson and Aouad, 1997).
Figure 3: A subset of the OSCONCAD architectural design model.
None of these models has a universal acceptance. This is due to the fact that a data model must be readily accessible to all construction-related applications and users. It must also be compatible with systems across AEC and other industries (Froese and Paulson, 1994). Recently a group of software companies and users have come together to form the Industry Alliance for Interoperability and define specifications for a set of Industry Foundation Classes (IFC, 1997). "Fig. 2" shows a subset of the classes of the IFC 1.2. The IFC is an industry object model, proposed as the standard for construction integrated systems. By its compliance to the IFC, the OSCONCAD models definition allows for a common interpretation of real world construction objects, whereby software applications using such objects do so in a consistent and interoperable manner. In this model, an object will incorporate physical and non-physical information such as weight, colour, electrical or thermal properties, cost, as well as a definition of how it should appear in space (from which applications software can derive an appropriate display representation). An object will also encapsulate procedures that can be used to describe how it should behave and relate to other objects. Thus, for instance, an object describing a light switch might give its height above a floor and the distance from a door and an object describing a window might 'cut itself' into a wall and create cavity closures and lintels automatically (IFC, 1997). The aim of the OSCONCAD system is not to recommend another design model but to derive classes from the IFC Generic classes and enhance them with knowledge about the detailed design stage that can be shared with other applications, such as cost and time planning. The detailed design model is a description of the design components and their specifications with very limited topological properties incorporated as shown in Fig. 3. The model offers the representation of several structural members:
1. Objects that reflect the database schema of the OSCONCAD architectural design model where instances of the design component are stored and
2. A set Abstract Design Classes, which provide abstractions that the design model classes can use to render themselves in a given display environment.
The AutoCAD and VRML front-end applications are discussed in individually in the fourth and fifth sections. They use the Abstract Design Classes to derive concrete classes to display instances of the design in the AutoCAD and VRML environments. However, the cost estimator (Esteem), the project manager (Mentor) and the project planner (Planner) applications, shown in Fig. 4., are discussed in the seventh section. These client applications are designed to access and manipulate instances of the specific domain and the OSCONCAD design distributed objects through CORBA Object Request Brokers.
At present, the data model of object-oriented systems are dictated by their underlying languages, databases, or programming environments. However, several major object-oriented data model standards currently are being developed. These standards provide all the characteristics needed to support integrated systems, and most commercial object-oriented databases are likely to adopt these standards as their fundamental data models. The OSCONCAD object model has adopted the CORBA (Common Object Request Broker Architecture) standard (CORBA, 1995) developed by the Object Management Group (OMG), an international organisation of information systems vendors, users, and researchers dedicated to promoting industry standards. By adopting CORBA standards, the OSCONCAD system allows inter-working between object-oriented software components within networked computing environments (Atwood, 1991). The compliance of the OSCONCAD objects with CORBA standards will allow not only the sharing and interchange of design data with linked applications of the integrated framework, but also the sharing of objects distributed amongst all the applications. To achieve this aim, CORBA IDL (Orbix, 1996) interfaces have been written for each of the OSCONCAD objects. The use of the IDL neutral interface for the OSCONCAD objects promotes language and platform independence, location transparency, modularity, and robustness of the system, which facilitates the integration of other construction applications that access and manipulate the OSCONCAD distributed objects.
Figure 4: OSCONCAD Architecture
Figure 5: The OSCONCAD Abstract ShapeFactory and Abstract Design
Model.
In practice, all classes in the OSCONCAD design model inherit from the base class IFCBuildingElement that defines a common interface. This interface includes a draw() operation which is redefined in subclasses to allow instances of the design classes to draw themselves in a given display environment. The draw() operation passes a pointer to the ShapeFactory (or Drawer) object which encapsulates a set of simple 3D drawing operations as virtual member functions. These functions are implemented in the ShapeFactory concrete subclasses i.e. AutoCAD_Drawer and VRML_Drawer, to display the drawing in specific environments. As a result of this modelling approach the design of building components stay independent of the prevailing CAD and other display environment and the AutoCAD 13 and the VRML applications discussed below become graphical front ends for the instances of the OSCONCAD architectural design model.
The application provides a simple environment for designing building components. The design primitives are walls, windows, doors, columns, beams, foundations etc. As design objects are created through the AutoCAD interface, corresponding objects are created in the object-database. The information that is used to draw these objects on the AutoCAD screen via an instance of the Drawer class comes directly from the instances in the database. This ensures that the screen reflects the design database. This also means that changes in a design can be immediately propagated to other areas of the project database. For instance, if an on-line project-planning package were accessing the database then it would be possible to change a part of the design and immediately see the impact these changes would have on the project schedule. The prototype provides the following functionality:
When the application is started it opens the ObjectStore database for storing the design data. As various commands are called to add design elements, each command invokes a function in the application which starts a transaction on the database, prompts the user for parameters and positioning of the design element and creates an instance of the element in the database. The display is then updated to reflect this addition, and the transaction is committed to the database. The graphical representations of the building elements are displayed in the AutoCAD drawing, with the user still having access to all the usual AutoCAD commands for changing the viewpoint, zooming, performing hidden line removal etc. Thus the user creates a design using design elements instead of creating a drawing using low level drawing primitives such as lines and 3D-faces. The advantage of this approach is that the application, and indeed the database objects themselves, can contain knowledge about how buildings are designed and can thus aid the designer during the design process. An example of this kind of knowledge built into the OSCONCAD database is that a column, which is supporting a beam, will be moved automatically if we stretch the beam. This knowledge is actually built into the database itself (in the implementations of the operations on beams) and so does not have to be programmed directly into any CAD applications that manipulate beams.
When the user leaves the application, the database is closed and the current drawing is discarded. There is no need to save the drawing as a standard AutoCAD file, because the next time the application is started it will recreate the drawing from the OSCONCAD object database.
Figure 6: OSCONCAD AutoCAD 13 Design prototype.
Most of the previous research on design-construction integration uses CAD database facilities, such as AutoCAD Block structures to store design information and also provide expensive and specialised interfaces to decode and transfer this information into an object-oriented database. The same process is repeated when the user performs any manipulation on the drawing. The difficulty with block structure is more apparent when performing an operation that involves more than one building element. For instance, to model the intersections between building elements, it is unavoidably necessary to explode the blocks, perform the editing task and then use the block facility once again. For complex projects, such as buildings, the use of blocks is not appropriate as the layout, orientation, shape, geometry, etc. of a building is not well defined. The OSCONCAD system avoids these complex and expensive interfaces. It uses exclusively the object-oriented database directly to store design information and to display the associated drawing in a CAD environment. To achieve this the OSCONCAD provides an alternative to block structure is used. An ID protocol is used to refer to objects previously created in the CAD environment. Each operation, which creates or updates a drawing entity, takes an ID string as a parameter. If this string is empty the operation creates a new drawing entity and returns a unique ID string for that entity. The ID string may then be used as a parameter to the same operation in which case the previously created entity is updated.
In OSCONCAD, VRML is not only used as a visualisation tool, but as a user interface as well. For instance, the user could interact with a 3D column in VRML rather than a column in a traditional database environment. This will allow the construction practitioners better access to information, which will motivate them to use integrated databases. Fig. 7. shows a screen shot taken from the OSCONCAD VRML application of a house and its related schedule from a VRML representation generated by the OSCON database. The user can navigate inside the building clicking on design objects and retrieving information about their properties, which include geometric, cost and time data.
This direct interaction with a VR environment has many advantages
over the use of a CAD package. In traditional CAD systems, the design is
relatively static which makes changes costly and time consuming as new
images are re-generated from sequences of fixed frames (Larijani,
1994). In a VR environment, changes are handled efficiently by the
technology. The ultimate benefit is the ability to create a walk-through,
which can facilitate collaboration between clients, designers, contractors
and suppliers. Such collaboration is the main objective of the OSCON database
and VR is a powerful medium for communication and convergence.
Figure 7: A VRML house model.
Figure 8: Cost displayed in Esteem.
Figure 9: A legacy application accessing the project database.
Figure 10: The Mentor process manager.
The OSCONCAD research described in this paper attempts to overcome the limitations of the previous research development in the area of design-construction integration. This is by using a model-based approach whose emphasis is on capturing graphical and textual information about building components and directly storing them into an object oriented database rather than the complex and ill-structured CAD database. This approach will avoid the expensive process of translating drawing data from the CAD database into the object-oriented database. Moreover, this approach allows the separation of the design data from the specific CAD package that is used to manipulate it. A consequence of this is that new CAD products will be much easier to incorporate and also other none CAD applications can have direct access to the design data from such a database. In addition, the OSCONCAD model classes compliance with the IFC will facilitate data exchanges between construction-related applications and allow common interpretation of the design objects amongst the applications in a consistent and interoperable manner. Also, in the context of project data interchange the OSCONCAD not only concentrates on the problem of integrating applications via data sharing but goes further into sharing objects distributed amongst applications by complying to CORBA standards.
This work does not address issues relating to management of the updating of objects. However, the authors believe that mechanisms to handle issues such as versioning, authorisation, and aggregation of changes will need to be addressed in relation to object-based approaches. Some of these issues are being addressed in a related COMMIT project (Cooper et al., 1997).
A system like OSCONCAD evolving in an integrated framework will have an impact on the ability of a construction project to meet client needs. It will also contribute to the quality of a construction project by improved design and better communication between the design and construction stages of the construction project. This will also reduce the AEC industry fragmentation and increase competitiveness. It also provides a vehicle for sharing different types of graphical and non-graphical data among the construction participants and across stages in the building life cycle. By using standards for different types of data, that data can be transferred among the project participants. By using the object-oriented technology, the OSCONCAD favour the reuse of object and applications and avoids expensive islands of automation and interfaces between construction applications. This will reinforce the important industry restructuring moves taking place and the trend towards greater use of design and build contracts as an organisational attempt to overcome the construction problems.
Ahmed S., Wong A., Sriram D. and Logcher R. (1991) . A comparison of object-oriented management systems for engineering applications. Resaerch Report No R19-12, Intelligent Engineering Systems Laboratory, Dept. of Civil Engineering, Massachusetts Institute of Technology, May 1991.
Alshawi M. (1996) . SPACE: integrated environment. Internal Paper, University of Salford.
Aouad G. (1991) . Integrated Planning Systems for the Construction Industry. PhD thesis, Loughborough University.
Aouad G., Marir F., Child T., Brandon P. and Kawooya A. (1995) . A Construction integrated databases- Linking Design, Planning and Estimating. International Conference on Rehabilitation and Development of civil engineering infrastructure systems, June 9-11, 1997, American University of Beirut, Lebanon.
Aouad, G,.Child, T, Marir, F & Brandon, P (1997) . Developing a conceptual model for an integrated database, Proceedings of intelligent information systems, Bahamas, December, pp 316-320.
Aouad G., Betts.M, Brandon.P, Brown.F, Child.T, Cooper.G, Ford.S, Kirkham.J, Oxman.R, Sarshar.M, and Young.B. (1994) . Integrated databases for design and construction: Final Report. University of Salford
Atkin B. and Gill M. (1986) . CAD and Management of Construction Projects. Journal of Construction Engineering and Management, Vol 112, No 4, December, ASCE, pp 557-565.
Atwood T. (1991) . Why the OMG object request broker should mean good news for object databases. Journal of Object-Oriented Programming, Vol. 4 No4, pp8-11.
Bjork B-C. and Wix J. (1991) . An introduction to STEP. VTT and Wix Mclelland Ltd. 1991.
Björk B-C. (1994) . RATAS Project - Developing an Infrastructure for Computer-Integrated Construction, Journal of Computing in Civil Engineering, Vol. 8, No. 4, 400-419. http://www.vtt.fi/cic/ratas/index.html
Bohms M., Tolman, F. and Storer G. (1994) . ATLAS, a STEP Towards Computer Integrated Large Scale Engineering, Revue internationale de CFAO, Vol. 9, No. 3, 325-337. http://www-uk.research.ec.org/esp-syn/text/7280.html
Booch G. (1994) . Object-Oriented Analysis and Design with Applications, 2/e. ISBN 0-8053-5340-2. Addison-Wesley.
Brandon P.S. (1993) . Integration of Construction Information. Proceedings of N+N Workshop. Lake District. UK.
Cherneff J., Logcher R. and Sriram, D. (1991) . Integrating CAD with construction schedule generation. J. Comput. Civil Engineering. 5(1), 64-84.
Cooper G.S., Aouad. G, and Brandon. P. (1994) . ICON Methodology Report, IT Institute and Department of Surveying, University of Salford.
Cooper G.S. (1995) Object-Oriented Databases from an Information Technology Viewpoint: Knocking Down Some Walls, in Object Technology and its Application in Engineering. ISBN 0902 376 209, Ed. Professor James Powell, proc. Conf. on Object Technology and its Application in Engineering, Glasgow, March 1995.
Cooper G.S., Rezgui Y., Yip J., and Brandon P. (1997) Notification and Change Propagation Support in a Multi-actor Environment, In proceedings of the CIB W78 Workshop on Information Technology Support for Construction Process Reengineering, Cairns, July.
CORBA (1995) . The: Common Object Request Broker: Architecture and specification. Revision 2.0.
Dubois, A.M., Flynn, J., Verhoef, M.H.G. and Augenbroe, F. (1995) Conceptual Modelling Approaches in the COMBINE Project, presented in the COMBINE final meeting, Dublin. http://erg.ucd.ie/combine/papers.html
Gamma, E., Richard Helm, Ralph Johnson, and John Vlissides. (1994) . Design Patterns: Elements of Reusable Object-Oriented Software. Addison Wesley.
Gorti, S., Sriram, D., and Ram, D. (1996) . From Symbol to Form: A Framework for Conceptual Design, Journal of CAD, Vol. 28, No. 11, 1996.
Ferguson I. (1989) . Buildability in Practice. Mitchell's Professional Library, London.
Froese, T. and Paulson B. (1994) . OPIS: An object model-based project information system. Microcomputers in Civil Engineering, 9, pp 13-28.
Howard, H.C. (1991) . Linking design data with knowledge-based construction systems. CIFE Spring Symposium.
IFC (1997) . Industry Foundation Classes, Release 1.0,Vol 1-4.
Ito, K. (1991) . Design and construction integration using object oriented project model with multiple views. In Proceedings of the Computer-Integrated Design and Construction Symposium, Construction Congress II, Cambridge, MA, April 1991.
Ito, K., Ueno, Y., Levitt, R. and Darwiche, A. (1989) . Linking knowledge-based systems to CAD design data with an object oriented building product model, CIFE Technical Report No. 17, Stanford University.
Kartam, N. (1994) . ISICAD: interactive system for integrating CAD and computer-based construction systems. Microcomputers in Civil Engineering 9, pp 41-51.
Larijani, L.C. (1994) . The virtual reality primer. McGraw-Hill, USA.
Latham, M. (1994) . Constructing the team, H.M.S.O.
Orbix (1996) . White Paper. IONA Technologies Ltd.
Pentilla, H.(1989) .A Scenario for The Development and Implementation of A Building Produce Model Standard. CIFE(Center for Integrated Facility Engineering-Stanford University) Symposium Proceedings.
Port, S. (1989) . The management of CAD for construction. BSP Professionals.
Rumbaugh, J., Blaha M., Premerlani W., Eddy F. and Lorensen, W. (1991) , Object-Oriented Modelling and Design. Englewood Cliffs, New Jersey: Prentice-Hall.
Sriram, D. and Logcher, R. (1993) . The MIT DICE Project, IEEE Computer, Special Issue on Concurrent Engineering, pp. 64-65, January 1993.
Sriram, D., Ram D., Wong, A., and Li, H. (1995) , GNOMES: An Object-Oriented Non-Manifold Geometric Engine, Journal of CAD, November.
Svensson K. and Aouad G. (1997) , Developing standardised architectural building product models for spaces and space enclosures. Submitted to Automation in Construction.
Tah J., Thorpe A.and McCaffer A.(1991) Decision making within large computer based construction management systems. SERC Report GR/E76957/E.
Timberline (1990) . Precision Estimating, Timberline Software Corporation, USA.
UML (1997) UML Document Set, Version 1.0, 13 January, 1997 (Rational Software Corporation). http://www.rational.com/uml/index.html
Wix, J. and Storer, G. (1996) . Building core models. Occasional paper.
Zabilski R. and Hall H.E. (1989) . Presented
in 3D. Civil Engineering, June, pp 48-50.
