2005:18

RESEARCH REPORT

Modelling and Simulation of Cast

in Place Concrete Constructions
Using N-Dimensions
SBUF Project 11333 ITstomme - Final Report

ROGER JONGELING
MATS EMBORG
MARTIN ASP
THOMAS OLOFSSON

Luleå University of Technology

Department of Civil and Environmental Engineering
Division of Structural Engineering

2005:18 • ISSN:1402-1528 • ISRN: LTU - FR -- 05⁄18 -- SE

 TECHNICAL REPORT

MODELLING AND SIMULATION OF CAST

IN PLACE CONCRETE CONSTRUCTION
USING N-DIMENSIONS
SBUF PROJECT 11333 ITSTOMME - FINAL REPORT

Rogier Jongeling, Mats Emborg, Martin Asp, Thomas Olofsson

Luleå 2005

Division of Structural Engineering

Department of Civil and Environmental Engineering
Luleå University of Technology
SE - 971 87 LULEÅ
www.cee.ltu.se
construction.project.ltu.se
 Preface

Preface

This report is the result of a research and development project on modelling

and simulation of construction with cast in place concrete by using product
modelling technology. The project is called ITstomme and was conducted in a
two-year period from the end of June 2003 to July 2005. The ITstomme project
was driven as a project within the eBygg centre for information technology in
construction at Luleå University of Technology.

Product modelling poses challenges to cast in place concrete construction, but

offers at the same time a number of opportunities. Within the ITstomme project
a number of dimensions of product models are presented. These dimensions
illustrate on the one hand the support for new and existing products and
processes for cast in place concrete construction and on the other hand show
the potential of product models beyond their common, but limited, use for 3D
graphics.

Product models are now starting to be adapted in projects by JM and

Betongindustri, which are two of the main participants in the ITstomme
project. We hope that the developments and pilot implementation efforts from
the ITstomme project can add to the value of product models in the business
processes of both companies. In addition, we hope that the work in the
ITstomme project forms a source of inspiration and a call for action for other
actors in the construction industry.

During our work we had the opportunity to cooperate with many different,
experienced and enthusiastic persons from Sweden and other countries. We
thank these persons for their support and especially value the implementation
efforts by the software developers at Enterprixe, Ceco and Concode-Intercopy.

I
SBUF project 11333 ITstomme - Final Report

Part of the research work has been performed by Rogier Jongeling while
visiting the Center for Integrated Facility Engineering (CIFE) at Stanford
University between March and July 2004. Professor Martin Fischer, Claudio
Mourgues and Jonghoon Kim of CIFE are acknowledged for their time and
input in this project.

We would like to express our gratitude to Svenska Byggbranschens

Utvecklingsfond (SBUF) for funding this research and development project.
We also thank the participating project members, Betongindustri, JM, Ceco,
Enterprixe and the eBygg centre at Luleå University of Technology, for their
contributions to the project.

Luleå, October 2005

The project team:

Rogier Jongeling, Luleå University of Technology*

Mats Emborg, Betongindustri & Luleå University of Technology

Martin Asp, JM

Thomas Olofsson, Luleå University of Technology

*
Contact person. Tel.: +46702702543; Email address: rogier.jongeling@ltu.se

II
 Abstract

Abstract

Product models have long been theoretical models within international research
communities, but are now starting to be adapted in the construction industry.
Product models are data models that can contain both product and process data,
such as geometry and planning data. Their current use in construction is mainly
limited to 3D graphics, but product models have many more than three
dimensions. In this research project we developed and applied different
dimensions of product models beyond their common, but limited, use in 3D.
Development and implementation efforts in the project are mainly driven by
the interests from a ready-mixed concrete supplier.

Product modelling systems imply an object oriented approach to cast in place

concrete construction, which is not object oriented as a result of the nature of
the product. Product models that are used in the design process differ from
models that are used for production simulations. These, and other conditions,
pose a number of challenges. However, product models offer also a number of
opportunities to cast in place concrete construction. Products and processes can
be simulated virtually for example, thereby facilitating the decision-making
process regarding the application of new or existing technologies that can
improve the construction process. Product models can also be used to link or
integrate product data, such as material specifications, casting sequences, etc.

Based on these, mainly theoretically identified, possibilities and constraints, a

number of dimensions of product models are developed and applied in practice
to different case study projects.

III
SBUF project 11333 ITstomme - Final Report

The following model dimensions are explored in the project:

- Traditional 2D drawings and documents are partly generated from 3D

models and hyperlinked to these models.
- Various types of 3D models from different disciplines are created by
using file-based data transfers and a collaborative client-server
environment.
- Multiple production plannings are linked to 3D models resulting in 4D
models.
- 5D cost estimation is developed by linking different 3D models to cost
estimation hierarchies.
- Model use and configuration of material parameters by a ready-mixed
concrete supplier provides an additional dimension to the use of product
models.

The different dimensions illustrate on the one hand the support for new and
existing products and processes for cast in place concrete construction and
show on the other hand the potential of product model use beyond 3D graphics.
Advantages and possibilities can be identified for the ready-mixed concrete
supplier in the different model dimensions and we believe that other actors in
construction can do the same. When actors have identified benefits of product
modelling for their own business process they might be more willing to
participate in shared product model use in search of benefits for the project
team as a whole.

Keywords: product model, cast in place concrete, design, construction

simulation

IV
 Abbreviations

Abbreviations

ABC Activity-Based Costing

ADT Architectural DeskTop

AEC Architecture Engineering Construction

BIM Building Information Model

BoM Bill of Materials

Byggandets Samordning AB. Classification System used in

BSAB
the Swedish Construction

CAD Computer Aided Design

CIFE Center for Integrated Facility Engineering

CNC Computer Numerical Control

CPM Critical Path Method

EPX Enterprixe

HVAC Heating Ventilation Air-Conditioning

IAI International Alliance for Interoperability

IFC Industry Foundation Classes

V
SBUF project 11333 ITstomme - Final Report

IT Information Technology

KPI Key Performance Indicators

LCI Lean Construction Institute

LoB Line-of-Balance

PE Project Explorer

PIO Project Information Officer

SBUF Svenska Byggbranschens UtvecklingsFond

SCC Self-Compacting Concrete

SCM Supply Chain Management

UDA User-Defined Attribute

VBE Virtual Building Environment

Visualisering i Projektering och Produktion (SBUF project

VIPP
no 11693)

VRML Virtual Reality Mark up Language

XML eXtensible Mark up Language

VI
 Table of Contents

Table of Contents

PREFACE ............................................................................................................I
ABSTRACT......................................................................................................III
ABBREVIATIONS ........................................................................................... V
TABLE OF CONTENTS................................................................................ VII
1 INTRODUCTION .....................................................................................1
1.1 Background ......................................................................................1
1.2 Objective ..........................................................................................4
1.3 Method .............................................................................................4
1.4 Organisation .....................................................................................6
1.5 Partners.............................................................................................7
1.6 Finances............................................................................................8
1.7 This report ........................................................................................8
1.8 Project publications ..........................................................................9
2 PRODUCT MODELLING......................................................................11
3 PRODUCT MODELLING ENVIRONMENT .......................................13
4 CAST IN PLACE CONCRETE ..............................................................17
4.1 Possibilities and challenges............................................................18
5 CASE STUDIES......................................................................................21
5.1 Hotellviken Marinan ......................................................................21
5.2 4D case study .................................................................................22

VII
SBUF project 11333 ITstomme - Final Report

6 THE SECOND AND THIRD DIMENSION.......................................... 25

6.1 Abstract .......................................................................................... 25
6.2 Introduction.................................................................................... 25
6.3 Results............................................................................................ 26
6.4 Discussion ...................................................................................... 28
7 THE FOURTH DIMENSION................................................................. 31
7.1 Abstract .......................................................................................... 31
7.2 Introduction.................................................................................... 31
7.3 Results............................................................................................ 32
7.4 Discussion ...................................................................................... 34
8 THE FIFTH DIMENSION...................................................................... 37
8.1 Abstract .......................................................................................... 37
8.2 Introduction.................................................................................... 37
8.3 Results............................................................................................ 38
8.4 Discussion ...................................................................................... 39
9 THE NTH DIMENSION .......................................................................... 41
9.1 Abstract .......................................................................................... 41
9.2 Introduction.................................................................................... 41
9.3 Results............................................................................................ 42
9.4 Discussion ...................................................................................... 44
10 CHALLENGES FOR PRODUCT MODELLING IN PRACTICE ........ 47
10.1 Education and culture .................................................................... 48
10.2 Implementation and integration ..................................................... 48
10.3 Demonstrating benefits .................................................................. 49
10.4 Data issues ..................................................................................... 50
10.5 Interoperability............................................................................... 50
11 DISCUSSION ......................................................................................... 53
11.1 Recommendations.......................................................................... 57
11.2 Model creation and access ............................................................. 59
11.3 Organizational................................................................................ 60
REFERENCES ................................................................................................. 63
Appendix A PROJECT MATRIX .......................................................... 67
Appendix B 4D CASE STUDY ............................................................. 69
B.1 Introduction.................................................................................... 69
B.2 Set up ............................................................................................. 70

VIII
 Table of Contents

B.3 Alternative I: 0 Reference ..............................................................71

B.4 Alternative II: Industrialized ..........................................................77
B.5 4D results .......................................................................................81
Appendix C 5D COST ESTIMATION ..................................................83
C.1 Introduction ....................................................................................83
C.2 Current praxis.................................................................................84
C.3 Proposed method............................................................................85
C.4 Implemented method......................................................................87

IX
 Introduction

1 INTRODUCTION

1.1 Background

The cast in place concrete industry is increasingly interested in applying new

processes and products to deliver value for its clients and to remain
competitive compared to other construction methods and other types of
materials. Thus, it is strived for the development of a more industrialized
construction process using cast in place concrete. Efforts are made to improve
the construction process on the one hand and the construction products on the
other hand. These efforts are necessary to create conditions by which projects
can be delivered on time, within budget and according to the requirements of
the client.

The technical research and development that is aimed at establishing these

conditions for the cast in place concrete industry can be grouped into five
areas:

- Improved concrete qualities that are customized for specific and

optimal use. For example, Self-Compacting Concrete (SCC) reduces
work on site and adds to the quality of the structure. Concrete with
reduced drying time can shorten the total construction time
significantly.
- Minimized reinforcement activities utilizing fibre reinforced concrete
and prefabricated reinforcement reduce on-site operations and increase
the production rate of the work.
- Permanent formwork minimizes site logistics. Form elements are
prefabricated in a factory under controlled conditions and erected on
site.

1
SBUF project 11333 ITstomme - Final Report

- Weather independent construction processes, e.g. by building a

structure inside a climate protective tent.
- Information Technology (IT) where different types of systems enable
the regeneration, flow and access of information in and between
different processes in the value chain of cast in place concrete
construction.

These topics are not totally isolated from each other, but are interrelated areas
and the benefits of the different technologies can be enhanced by applying
them in different combinations. A practical prerequisite for the application of
these technologies is that they can be applied in combination with other,
existing, building systems and building processes that are part of the total
construction process. As an example the application of permanent formwork
systems can be given. Prefabricated permanent formwork elements for walls
can contain openings for windows and doors or have these elements integrated
in a factory. This requires coordination of different design disciplines, such as
architects and structural engineers. It also requires detailed planning of
logistics and erection on site due to the fact that prefabricated elements can
demand a certain order of installation. From this example it becomes clear that
the application of new technology demands a multi-disciplinary approach to
design, planning and construction. However, the technical and organisational
infrastructure for such a multi-disciplinary approach is currently not available
or a common practice in construction projects.

A potential technology that can be part of this technical infrastructure is

product modelling technology, which is discussed in Chapters 2 and 3. Product
models are data models that can contain both product and process data, such as
geometry, planning data, material specifications, etc. Product models for
construction are nothing new and have been an area of research for many
years. Much of the inspiration and experience for the application of product
models comes from the manufacturing industry, which is far ahead of the
construction industry in applying model-based design and production
processes. Products and processes literally changed in the manufacturing
industry partly as a result of the application product modelling (Olofsson
2004).

The construction industry has not come very far with the application of product
models, with the exception of a few visionary and pioneering companies.
Current application of product modelling in construction is often synonym to
3D modelling and 3D graphics, used in early stages of projects. There are

2
 Introduction

many reasons and explanations that can be given why the construction industry
is still relying on a non-model-based information flow, e.g. 2D drawings.
Many researchers state that the industry is fragmented and conservative.
Professionals in the industry often agree with these statements, but note that
research regarding product modelling has been too technical. These
practitioners also experience that the application requires too many changes in
today’s processes, too much expertise to work with in practice, and without
any clear and obvious advantages beyond the use of 3D graphics. These
conditions result in prerequisites and challenges for the use of product models
in construction, which are discussed in Chapter 10 ‘Challenges for Product
Modelling in Practice’.

The challenges, but also the potential, for the use of product modelling for cast
in place concrete are discussed in Chapter 4. This section is followed by a
number of practical uses of product models for cast in place concrete structures
according to the different dimensions of product models. These applications
illustrate that product models can be applied in today’s processes with minimal
changes by using commercial and prototype software packages. The
applications also illustrate the use of product models for cast in place concrete
construction beyond 3D modelling and 3D graphics.

The application of product models to cast in place concrete construction is

explored in a prestudy at Luleå University of Technology in which the current
value chain from a contractor and concrete supplier are analyzed (Jongeling
2003). The main actors in the project are Betongindustri, a ready-mixed
concrete supplier, and JM which is a residential project developer. In the
prestudy it is shown that product modelling has the potential to improve the
business process as well as the information management and cooperation in
projects between Betongindustri and JM, by using product models beyond 3D
graphics. As a result, both companies joined this project to study the
application of product models with a broader scope and with application in real
and virtual construction projects. Starting point for the application of product
models is to support existing and new construction products and processes that
contribute to the use of cast in place concrete and that suit the operations by
contractors and material suppliers, such as JM and Betongindustri.

3
SBUF project 11333 ITstomme - Final Report

1.2 Objective

The objective of this study is to support existing and new products and
processes for cast in place concrete construction by developing and applying
product modelling technology beyond 3D graphics.

The aim is to establish and analyze different model dimensions by the

development and application of modelling tools to residential construction
projects. This study does not address the structural design of concrete
structures with product modelling systems, but focuses on the use of product
models throughout a building process from the perspective of a contractor and
a ready-mixed concrete supplier.

1.3 Method

1.3.1 Model dimensions

The implementation of the project is performed according to the nD modelling

framework (Lee 2005) in which different dimensions of a product model are
identified. This report is structured according to these model dimensions. The
dimensions include:

- 2D drawings and documents in combination with 3D modelling

- 3D models of different design disciplines
- 4D modelling and simulation of alternative construction methods
- Cost estimation hierarchies are mapped to 3D models, providing a 5th
dimension
- An example is given of a possible nth dimension of a product model by
integrating input from a concrete supplier into 3D models.

1.3.2 Project phases – as planned

When initiating the study in 2003 the plan was to approach the project in five
phases:

1. A prestudy and overview of product modelling tools for cast in place

concrete
2. Pilot study in which existing tools are applied to the construction
project Hotellviken Marinan by JM

4
 Introduction

3. Evaluation and development work that form the basis for the next phase
in which new tools are applied
4. Application of modelling tools in full scale in the next phase of the
Hotellviken project by JM, called Ringvägen
5. Evaluation and reporting

1.3.3 Project phases – as implemented

A number of factors resulted in a slightly different implementation of project

phase two and four:

- The pilot study project Hotellviken Marinan was delayed due to a long
process to obtain a building permit. Therefore a virtual project, with a
similar type of construction as Hotellviken Marinan, was used as a test
object in phase two of the project. This project was used as a basis for
the development and application of 4D modelling tools while the
building permit for Hotellviken Marinan was pending.
- When JM received the building permit in 2004 it was decided to treat
Hotellviken Marinan as the full scale project in phase four of this
project and to include the next phase of the project, Hotellviken
Ringvägen, in the SBUF project no. 11693, called VIPP (Visualisering
i Projektering och Produktion. SBUF project no. 11693).
- During the prestudy it became clear that the application of modelling
tools connected to a model server by all consultants in full scale was
not yet feasible, due to technical and organisational reasons. Parts of
the 3D modelling work for Hotellviken Marinan were performed
instead with stand alone tools by consultants and other parts of the
modelling work were performed in a client-server environment by the
project team of this project.
- The prestudy showed that there was a need to develop and use a model
viewer in addition to the use of CAD systems to access and manipulate
model data. Users at JM and Betongindustri are not used to CAD
environments and model access by a light and user-friendly model
viewer is preferred. The development of the model viewer required
considerable effort and resources, and has delayed the actual
application of the viewers in practice. However, model viewers have
started to be adapted in both organisations and are now part of newly
initiated research and development projects funded by the SBUF
(LKAB MK3 project no. 11636 and VIPP project no. 11693).

5
SBUF project 11333 ITstomme - Final Report

- The process of making cost estimates based on product models

appeared to be a challenging and time-consuming development process.
The development of cost estimation tools, also-called 5D functionality,
was successfully implemented in the model viewer. The 5D estimation
has not been applied in full scale in the ITstomme project.
- A prototype program, called BiDry, has been developed for the
calculation of the drying process of concrete slabs. With the final
version of the program it will be possible to interact with product model
data, such as material parameters, concrete volumes and thicknesses,
etc.

1.3.4 To be implemented

The following activities were originally planned for the ITstomme project, but
have not been implemented yet. These activities will be implemented during
autumn 2005 and spring 2006:

- 4D modelling and simulation will be performed in full scale by

construction planners at JM itself and will used to communicate and
optimize the schedule for the Hotellviken Ringvägen project.
- 5D estimation will be evaluated in a future project by JM in full scale.
- A study will be made at JM and Betongindustri of the direct and
indirect benefits and costs of model-based construction processes.
- The BiDry program for calculation of the drying process in concrete
slabs will be developed as a stand-alone program, with the possibility to
interact with product model data.

1.4 Organisation

A publicly available project website has been established, which contains a

short description of the project. The website can be found at:
http://construction.project.ltu.se/~ebygg. A project intranet site has been
established in addition to the public website for document management and
project planning.

The issues addressed in the project are rather diverse and many of the
implementing actors come from different organizations with varying
backgrounds. The project is therefore decomposed in several project
components. Every project component addresses a specific and manageable

6
 Introduction

topic of the overall project framework in each phase of the project. This led to
the development of a project matrix to define and manage the various parts of
the project. A schematic overview of this matrix is included in Appendix A.

1.5 Partners

The main project participants are:

- Betongindustri (Curt Arne Carlsson, Mats Emborg, Thorbjörn Dorbell)

- JM (Martin Asp, Niclas Engdahl)
- Luleå University of Technology (Rogier Jongeling, Thomas Olofsson,
Jan-Erik Johansson)
- Ceco (Daniel Thall, Anders Pettersson)
- Enterprixe (Jan-Olof Edgar, Pekka Karjalainen, Tapio Ristimäki)

During the research project Rogier Jongeling visited the Center for Integrated
Facility Engineering (CIFE) at Stanford University (March 2004 – July 2004)
where case study projects from this research project were used to perform
studies of 4D modelling and 4D analyses. This research was performed
together with Professor Martin Fischer and two of his PhD students; Claudio
Mourgues and Jonghoon Kim.

7
SBUF project 11333 ITstomme - Final Report

1.6 Finances

The project was financed by the SBUF and Betongindustri. All the project
participants contributed to this project with their own time. An overview of the
funding and own contribution for this project is included in the table below.

Table 1: Funding and contribution for the ITstomme project (KSEK).

Participants Own contribution Funding
JM 200

eBygg-LTU 150

Betongindustri 250 260

Enterprixe-DDD Sweden 70

Enterprixe Finland 70

Heidelberg Zement Group 100

SBUF 1040

Total 840 1300

1.7 This report

The definition and interpretation of product modelling are first discussed in

this report. Secondly, the product modelling system used in this study is
presented after which concrete supplier’s motivation to study product
modelling is discussed. Thirdly, different model dimensions of product
modelling are presented, illustrated by examples from case study projects. For
every model dimension a short abstract and introduction is given, followed by
a more detailed description of the results and analyses of the application and
development work. The chapters about model dimensions are followed by a
discussion of the main challenges for the use of product models in
construction. Finally, recommendations are given for further work with
product modelling for cast in place concrete construction.

8
 Introduction

1.8 Project publications

This project has led to a number of publications that are part of the PhD work
by Rogier Jongeling at Luleå University of Technology. The publications
provide detailed descriptions of the research and development work performed
for different parts of the project. This report summarizes and combines this
work and for further reading about specific topics we recommend reading the
publications (i.e. they can be denoted as project deliverables):

Jongeling, R., Olofsson, T., and Emborg, M. (2004) ‘Product modelling for
industrialized cast-in-place concrete structures.’ In P. Brandon, Li, H., Shaffii,
N., Shen, Q. (eds) INCITE 2004 - International Conference on Information
Technology in Design and Construction, Langkawi, Malaysia, 103-110.

Jongeling, R., Olofsson, T., and Emborg, M. (2004) ‘Modelling cast in place
concrete construction alternatives with 4D CAD.’ In A. Dikbas, Scherer, R.
(eds) ECPPM 2004 - eWork and eBusiness in Architecture, Engineering and
Construction, Istanbul, Turkey, 109-116.

Jongeling, R. (2004). ‘IT-stomme - produktmodeller i

stombyggnadsprocesser.’ Betong -Svenska Betongföreningens Tidskrift, 4, 40-
42

Jongeling R, Emborg M and Olofsson T (2005) ‘nD modelling in the

development of cast in place concrete structures’, ITcon Vol. 10, Special Issue
From 3D to nD modelling, 27-41, http://www.itcon.org/2005/4

Jongeling, R., Kim, J., Mourgues, C., Fischer, M., Olofsson T. (2005)
‘Quantitative Analyses Using 4D Models – An Explorative Study’, In Park
(eds) ICCEM 2005 - International Conference on Construction Engineering
and Management, Seoul, Korea, 830-835.

Jongeling, R., Kim, J., Mourgues, C., Fischer, M., Olofsson T. (2005)
‘Analyses of Extracted 4D Model Contents’, To be submitted to the Journal of
Automation in Construction.

The following publications are the result of research that has been conducted in
conjunction with this project:

Blokpoel, S. (2003). ‘Cooperation and Product Modelling Systems - The

Application of Product Modelling Systems in the Building Process.’ 2003:17,
Luleå University of Technology, Luleå, Sweden, pp. 126.

9
SBUF project 11333 ITstomme - Final Report

Blokpoel, S., Olofsson, T., and Jongeling, R. (2004) ‘Cooperation and product
modelling systems’ In A. Dikbas, Scherer, R. (eds) ECPPM 2004 - eWork and
eBusiness in Architecture, Engineering and Construction, Istanbul, Turkey,
423-429.

Carlsson, A., Helgesson, L. (2005). ‘Modellering och representation av

konstruktionsmoment med 4D-CAD relaterade till platsgjuten betong,’
Diploma Work, Luleå University of Technology, Luleå, Sweden, pp. 124.

10
 Product Modelling

2 PRODUCT MODELLING

The development and use of computer-based models for the AEC

(Architecture, Engineering and Construction) industry has been discussed
within international research and development communities for some time
(Eastman 1992; Fischer 2004; Gielingh 1989; Laitinen 1998). Different terms
and concepts are used in discussions to denote these models and modelling
systems. Recently, the concepts of a Building Information Model (BIM), nD
modelling and Virtual Building Environment (VBE) have been added to the
terminology describing information models for the AEC industry.

BIM was launched by major vendors of CAD applications such as Autodesk,

Graphisoft, Bentley and Nemetschek. A BIM is a computer model database of
building design information, which may also contain information about the
building’s construction, management, operations and maintenance (Graphisoft
2002). Research and development regarding nD modelling is currently mainly
conducted at the University of Salford within the 3D to nD modelling research
project (Lee 2004; Lee 2005). A nD model is an extension of the BIM, which
incorporates multi-aspects of design information required at each stage of the
lifecycle of a building facility (Lee 2003). A VBE is a ‘place’ where building
industry project staffs can get help in creating BIMs and in the use of virtual
buildings (Bazjanac 2004). A virtual building is a BIM, or an nD model,
deployed in software. A VBE consists of a group of industry software that is
operated by industry experts who are also experts in the use of that software
(Bazjanac 2004).

11
SBUF project 11333 ITstomme - Final Report

Throughout this report we use the following terminology:

- A product model refers to data models that contain both product and
process data supporting a building’s life cycle. Examples of such data
are geometry, planning and cost data. According to this definition we
define BIMs and nD models as product models.
- Product modelling systems are used to access, manipulate and store
information from exchange files and databases. Examples of such
systems are CAD applications, but also project planning software and
product model servers.
- We define the collection of product modelling systems used in a
project, including the professionals that operate the systems, as the
product modelling environment. We define VBEs as product modelling
environments.

Although definitions of product models and nomenclature vary within the

research and development communities, most actors agree that the main
advantage of product models lies in tasks beyond 3D modelling and generation
of drawings for a building (Fischer 2004). It is in this spirit that we explored
different dimensions of product models applied to cast in place concrete
construction beyond 3D modelling and 3D graphics. For this purpose we
applied a client-server modelling system to two case study projects. Before we
describe the motivation of the cast in place concrete industry to adapt product
modelling, we discuss the product modelling environment that is used in the
case studies of the ITstomme project.

12
 Product Modelling Environment

3 PRODUCT MODELLING ENVIRONMENT

Within the ITstomme project the product modelling work is performed with an
Internet-based product modelling system developed by Enterprixe Software
Limited (Enterprixe 2002). The system (i.e. the Model Server) uses a central
database in which the product model is stored. Additional databases containing
for example cost data or documents can be linked to the database, Figure 1.

When logged in, a project can be selected from the product model server and
specific client software to view and edit the product model. AutoCAD-based
software and a model viewer embedded in an Internet browser are used as
software clients to the product model server in the ITstomme project. Many
actors are already familiar with AutoCAD in Sweden and the transition to a
model-based practice using AutoCAD can possibly facilitate this process.

The model viewer is used for navigation in 3D and for modelling of 4D models
and 5D cost estimation. The model viewer is using VRML-graphics and was
specially developed within the ITstomme project by Ceco AB (Ceco 2005).
With this model viewer users can access product model data without having
AutoCAD installed on their computer.

Exchange file import and export is used in addition to direct client-server

connection to extend the product modelling environment. For example, IFC
files are imported from ArchiCAD to AutoCAD Architectural Desktop (ADT),
which in its turn is a client to the product model server. Industry Foundation
Classes (IFC) is a product model data standard developed by the International
Alliance for Interoperability (IAI) (IAI 2005). IFC objects are uploaded from
ADT to the central database, where they become available for all project
participants.

13
SBUF project 11333 ITstomme - Final Report

Figure 1: The product modelling environment of the ITstomme project. The

model server and document server are two different systems, but have been
integrated in the ITstomme project and act as one system. Different software
systems can be used to create, edit and use product model data. Examples of
these systems are AutoCAD, Microsoft Excel, Microsoft Project and Ceco
viewer.

The product model is loaded from the product model server to local computers
and constantly updated when connected to the server. The product model of
one of the case study projects contains approximately 3000 objects, which is a
workable amount of data to work with over the Internet. It is possible to load or
unload parts of the product model and in this way it is not necessary to have all
model data loaded to the local computer.

14
 Product Modelling Environment

The product model server keeps track of users who create an element or who
edit one. Figure 2 illustrates the partial model check out and check in
functionality. Product model objects that are checked in appear as normal
elements. Elements that are checked out by the current user carry a check-mark
(5). Checked out elements are locked for editing by other users. Elements
checked out by other users, cross-marked (6), are locked and cannot be edited.

Figure 2: A tree view of the product model at the product model server. Partial
product model check in and check out functionality is used to reduce the risk
for double work and product model inconsistency.

Part of the modelling work is concurrently carried out by multiple distributed

users within the ITstomme project. Project participants consider the partial
product model check in and check out one the hand a valuable product model
server functionality that reduces the risk for double work and product model
inconsistency, but consider it on the other hand constraining to be forced to
work online when editing product model data. On building sites or in project
meetings for example there is often not enough time or no connection available
to connect to the model server and then download specific parts of the project.
This is further constrained by system restrictions, such as firewalls.

One of the case study projects includes about 50000 objects and when all
loaded the performance of the system slows down considerably. The system
performance becomes insufficient to work with. These performance and
usability problems can partly be solved by improving the computing
performance and partly be developing an offline mode for the system.

The offline mode will allow users to work on the product model without being
connected to the model server. After their work the users can synchronize their

15
SBUF project 11333 ITstomme - Final Report

work with the central model at the model server. The offline mode
functionality is expected during the autumn of 2005 and is expected to
facilitate the work considerably.

16
 Cast in Place Concrete

4 CAST IN PLACE CONCRETE

There is an increasing interest of the ready mixed concrete suppliers in Sweden

in product modelling technology. The interest is the result of the identification
of the potential of product modelling technology for cast in place concrete. As
an example in this respect the process of virtual prototyping can be given of
innovative construction products and processes that are aimed at industrializing
cast in place concrete construction (Jongeling 2004a). The interest is also partly
the result of developments that are ongoing at competing building materials
and building systems, such as steel and prefabricated concrete.

The steel and prefabricated concrete industries have already started to use a
model-based construction approach. The main product modelling system
developments and standardization efforts to date have focused on these sectors.
This focus can partly be explained by the nature of the building technology.
Steel and prefabricated concrete are strongly component oriented in both
design and production as opposed to the design and production process of cast
in place concrete structures. Product modelling systems imply an object
oriented approach and are therefore well-suited for object oriented building
systems, such as the ones based on steel and prefabricated concrete
components.

One of the strengths of cast in place concrete is customisation of shapes and

material properties. Basically any shape can be created by forming the concrete
on site in tailor-made forms. The customisation and freedom of design partly
explain the absence of standard product catalogues for cast in place concrete
objects. However, the main reason for this absence is the fact that the product
is a material and not a building object. Standard product libraries that are
available in CAD applications for steel and prefabricated concrete objects do

17
SBUF project 11333 ITstomme - Final Report

not exist for the cast in place concrete industry. There is in other words no
explicit integration of the design object and supplied product in CAD systems
for cast in place concrete.

In addition to difficulties with product specification and design in CAD

systems there is no single definition of an object. Design objects differ from
production objects of which the latter is often not explicitly modelled. An
example of a production object is a concrete pour, i.e. a cast sequence, which
can be part of a design object. The design is used as a bounding box for
production solutions, as opposed to steel construction where design
information is used to direct production by Computer Numerical Control
(CNC). Most of today’s product models do not contain objects that are suitable
for detailed construction simulation and optimisation of cast in place concrete
structures.

Cast in place concrete construction is information intensive. Object properties

such as concrete quality and Water to Cement ratio are continuously updated
depending on changes in production planning and weather conditions. This
information is in our experience currently not integrated or associated with
production models by concrete suppliers.

4.1 Possibilities and challenges

We have identified a number of possibilities and challenges for the application

of product models for cast in place concrete construction. For every issue we
identified a dimension of the product model to which the issue applies. Most of
these issues are specific to the cast in place concrete industry, but some have a
more general character. However, especially the challenges have to be
addressed, specific to cast in place concrete or not, in order to apply and benefit
from product modelling technology.

18
 Cast in Place Concrete

Table 2: Possibilities for the application of product models for cast in place
construction
Possibilities Dimension
- Modelling and review in 3D improves constructability and 3D
quality of the structure.
- Product models can be linked and associated with 2D
drawings and documents, such as the results of technical 2D + 3D
calculations. This facilitates the process of document
retrieval.
- Simulation of innovative products and production
4D
processes facilitates the application of new technologies in
practice.
- (semi-)automated quantity take-off of concrete volumes
5D
and material specifications can save time and can reduce
errors in quantity take-off.
- Material specifications can be integrated in product
nD
models and are as a result a natural part of objects used in
design and production.

19
SBUF project 11333 ITstomme - Final Report

Table 3: Challenges for the application of product models for cast in place
construction
Challenges Dimension
- The architectural and structural views on modelling are 3D + 3D
different.
- There is limited support in CAD applications for
modelling of objects related to cast in place concrete
3D
construction. This applies to concrete shapes, definition of
concrete pours, reinforcement bars, isolation, drainage,
etc.
- Design models differ from production models in which
3D + 4D
the latter is more detailed and includes non-building
objects, such as scaffolding, work spaces, etc.
- Material specifications are often not specified in product nD
models
- There are very few actors that can use CAD which limits All
the use of product models in practice.
- There is currently no role defined for someone who
should coordinate the consistency of models from the All
architect and structural engineer and the models used for
cost estimation, planning and production.

These possibilities and challenges serve as a starting point for the development
and application of different model dimensions for the cast in place concrete
industry. Before these developments and applications are discussed, we will
introduce the case studies to which product modelling technology is applied in
the ITstomme project.

20
 Case Studies

5 CASE STUDIES

The main case study of the ITstomme project is a project by JM called

Hotellviken Marinan. In addition, a virtual project is studied that is mainly used
for the development and application of 4D simulations and analyses.

5.1 Hotellviken Marinan

The Hotellviken project comprises the construction of 120 apartments of which

the first phase, Hotellviken Marinan, is addressed in this report. Phase one
consists of five multi-storey apartment blocks (in total 25 apartments) of which
two are connected with a parking garage. Figure 3 shows a 3D model of the
concrete structure. The bearing structures are cast in place concrete. Facades
consist of standardized non-bearing prefabricated plastered elements.
Prefabricated lattice girder elements are used for slabs on which cast in place
concrete is poured. Traditional formwork is used for the inner walls in which
concrete is poured on site. The roof of the connecting garage consists of hollow
core slabs.

21
SBUF project 11333 ITstomme - Final Report

Figure 3: Bearing structure of the Hotellviken Marinan project.

Hotellviken is used to study and develop the following product model

applications:

- Traditional 2D drawings and documents are partly generated from 3D

models and hyperlinked to these models.
- Various types of 3D models from different disciplines are created by
using file-based data transfers and a collaborative client-server
environment.
- 3D models are mapped to cost estimation hierarchies providing a 5th
dimension.
- Model use and configuration by the ready-mixed concrete supplier
provides an additional dimension.

5.2 4D case study

In an effort to adequately support the introduction and evaluation of

construction innovations, Betongindustri initiated a pilot study using 4D CAD
simulations to evaluate two construction alternatives of a residential

22
 Case Studies

construction project, Figure 4. The 4D case study is used for the development
and application of 4D CAD.

Betongindustri conducted the experiments after the actual construction of the

project was finished. The experiments are realistic and, where possible, are
based on actual site data, but had no direct impact on the construction process
performed by the contractor.

Figure 4: The 3D model that is used in the 4D case study by Betongindustri.

The building measures 60 meters in length and is 12 meters wide. All displayed
slabs and walls are cast in place concrete components, except for balconies that
are prefabricated components.

The method and results of the 4D case study are discussed in Chapter 7. In the
next section, Chapter 6, we describe our application and development efforts
within this project regarding the second and third dimension of a product
model.

23
 The Second and Third Dimension

6 THE SECOND AND THIRD DIMENSION

6.1 Abstract

The application and developments regarding the second and third dimension of
product models for cast in place concrete focus on the use of multiple 3D
models that are combined with 2D drawings and documents. This combination
allows product models to be created by traditional design disciplines and
allows them to be used in conjunction with traditional documents that are used
in today’s processes. In this way the use of product models is complementary
to today’s document-based information flow. The main benefits of this
approach are the facilitation of design reviews and design coordination in 3D.
In addition the 3D models can be used for document retrieval and linkage of
information by different stakeholders in the projects.

6.2 Introduction

Researchers (Lee 2004) and software developers (Autodesk 2002; Graphisoft

2002) envision a database constructed with intelligent objects from which
different views of the information can be generated automatically; views that
correspond to traditional design documents such as plans, sections, elevations,
schedules, et cetera. As the documents are derived from the same central
database, they are all coordinated and accurate.

We identify a number of issues that currently limit the use of product models to
the extent envisioned in the above:

- First of all, generating views from product models is currently partly

possible. Product models do not necessarily contain all information that

25
SBUF project 11333 ITstomme - Final Report

is required to produce design views. Absence of information is due to

unavailability of adequate modelling tools, required effort to add this
information to product models and the effort to extract the information.
For example, modelling work of certain reinforcement bars is possible
in a limited number of CAD systems. Generating views from these
systems is constrained by national and local preferences of
reinforcement bar detailing in shop drawings.
- Secondly, views of product models differ between actors. For example,
a structural engineer models building objects differently from objects
modelled by an architect, Figure 5 B-C. Generating specific views from
a multi-disciplinary central model that contains all information is
constrained by these different views.
- Thirdly, certain information is associated with a model, but not
necessarily part of a model. Even in the most optimistic scenario for
model-based approaches, the vast majority of current project
information exits in the form of unstructured documents (Froese
2004b). At present there is very little linkage between information
technologies for working with unstructured document-based
technologies and model-based technologies.
- Finally, the number of actors in a construction project that can access
and operate software tools to generate database views is mostly limited
to actual product modellers. In addition to modellers we see a number
of actors that are merely viewers of product models, such as estimators,
planners, suppliers, subcontractors, customers, et cetera. These actors
do not necessarily have product modelling systems installed and lack
the knowledge to generate specific views from a product model.

6.3 Results

The following solutions are applied in the ITstomme project to combine 2D

data with multiple product models and to make this data and product models
available for all project participants:

- Separate architectural and structural models are created of Hotellviken

instead of an all-including single product model. All other design, such
as building services design, is done according to a traditional 2D
practice. The architectural model is modelled in ArchiCAD and
exported as an IFC2x file to AutoCAD ADT. The IFC2x file is
uploaded from AutoCAD ADT to the central database where it is

26
 The Second and Third Dimension

available for all project participants. The structural model is modelled

in AutoCAD ADT, used as client-software to the central database.
- Views are generated from 3D models to which 2D geometric primitives
are added in paper space. We call this a hybrid design document type.
The views are saved at the model server and can automatically be
updated with product model data when required. 2D data, such as
reinforcement bar detailing, can only manually be updated per view.
- Product model views and other documents are located at a document
server and are hyperlinked to the product model. Links are added to
specific model objects, but also to parts of an object or to specific
sections of a product model. For this purpose different types of pointer
objects are used in the models that contain links to the document server.
Different pointer objects are available for different disciplines to
facilitate information management, Figure 5 A.
- A model viewer is used as client software to view the product model in
the central database and to browse through hyperlinked data.

27
SBUF project 11333 ITstomme - Final Report

Figure 5: (A) A product model in AutoCAD ADT connected to the Enterprixe

product model server. Detail 1 illustrates different views on modelling by a
structural engineer (B) and an architect (C). The architect models the slab as
one volume, whereas the structural engineer models different concrete slabs
individually. (A) Detail 2 is an example of pointer object for a concrete slab to
which results from technical calculations are linked, which are located at a
document server.

6.4 Discussion

Working with two separate product models is considered to be beneficial, but

also shows limitations. Design coordination and design reviews are facilitated
by the use of 3D models. An advantage of a separate product model per design

28
 The Second and Third Dimension

discipline is that both the architect and structural engineer can have their own
view on their design practice, which they are familiar with. Legal concerns by
project participants are minimized with this approach, which can facilitate the
acceptance and uptake of 3D modelling. A disadvantage of this approach is the
lack of coordination between different models. Updates in the architectural
model that affect the structural model have to be propagated manually in the
structural model. Product model clash detection software (Commonpoint 2005;
Navisworks 2005) or model checking software (Solibri 2005) is not used in the
ITstomme project, but is believed to save time and increase the accuracy in the
process of coordination product models from different disciplines.

The process of generating views from 3D models and adding 2D data proves to
be feasible for the Hotellviken project. Difficulties are experienced when
updates are made in the central product model. Ensuring up-to-date 2D data in
all separate model views of for example reinforcement bars is a process that
does not provide significant advantages compared to the traditional 2D
structural design process.

Project actors that do not have CAD software installed can view and browse
the product model with inexpensive viewing clients. To illustrate: at the start of
the ITstomme project there was one CAD system available at JM and one at
Betongindustri. No experienced CAD personnel is available to operate CAD
systems in the two organizations. Using Internet explorer-based product model
viewers facilitates the uptake of product model use in both organisations.
Model views and other documents, located at a document server, become
centrally available by using product model viewers. Using the 3D model is
believed to add to project participants’ understanding of to what part of the
model the different documents and views are related.

It became clear during the project that there was a need for a central person that
would coordinate and ensure proper linking of the product model with
information located elsewhere. A number of other tasks and responsibilities
have been identified, in addition to management of linked information to the
product model. Examples of these tasks and responsibilities are:

- Management of the product model server

- Coordinating and ensuring the use of templates for modelling work
- Integration of product models from different disciplines
- Education and knowledge management of (potential) model users

29
SBUF project 11333 ITstomme - Final Report

Research efforts by Froese on the integration of product models with

document-based information (Froese 2004b) and on the definition of a Project
Information Officer (PIO) (Froese 2004a) can be mentioned here as potential
future developments that are relevant to integration and management of
product models with 2D data.

30
 The Fourth Dimension

7 THE FOURTH DIMENSION

7.1 Abstract

Within the ITstomme project 4D CAD is used as an instrument to introduce

construction innovations and to evaluate construction alternatives for cast in
place concrete construction. Architectural and structural 3D models serve as a
basis for the 4D modelling process in which these models are linked to
production schedules. The main benefits of this approach are that fact that
project stakeholders can simulate and analyze what-if scenarios before
commencing work on site. New construction materials or processes can be
tested virtually under controlled conditions before committing resources and
without risks for expensive failures.

7.2 Introduction

4D modelling is an increasingly used process method in which 3D CAD

models are visualized in a 4-dimensional environment. A currently widely used
method for process modelling is the Critical Path Method (CPM) which is
often graphically represented as a Gantt schema, i.e. a bar chart. The method
concentrates mainly on the temporal aspect of construction planning (i.e. the
time-aspect) and is seen as one-dimensional (Heesom 2004). Construction
projects have unique spatial configurations and the spatial nature of projects is
very important for planning decisions (Akbas 2004). CPM schedules do not
provide any information pertaining to the spatial context of project components
and requires users to look at 2D drawings to conceptually associate
components with the related activities (Koo 2000; Koo 2003). This approach
limits evaluation and comparison of alternative solutions. Construction plans

31
SBUF project 11333 ITstomme - Final Report

can be represented graphically by adding the time dimension to 3D models to

allow project planners to simulate and analyze what-if scenarios before
commencing work execution on site (Mallasi 2002). 4D modelling is identified
as a tool to convey planning information (visualization tool), enhance
collaboration among project participants (integration tool), and to support users
to conduct additional analyses (analysis tool) (Koo 2000).

4D CAD is used in the ITstomme project to evaluate today’s construction

practice and to introduce innovations construction materials and construction
processes that can add to the industrialization of cast in place concrete
construction.

Appendix B includes a detailed description the application of 4D CAD for cast

in place concrete in the ITstomme project. The main results are summarized
below. For a more detailed overview and discussion of the results we
recommend reading (Jongeling 2005b; Jongeling 2004a).

7.3 Results

Two alternative models are modelled of a 4D case study project (see section 0)
in order to compare different construction methods, Figure 6. The first
alternative, the 0-Reference scenario, represents today’s common practice for
cast in place concrete construction. The objective of this scenario is to
represent typical sequenced and concurrent activities on a construction site that
are related to casting of concrete walls and slabs. The second alternative
provides an industrialized approach to cast in place concrete construction. A
number of innovative production technologies form the basis for this
alternative. The objective is to visualize the potential for permanent formwork
systems in combination with the use of prefabricated carpets of reinforcement
and self-compacting concrete. Such a combination of innovative production
technologies has not been applied in actual projects in Sweden and the
possibility to evaluate these methods in a virtual environment is considered to
be useful.

32
 The Fourth Dimension

Figure 6: Parallel simulation and comparison of construction alternatives in 4D

CAD. (Left) Traditional production technology. (Right) Industrialized practice
combining innovative technologies, such as permanent formwork systems,
reinforcement carpets and self-compacting concrete. Colour settings in
simulations: Red = in activity, grey = finished elements, yellow = traditional
formwork and shoring elements. See also Appendix B.

Production models are used, in addition to architectural and structural models,

to suit detailed production simulation of both construction alternatives. CPM
schedules are imported to the database and linked to production model objects.
The architectural model and structural model are too abstract to be used for
specific construction operations. In order to create production models from
architectural and structural models:

- Changes are made in the 3D CAD object hierarchy. 3D CAD objects

are regrouped to represent work packages and activities
- Certain objects have to be split and regrouped. This especially applies
to large objects, such as slabs
- 3D CAD objects are added that are not present in the architectural
model. As an example traditional formwork can be given. The objects
solely serve a visual purpose and are abstract representations of actual
formwork elements

33
SBUF project 11333 ITstomme - Final Report

7.4 Discussion

Both construction alternatives can be simulated in parallel in 4D CAD. A

number of time-space conflicts were detected during the 4D modelling process
that had not been detected in the CPM schedule. The parallel visualization is
considered effective to visually explain the differences between the two
construction alternatives. The simulations have been evaluated by a number of
professionals and they generally agree that the 4D models help to understand
the different construction processes, but they also note that the models are
limited in scope and non-interactive.

Evaluation of alternative work flow strategies or changes in productivity can

not easily be managed in the 4D models. The 3D CAD objects in the
production models that are created and grouped to represent specific activities,
i.e. formwork, reinforcement and concreting, constrain the rapid evaluations of
alternatives. Changes in the architectural and structural model and changes in
the schedule imply often major changes in the production model. The detailed
production models require considerable modelling effort and are due to their
complexity and interdependencies difficult to manage. The complexity of the
CPM schedules that contains a large number of dependencies between
activities further constrains the management of the production models. Akbas
(2004) notes that when models accurately represent construction operations,
the model complexity increases significantly and consequently the effort to
create and maintain these process models.

The 4D models show activities related to building components, but no other

aspects that are related to such an activity. As an example the space usage of a
construction activity can be given. This is especially relevant for cast in place
concrete construction, where the work with forms, reinforcement and casting
on site implies different types of space usage, such as materials space, labour
space, space for safety measures, etc. The application of 4D CAD to model and
represent the space usage for cast in place concrete is studied within the
ITstomme project by students at Luleå University of Technology (Carlsson
2005). The results of the work are manuals for production planners and 4D
CAD modellers that include recommendations about what types of spaces that
should be represented for different types of activities during the casting process
of concrete walls and slabs.

34
 The Fourth Dimension

We consider a number of developments that can possibly address the issues

related to the creation and maintenance of space-loaded 4D CAD production
models (Jongeling 2004a):

- Adding CAD objects to represent for example temporary structures

could be partly automated by using feature-based 4D models. Feature
modelling is an approach whereby modelling entities termed features
are utilized to provide improvements for common geometric modelling
techniques (Kim 2004, unpubl.). The application of feature-based 4D
models could have reduced the modelling work and could have
contributed to the quality of the temporary structures plans.
- Although 4D simulations are a promising method to evaluate complex
schedules, the method still relies on CPM schedules. As noted in the
above, CPM schedules do not provide information about the spatial
context of production processes. An additional shortcoming of CPM is
the difficulty of identifying work flow in a production system. Research
initiatives within the Lean Construction Institute (LCI 2004) on the
development and application of Line-of-Balance (LoB) scheduling
techniques (Kenley 2004) can possibly provide an alternative for or
addition to CPM based scheduling for 4D CAD.
- Space objects could be used to represent activities and space-usage,
rather than using building objects as a representation. Space objects can
provide a better visualization than building components of the space
that is required for ongoing work in a specific section of a building.
Spaces can also help to minimize the 4D modelling work as these
spaces can include many different building components.

As a future extension of the ITstomme project and conducted 4D simulations,

we plan a study to evaluate feature-based 4D models and integration of LoB
planning software (DSS 2005) with 4D CAD. The combination of 4D CAD
models and LoB diagrams has been used in a PhD and Industry course, called
Virtual Construction, at Luleå University of Technology, during spring 2005
(Jongeling 2005a). The participants in the course considered the combination a
useful mechanism to understand the spatial context of the production planning
and to manage the flow of resources and materials through different locations
on the production site. An article has been submitted to a scientific journal in
which the combined use of 4D CAD models and the LoB planning method is
explored (Jongeling 2005c).

35
 The Fifth Dimension

8 THE FIFTH DIMENSION

8.1 Abstract

In addition to the 2nd, 3rd and 4th dimension, we studied and developed the
integration of product models with cost data to facilitate the cost estimation
process for cast in place concrete structures. This process is referred to as 5D
modelling (Edgar 2002), in which project model objects or assemblies of
objects are mapped with recipes that specify the activities and resources needed
for construction of these objects. The objects with a specified recipe are
subsequently used in cost estimation software to generate cost estimates. The
main benefits of this 5D approach are the savings in time for cost estimators in
the quantity take-off process and the reduced risks for omissions or mistakes in
the quantity take-off. The savings in time and the fact that cost estimates are
based on virtual models can possibly lead to a decision-making process in
which more alternative (innovative) cast in place concrete products and
processes are considered, which can lead to better informed decisions and
ultimately to a better final product.

8.2 Introduction

Currently the quantity take-off is done manually from paper drawings, which is
time-consuming and prone to errors. Each time the design is revised the
process has to be repeated and started over again. This is a cumbersome and
time-consuming process, and limits the consideration of alternative products
and processes during the cost estimation process. The goal of the development
and application of a 5D cost estimation process is to partly automate the cost

37
SBUF project 11333 ITstomme - Final Report

estimation process in order to facilitate the consideration of multiple

production alternatives.

In a typical project three types of cost estimates are made:

1. Pre cost estimate, which is based on experience ‘key values’

(Förhandskalkyl)
2. Base cost estimate, which is based on early stage drawings
(Grundkalkyl)
3. Production cost estimate, which is based on building drawings.

In the ITstomme project we focussed on the base (2) and production (3) cost
estimate. A detailed description of the research and development regarding the
5D cost estimation process in the ITstomme project is included in Appendix C.
The main results are summarized below.

8.3 Results

The cost estimation process with the use of product models is not as
straightforward as multiplying model object quantities with unit costs. Most of
the product model objects from the architectural and structural engineering
models cannot directly be used cost estimation. The objects require mapping
with recipes codes. A recipe code is an identifier of a recipe in the cost
estimation software used by the cost estimator. A recipe includes the working
methods, or activities, for the construction of a specific object and the needed
resources.

The structure, or hierarchy, of the cost estimation is in many cases different

than the structure of a product model and for this purpose, a separate estimation
hierarchy is created in the database that is mapped to model objects from the
architectural and structural object hierarchies.

The process of mapping includes the creation of an estimation hierarchy to

which product model objects are linked, Figure 7. The objects from the product
models can be selected from the product model hierarchy by filtering the
product models on specific parameters, such as ‘wall width > 200mm.’, or by
picking the objects with a computer mouse from the 3D models on a screen.
For every cost item in this estimation hierarchy the cost estimator can specify a
recipe code and the type of object, such as a slab, to which this code applies.
The recipe codes are available in a user-defined XML file and are created

38
 The Fifth Dimension

according to a SBEF classification system, which is in use at JM. SBEF is a

classification system and file format for exchange of cost calculation data.
After having specified a recipe code for all model objects the estimator can
export the quantities and the recipe codes to cost estimation software where the
cost estimates are completed.

Figure 7: Screen capture from the 5D cost estimation process by using a model
viewer developed in the ITstomme project. 3D objects are selected by an
estimator in the 3D scene (left) or via the PE (Project Explorer) (right) after
which a cost estimation post is created in a so-called 5D hierarchy (on the left
of the PE) for which a recipe code is defined. A cost estimation file is created
from the 5D hierarchy and is used in cost estimation software. See also Figure
22 and Figure 23 in Appendix C.

8.4 Discussion

The process of selecting objects, adding recipe codes and exporting the cost
estimation files to cost estimation software requires configuration of recipe
files according to a company standard and requires training of cost estimators
in the use of the software. Technically there are no major obstacles to start
using the 5D technology.

39
SBUF project 11333 ITstomme - Final Report

Different approaches to the 5th dimension of product models have been

demonstrated in a number of research initiatives (Edgar 2002; Kam 2002;
Laitinen 1998) and are starting to be commercially applied (Graphisoft 2005).
A number of companies currently emerges in Sweden that provide cost
estimation services based on product modelling for project developers and
construction companies. These companies use product models for a variety of
analyses in addition to 5D modelling, such as energy simulation, lighting,
accessibility, et cetera. Creating a product model solely for estimation purposes
is considered a viable business case for these companies. Additional analyses
that can be performed with product models add to the value of the models.

The use of recipes for cost estimates in projects has advantages and
disadvantages for cast in place concrete construction. By using standardized
object types and recipes one can enhance standardization of products and
processes, and can semi-automatically generate a 4th dimension (Jongeling
2004b). However, cast in place concrete is not a single standard product and
has no standard process. The products and processes are to a certain extent
standard, but can require customization of for example material properties.
With standard recipes for cast in place concrete one can only use known
product and process information resulting in solutions not adapted to the actual
construction site. There is a risk that models will contain generic information
from for example cost estimation databases that is not tailored and checked by
specialists. In this spirit we suggest that expert knowledge from concrete
suppliers is integrated in product models. We discuss this process in the next
chapter.

40
 The Nth Dimension

9 THE NTH DIMENSION

9.1 Abstract

Within the ITstomme project we illustrated a product model’s nth dimension by

the development of a program called BiDry. BiDry serves as a typical example
for the integration of what we call production data into a product model. The
program is used to calculate the optimal drying process of concrete slabs. The
main benefit of this approach is the explicit integration of design and
production data. Production data can easily be accessed via a product model
and the production data itself is based on parameters from the product model,
thus minimizing the risks for communication and data errors and ensuring the
use of the right concrete products and production processes.

9.2 Introduction

The drying time of slabs often proves to be a bottleneck in projects, leading to

costly delays. Slabs have to have reached a relative humidity of 85%, or
sometimes 90%, before they can be covered by other floor material. A relative
humidity higher than 85% or 90% can result in moisture damage, such as
mould, which is negatively affecting the indoor climate of a building
(Hedenblad 1996). Different types of concrete can result in faster or slower
drying processes of slabs and can be delivered according to the requirements of
the contractor. It is thus important to use the right type of concrete at the right
place in a project and not to cover the slabs too early.

The process of ordering the right concrete and monitoring the relative humidity
requires close communication between the contractor and concrete supplier. In
today’s projects this communication is done by telephone, sending faxes,

41
SBUF project 11333 ITstomme - Final Report

emailing, etc. In this process there is no explicit integration between the results
from technical calculations and the design data that is produced by consultants,
such as architectural and structural models. In the ITstomme project we
illustrate the integration of production data and product data, by integrating
BiDry and a model viewer.

9.3 Results

The output of BiDry consists of time frames for the drying process of slabs and
material properties, such as specific concrete classes and Water to Cement
ratios, Figure 8. The material properties, determining the drying process, are
configured in the product model of Hotellviken by using model viewers to the
central product model server. The time frames are integrated in the
construction planning and linked to the product model. The model can be
browsed in 4D, using the model viewers. Standardized property reports,
including product data specifications, are linked to slabs in the product model,
Figure 5 A and Figure 9, in addition to configured slab properties and 4D
simulations.

42
 The Nth Dimension

Figure 8: BiDry, a program to calculate the optimal drying process of concrete

slabs, is used as an example of the integration of product data into a product
model. The output from the BiDry program is used to configure product model
object properties. Planning data from BiDry is linked to the product model and
used for 4D simulations. Reports from BiDry are located at a document server
and linked to the product model.

43
SBUF project 11333 ITstomme - Final Report

Figure 9: Properties of a selected (yellow) slab in the product model by the

structural engineer. The slab contains a link to a document server on which the
results of a technical calculation by the concrete supplier are located. The slab
contains also time stamps, indicating when construction work will be ongoing.

9.4 Discussion

The relevance of configuring product model contents by a ready-mixed

concrete supplier is illustrated by using an expert program for calculation of
concrete slab properties. The resulting model properties and 4D simulations
that follow from this process are an example of a potential business case for the
ready-mixed concrete supplier.

Technically there are no major obstacles to start cooperating in projects by

using product models in a similar or identical way as illustrated in the example

44
 The Nth Dimension

above. However, it requires training from different stakeholders in a project.

Furthermore it requires also clear formulation of the objectives for the use of
product models.

We illustrated a number of use cases that can be based on the use of product
models, but despite the potential benefits and possibilities the uptake is a slow
process. In the next chapter we discuss the challenges for applying product
models in practice.

45
 Challenges for Product Modelling in Practice

10 CHALLENGES FOR PRODUCT MODELLING

IN PRACTICE

Despite the possibilities of different model dimensions reported in this report,

the uptake is slow. Most of the modelling and coordination work was
performed by a few individuals outside the core project team of Hotellviken.
Motivating participating companies to commit resources for product modelling
in future projects is hard due to difficulties in explicitly communicating
benefits from model development and applications in the Hotellviken and 4D
case study project.

Difficulties in explicitly demonstrating the benefits of product modelling is one

reason why the uptake is slow, but there is a variety of other reasons why the
comprehensive use of product modelling is limited to date. These reasons, or
rather challenges, were discussed at an international workshop held in January
2003, as part of the 3D to nD modelling project at the University of Salford.
Thus, the five biggest challenges for the use of product modelling according to
the workshop are the following (in prioritized order) (Lee 2004):

- Improving education and changing the industry’s culture

- Implementation and integration
- Demonstrating the benefits and value of a product modelling system
- Data issues, such as multiple design perspectives
- Developing a common data standard for interoperability

We will discuss these five challenges in relation to the findings and efforts
from the ITstomme project.

47
SBUF project 11333 ITstomme - Final Report

10.1 Education and culture

Improving education and changing the industry’s culture were seen as the
biggest challenges by workshop participants for an uptake of product
modelling. Getting industry convinced of benefits and motivating professionals
to increase their knowledge about product modelling would greatly facilitate
the industry uptake.

If was found that professionals have difficulties in allocating time for education
and implementation in addition to their ongoing project work. Due to a lack of
education and implemented use cases it was difficult to motivate project actors
to consider product modelling as an alternative or improved practice compared
to their traditional 2D practice. Most of the modelling work and
implementation was therefore performed by a few individuals that limited the
learning experience for the total project team.

We suggest the allocation of resources within an umbrella organization, rather

than in a single project, to facilitate the uptake of product modelling.
Experience and gained knowledge from product modelling has a bigger chance
to sustain within organizations, than in a project team that in many cases ceases
to exist after a project has been finished.

In addition to strategic uptake of product modelling at a company level, we

suggest partnering with other organisations and actors to form a product
modelling environment, envisioned by Bazjanac (2004) as a Virtual Building
Environment (VBE). A ‘place’ is needed where companies and individuals can
get help in creating and using product models across different disciplines.

10.2 Implementation and integration

Implementation and integration was seen as the second biggest challenge by

the workshop participants. There are almost no commercial software
applications that work with and add to product models beyond 3D modelling
(Fischer 2004).

We illustrate different dimensions of product models by linking different

applications and by combining results from these applications. For example,
2D data located at a document server is linked to 3D models. 3D models are
used for 4D simulations and 5D cost estimation. In addition to 5D modelling
we illustrate a potential nth dimension by integrating property data from a

48
 Challenges for Product Modelling in Practice

supplier in a product model. These examples illustrate the use of product

models beyond 3D modelling.

Although commercially available applications are used for most of the

modelling work, there are a number of work-arounds and implementations
necessary to enable the transition from 3D to nD product model use. Project
participants should be able to access the product models by using software,
which they already are familiar with or with software which is very easy to
learn. This can greatly facilitate the uptake and culture change in the project
team towards product modelling.

We suggest starting implementation and integration efforts with existing and

commercially available systems as a basis. Sophisticated applications are
already available for 3D modelling, production scheduling, cost estimates, et
cetera. The challenge is to combine these applications, rather than to develop
an all-inclusive modelling system. We illustrated the combination of different
commercial applications in the ITstomme project, instead of developing an
own application tailored to our needs. An incremental approach should be
adapted in combining different applications with end-user participation in the
development process.

10.3 Demonstrating benefits

Demonstrating the benefits and value of a product modelling system was

considered the third biggest challenge. Business cases are needed that outlay
the needs for product modelling in projects and within organisations.

As noted in the above, motivating participating companies in the ITstomme

project to commit resources for product modelling in future projects was
difficult due to difficulties in explicitly communicating benefits from model
development and applications in the Hotellviken project. We plan to address
this shortcoming by adapting formalized cost benefit analyses developed for
the construction industry (Fox 2004).

In addition to cost benefit analyses we suggest performance measurement by

Key Performance Indicators (KPI) (Blokpoel 2003), illustrated by a number of
product models from different disciplines. One of the tasks of a possible
Project Information Officer (PIO) (Froese 2004a) or product model manager
could be the collection and analyses of model data to support performance
measurement and demonstration of product model benefits.

49
SBUF project 11333 ITstomme - Final Report

10.4 Data issues

Data issues were considered a big challenge, but not a high priority to address
in research and development of product modelling. Examples of data issues are
multiple design perspectives and the information flow, exchange and accuracy
in projects.

As a result of different design perspectives it was decided to work with two

separate product models; one architectural and one structural model. A hybrid
product model view is adapted in addition to these different models in which
3D model data is combined with manually added 2D data. The architectural
and structural models are combined in two production models, in which both
models are extended and further detailed.

The multi-model approach implies a number of advantages and disadvantages.

We believe that separate models per discipline are feasible, but only if
adequate tools are used to coordinate these models and to check the
consistency of these models. Highly detailed production models are labour
intensive to create and to maintain, but are considered valuable instruments to
communicate process intend and to evaluate production strategies in a virtual
environment, without committing resources on site.

The hybrid document approach proves to be more problematic than the multi-
model approach, due to errors in updates of 2D data. With constantly
improving commercial 3D modelling tools we believe that a hybrid approach is
not a necessity in the future.

10.5 Interoperability

Developing a common data standard for interoperability was perceived by

workshop participants as a big challenge, but not a high priority to address.
Considerable efforts by the IAI (IAI 2005) to create the IFC product model
standard have been made in the last years. Although promising (Kam 2002),
the IFCs have not yet found wide acceptance among software vendors and
construction companies (Fischer 2004).

Interoperability is not a major issue in the ITstomme project as a result of the

use of one environment. An architectural model was exported from ArchiCAD
and imported to AutoCAD ADT by using IFC2x. The model contains basic
objects, such as walls, doors, windows and slabs, and did not result in major
data losses. The import from ADT to the product model server proved to be

50
 Challenges for Product Modelling in Practice

more problematic, but was solved in the course of the project. Interoperability
is considered essential to start using product models beyond 3D CAD
applications. However, the ITstomme project also shows that different
dimensions can be added to a product model independently from standardized
data schemas.

51
 Discussion

11 DISCUSSION

The ITstomme project illustrates a number of dimensions of product models

that on the one hand are aimed at supporting new and existing products and
processes for cast in place concrete construction and on the other hand are
aimed at showing the potential of using product models beyond 3D graphics.
The development of support for cast in place concrete construction is mainly
studied in the 4th and nth dimension of this study. The potential of using product
models beyond 3D graphics is shown by combining 3D models from different
design disciplines that form the basis for document retrieval, 4D modelling and
5D cost estimation.

Developments and applications in the project are mainly driven by the interests
of the ready-mixed concrete supplier, Betongindustri, who, in a prestudy
(Jongeling 2003) identified product modelling as a threat and as an opportunity
for its business process. Modelling cast in place concrete structures implies a
number of modelling challenges compared to steel and prefabricated concrete
structures, such as the differences in definition of design and production
objects. However, the case study projects of this study show that these
challenges do not exclude the use of product models for cast in place concrete
structures. Possibilities and challenges that are listed in paragraph 4.1(Table 2
and Table 3) are included in Table 4 and Table 5 below, and commented on
based on results from the different model dimensions of this study. The
challenges as listed in Table 5 are complementary to the challenges that are
identified in Chapter 10.

53
SBUF project 11333 ITstomme - Final Report

Table 4: Findings from the ITstomme project that are related to the identified
possibilities for the application of product models for cast in place construction
Possibilities Findings from the ITstomme project
By using the 3D structural model in the Hotellviken
Marinan project we could identify in an early stage a
bearing wall that was not located in the same vertical
Modelling and review
plane as a bearing wall at a lower floor. Although
in 3D improves
this issue also was identified by the structural
constructability and
engineer with 2D drawings, the issue became much
quality of the
more explicit and could easily be detected in the 3D
structure
model. In general the use of 3D models is believed to
increase the accuracy and quality of the design and
provides a clear basis for decision-making.

Special markers are used that carry different colours

for different types of documents, Figure 5. The
Product models can
markers, or cone objects, link the structured
be linked and
information from the product models to unstructured
associated with 2D
information in documents and drawings. In the
drawings and
ITstomme project we linked information from the
documents, such as
architect, structural engineer and concrete supplier to
the results of technical
product models. This approach facilitates the
calculations. This
combined use of 2D drawings and 3D models, and
facilitates the process
helps stakeholders in the project to quickly find
of document retrieval
documents that are related to specific objects in the
product model.

Parallel visualization of different construction

Simulation in 4D of alternatives is considered an effective way to visually
innovative products explain differences between alternative construction
and production approaches, Figure 6. In the 4D case study we were
processes facilitates able to identify a number of time-space conflicts in
the application of new the 4D models that we had overlooked in our CPM
technologies in schedules. The 4D models require considerable
practice. modelling efforts and could possibly be improved by
using different scheduling and modelling methods.

54
 Discussion

After having prepared property sets for all objects the

quantity take-off is literally a process of a few
minutes. These quantities provide a quick and
(semi-)automated
reliable overview of the total amounts of concrete,
quantity take-off of
windows, doors, etc, that are needed in a project.
concrete volumes and
However, quantities also have to carry a recipe code
material specifications
and location to be used for cost estimation purposes.
can save time and can
This process is developed in the 5D cost estimation
reduce errors in
process. The 5D process saves time in quantity take-
quantity take-off
off and reduces risks for calculation errors, but still
requires input from the cost estimator about how and
where objects will be calculated.

Material By using model viewers one can navigate to specific

specifications can be components in the 3D models and can retrieve object
integrated in product properties by selecting objects in the 3D scene.
models and are as a Available information for objects, such as the
result a natural part of volume, length, material, construction dates, etc., is
objects used in the easily accessible and is considered useful by
design and production participants in the ITstomme project for planning of
process production and ordering concrete deliveries.

Table 5: Findings from the ITstomme project that are related to the identified
challenges for the application of product models for cast in place construction
Challenges Findings from the ITstomme project
Separate architectural and structural models are used
in the ITstomme project. We believe this approach
The architectural and
minimizes the need for radical process changes and
structural views on
minimizes legal concerns. A disadvantage of this
modelling are
approach is the lack of coordination between
different
different models. Product model clash detection
software is recommended in this respect.

55
SBUF project 11333 ITstomme - Final Report

There is limited
The structural detailing of cast in place concrete
support in CAD
structures is not part of the ITstomme project.
applications for
Therefore the main structural modelling work is
modelling of objects
carried out with standard and general modelling
related to cast in place
tools, such as AutoCAD ADT. Only the main
concrete construction.
concrete components are modelled of which the
This applies to
results are satisfying. All shapes can be represented
concrete shapes,
by using standard functionality in the program. For
definition of concrete
detailed modelling we recommend the evaluation of
pours, reinforcement
dedicated software packages for structural design of
bars, isolation,
cast in place concrete structures.
drainage, etc.

Production models consist of objects from

architectural and structural models to which
Design models differ additional objects are added to represent non-building
from production objects. We created detailed 4D models (production
models in which the models) that required considerable modelling efforts.
latter is more detailed It is not clear how the relations between the different
and includes non- models should be managed, in case the architectural
building objects, such model for example is updated. We plan to study the
as scaffolding, work use of a model server that can manage these relations.
spaces, etc In addition we plan to study planning and modelling
approaches that minimize the required modelling
efforts for production models.

Material We illustrate the possible integration of results from

specifications are technical calculations by a concrete supplier with
often not linked to or design models. Although the implementation is
specified in product limited in scope it provides an example of a possible
models business case for a ready-mixed concrete supplier.

At the start of the ITstomme project there was one

There are very few
CAD system available at JM and one at
actors that can use
Betongindustri, thereby strongly limiting the access
CAD which limits the
to product models. Using product model viewers
use of product models
facilitates the uptake of product model use in both
in practice
organisations.

56
 Discussion

There is a need for It became clear during the project that there was a
someone who need for a central person that would coordinate the
coordinates the linking, updating and use of multiple product models.
models from and for Within the ITstomme project this person was a
different stakeholders member of the research team. In construction projects
in a project. this person should be a part of the project team.

From the tables it becomes clear that advantages and possibilities can be
identified for the ready-mixed concrete supplier and contractor. We believe
that other actors in construction also can identify advantages and possibilities
in the different model dimensions of this study. When actors have identified
benefits of product modelling for their own business process they might be
more willing to participate in shared product model use in search of benefits
for the project team as a whole.

The project developer in this study, JM, started to identify benefits of product
modelling for its business process during the ITstomme project and is now
actively searching for additional benefits by participating in research projects,
such as the VIPP project, and by applying product modelling in practice.
Examples of the currently identified benefits by JM are:

- Improvement of the quality and effectiveness of product marketing by

using 3D models.
- The quality of design reviews and design coordination improves, which
in its turn is believed to provide a better basis for planning and
production, with fewer design mistakes and less rework.
- 5D cost estimation based on product models reduces the time needed
for quantity take-offs and reduces the risks for mistakes in the
estimates.
- Communication with project stakeholders improves by using 3D and
4D models. Especially the communication with the municipality about
building permits and the communication with clients benefits from
using 3D and 4D models.

11.1 Recommendations

Throughout the report a number of technical and organisational

recommendations are made. In the following section we summarize these

57
SBUF project 11333 ITstomme - Final Report

recommendations and provide suggestions for further research and for use in
practice. This report is written for audiences with differences in background
and expertise. These differences are also reflected in the recommendations that
we provide, of which some are general and others are more case- or actor-
specific.

11.1.1 Model management

Separate product models are created in the ITstomme project per design
discipline. We believe this approach minimizes the need for radical process
changes and minimizes legal concerns. A disadvantage of this approach is the
lack of coordination between different models. Product model clash detection
software (Commonpoint 2005; Navisworks 2005) or model checking software
(Solibri 2005) is recommended for evaluation with the objective to save time
and increase the accuracy of the process of coordinating and updating product
models from different disciplines.

In a number of companies the roles of the architects and structural engineers

have changed as a result of a model-based practice. The structural engineer’s
role has become broader and can be described as a building information
modeller. This includes the modelling of structural and architectural objects in
one single model, including the preparation of the model for cost estimation
and production simulation. The role of the architect is more focussed on
capturing client requirements and communicating these to the building
information modeller. The companies experience this change as a satisfying
approach. However, the approach raises many questions and we recommend
case studies of companies where such changes in processes have been made.
These case studies can provide a better picture of the advantages and
disadvantages of this new process.

It became clear during the project that there was a need for a central person that
would coordinate the linking, updating and use of multiple product models.
Additional tasks and responsibilities for this person are:

- Management of the product model server, such as model exchange and

updates
- Coordinating and ensuring the use of templates for modelling work
- Integration of product models from different disciplines
- Education and knowledge management of (potential) model users
- Providing an (up-to-date) information base for 4D and 5D modelling

58
 Discussion

Research efforts by Thomas Froese on the integration of product models with

document-based information (Froese 2004b) and on the definition of a Project
Information Officer (PIO) (Froese 2004a) should be further studied in this
respect.

11.2 Model creation and access

Structural design was performed with AutoCAD ADT in the project with
satisfying results. However, we also recommend evaluating dedicated software
packages for structural design of cast in place concrete structures, such as
Tekla Structures (Tekla 2005), Enterprixe Structural (Enterprixe 2005), Impact
and FEM-design (Strusoft 2005).

It is recommended to broaden the scope of research for product modelling

applied to cast in place concrete construction by including civil works, such as
bridges and tunnels.

Advanced 3D modelling tools are available for the design of HVAC and
electric installations in buildings. We recommend studying the integration of
models that are produced by these design disciplines with models from
architects and structural engineers.

Model viewers are used in the ITstomme project to facilitate the access and use
of product models by non-CAD users. These model viewers have a great
potential to facilitate the uptake of model use and more research is
recommended to design and optimize suitable user-interfaces, by for example
studying game consoles, available tools in the manufacturing industry, etc.

The 4D modelling case study that is presented in this report shows a number of
benefits for the ready-mixed concrete supplier and we believe there are many
more benefits for other actors in the construction industry. Examples of these
potential benefits are improving the instruction of sub-contractors, improving
Supply Chain Management (SCM), etc. We recommend applying 4D
modelling in practice at construction sites to further explore the benefits of this
technology. Reviewing the research and application efforts by Johan Sjögren
are recommend as a point of departure and source of inspiration (Sjögren
2005).

59
SBUF project 11333 ITstomme - Final Report

A number of research and development efforts are recommended to facilitate

the 4D modelling process:

- Feature-based 4D models could minimize and rationalize the 4D

modelling process. Practical test are necessary with the research results
from Kim (Kim 2004, unpubl.).
- Space objects could be used to represent different types of spaces and
locations that are required during the construction process by different
actors on a construction site. The use of spaces to model and represent
different types of construction activities should be further explored.
- The Line-of-Balance (LoB) planning method is a promising planning
method to be used in combination with 4D CAD models. The combined
use of 4D CAD models and the LoB planning method should be further
researched.

Application of the 5th and nth dimension in a similar or identical way as

illustrated in the report is not limited by technical obstacles. It requires training
from different stakeholders in a project and a number of test cases to make the
technology robust enough for different use cases. In addition to training and
test cases, it requires well-documented protocols how the technology can and
should be applied.

The 5D cost estimation process could be improved by creating the product

models from the start according to company standards for the cost estimation
process. This requires that consultants, such as architects and structural
engineers, create their models with objects from a customized and company-
specific library that includes the right codes and objects styles that match to
styles and codes in the cost estimation database.

11.3 Organizational

We identify a number of organizational challenges for the use of product

modelling. The first challenge is education and culture. In order to have
practitioners adapting product modelling technology they should have the
resources and power to do so. This requires strategic decisions in companies
that allow employees to try and learn the use of product modelling technology.
To facilitate this process we suggest the allocation of resources within
organizations, rather than in a single project to facilitate the uptake of product
modelling.

60
 Discussion

We suggest starting the application of product modelling with small and well-
defined examples internally. When benefits are identified for the internal
processes, actors can participate in shared product model use in search of
benefits for the project team as a whole. We suggest using formalized cost
benefit analyses to explicitly express the benefits from model developments
and model application (Blokpoel 2003; Fox 2004).

In addition to a strategic uptake of product modelling at a company level, we

suggest partnering with other organisations and actors to form a product
modelling environment where experiences can be shared and where companies
and individuals can get help in creating and using product models across
different disciplines.

A key factor and main challenge in the uptake of product modelling technology
is education of practitioners. A course on Virtual Construction was organized
at Luleå University of Technology during spring 2005 in an effort to provide
such education. We recommend practitioners from different disciplines to
participate in such courses. We also recommend setting up additional and
dedicated courses for different types of model creators and users.

In addition to courses organized by academic institutions or by companies,

there are many different seminars to choose from or publications to read. These
are recommended to be further informed and updated about the development
and use of product modelling technology.

Finally, we suggest considering product modelling technology in conjunction

with research and development efforts that are made regarding
industrialization, partnering and lean construction. These three areas of
research are popular themes in the discourse about methods to improve
construction and we believe that product modelling technology plays a
prominent and facilitating role across these themes.

61
 References

References

Akbas, R. (2004). "Geometry-based modeling and simulation of construction

processes," PhD Thesis, Stanford University, Stanford, CA.
Autodesk, Inc. (2002). "Building Information Modeling - White Paper."
www.autodesk.com/buildinginformation.
Bazjanac, V. (2004) "Virtual building environments (VBE) - applying
information modeling to buildings." In A. Dikbas, Scherer, R. (eds)
ECPPM 2004 - eWork and eBusiness in Architecture, Engineering and
Construction, Istanbul, Turkey.
Blokpoel, S. (2003). "Cooperation and Product Modelling Systems - The
Application of Product Modelling Systems in the Building Process."
2003:17, Luleå University of Technology, Luleå, Sweden.
Carlsson, A., Helgesson, L. (2005). "Modellering och representation av
konstruktionsmoment med 4D-CAD relaterade till platsgjuten betong,"
Luleå University of Technology, Luleå, Sweden.
Ceco. (2005). "Ceco Interactive Design - Model Publisher & Viewer."
www.ceco.se.
Commonpoint, Inc. (2005). "Commonpoint Project 4D."
www.commonpointinc.com.
DSS. (2005). "Dynamic System Solutions, Line-of-Balance planning systems."
www.dss.fi.
Eastman, C.M. (1992). "Modeling of buildings: evolution and concepts."
Automation in Construction, 1(2), 99-109.
Edgar, J-O. (2002). "Prod-IT (98324) 3D Produktmodell som 4D-
Produktionsmodell (in Swedish)." IT Bygg och Fastighet 2002,
Sweden.
Enterprixe, Ltd. (2002). "Enterprixe White Paper." www.enterprixe.com.

63
SBUF project 11333 ITstomme - Final Report

Enterprixe, Ltd. (2005). "Enterprixe Model Server & Enterprixe Structural

Design Software." www.enterprixe.com.
Fischer, M., Hartmann, T., Rank, E., Neuberg, F., Schreyer, M., Liston, K., and
Kunz, J. (2004) "Combining different project modelling approaches for
effective support of multi-disciplinary engineering tasks." In P.
Brandon, Li, H., Shaffii, N., Shen, Q. (eds) INCITE 2004 -
International Conference on Information Technology in Design and
Construction, Langkawi, Malaysia, 167-182.
Fox, S. (2004). "The ICT Benefit Formula - First Interim Report." VTT
Building and Transport, Espoo, Finland.
Froese, T.M. (2004a) "Help wanted: project information officer." In A. Dikbas,
Scherer, R. (eds) ECPPM 2004 - eWork and eBusiness in Architecture,
Engineering and Construction, Istanbul, Turkey, 19-24.
Froese, T.M. (2004b) "Integration of product models with document-based
information." In A. Dikbas, Scherer, R. (eds) ECPPM 2004 - eWork
and eBusiness in Architecture, Engineering and Construction, Istanbul,
Turkey, 85-89.
Gielingh, W. F. (1989). "General AEC Reference Model (GARM) ISO TC
184/SC4/WG1 Document N329."
Graphisoft, Inc. (2002). "The Graphisoft Virtual Building: Bringing the
Information Model from Concept into Reality - Graphisoft
Whitepaper." www.graphisoft.com.
Graphisoft, Inc. (2005). "Graphisoft 5D Estimator." www.graphisoft.com.
Hedenblad, G. (1996). "Drying of construction water in concrete - Drying
times and moisture measurement." The Moisture Research Group, Lund
Institute of Technology, Lund, Sweden.
Heesom, D., and Mahdjoubi, L. (2004). "Trends of 4D CAD applications for
construction planning." Construction Management and Economics,
February 2004, 22, 171-182, February 2004(22), 171-182.
IAI. (2005). "International Alliance for Interoperability." http://www.iai-
international.org.
Jongeling, R. (2003). "Product modelling for cast in place concrete - a
feasibility study." Luleå University of Technology, Luleå, Sweden.
Jongeling, R. (2005a). "Nya planeringsverktyg för effektivt betongbyggande."
Bygg & Teknik, 15-17.
Jongeling, R., Kim, J., Mourgues, C., Fischer, M., Olofsson T. (2005b).
"Quantitative Analysis Using 4D Models." To be submitted to the
Journal of Automation in Construction.
Jongeling, R., Olofsson T. (2005c). "A method for planning of work-flow by
combined use of location-based scheduling and 4D CAD." Submitted to
Automation in Construction.

64
 References

Jongeling, R., Olofsson, T., and Emborg, M. (2004a) "Modeling cast in place
concrete construction alternatives with 4D CAD." In A. Dikbas,
Scherer, R. (eds) ECPPM 2004 - eWork and eBusiness in Architecture,
Engineering and Construction, Istanbul, Turkey, 109-116.
Jongeling, R., Olofsson, T., and Emborg, M. (2004b) "Product modelling for
industrialized cast-in-place concrete structures." In P. Brandon, Li, H.,
Shaffii, N., Shen, Q. (eds) INCITE 2004 - International Conference on
Information Technology in Design and Construction, Langkawi,
Malaysia, 103-110.
Kam, K., Fischer, M. (2002). "Product Model & 4D CAD - Final Report."
Center for Integrated Facility Engineering, Stanford University,
Stanford, CA.
Kenley, R. (2004) "Project micromanagement: practical site planning and
management of work flow." In S. Bertelsen, Formoso, C.T. (eds) IGLC-
12, 12th Conference of the International Group for Lean Construction,
Helsingor, Denmark, 194-205.
Kim, J. (2004, unpubl.). "Generating temporary structures with feature-based
4D models." Department of Civil and Environmental Engineering,
Stanford University, Stanford, CA.
Koo, B., and Fischer, M. (2000). "Feasibility Study of 4D CAD in Commercial
Construction." Journal of Construction Engineering and Management,
126(4), 251-260.
Koo, B., and Fischer, M. (2003). "Formalizing Construction Sequencing
Constraints for Rapid Generation of Schedule Alternatives." 75, Center
for Integrated Facility Engineering, Stanford University, Stanford, CA.
Laitinen, J. (1998). "Model based construction process management," PhD
thesis, Royal Institute of Technology (KTH), Stockholm, Sweden.
LCI. (2004). "Lean Construction Institute." www.leanconstruction.org.
Lee, A., Marshall-Ponting, A.j., Aouad, G., Wu, S., Koh, I., Fu, C., Cooper, R.,
Betts, M., Kagioglou, M., Fischer, M. (2003). "Developing a Vision of
nD-Enabled Construction - Construction IT Report." University of
Salford, UK.
Lee, A., Wu, S., Aouad, G., Fu, C. (2004) "nD modelling in construction -
buzzword or reality?" In (eds) INCITE 2004 - International Conference
on Information Technology in Design and Construction, Langkawi,
Malaysia.
Lee, A., Wu, S., Marshall-Ponting, A.j., Aouad, G., Cooper, R., Tah, J.H.M.,
Abbott, C., Barrett, P.S. (2005). "nD Modelling Roadmap - A Vision
for nD-Enabled Construction." University of Salford, UK.
Mallasi, Z., and Dawood, N. (2002) "Registering Space Requirements of
Construction Operations Using Site-PECASO Model." In K. Agger,

65
SBUF project 11333 ITstomme - Final Report

Christiansson, P., Howard, R. (eds) CIB w78 conference 2002 -

Distributing Knowledge in Building, Aarhus School of Architecture,
Denmark, 1-8.
Navisworks, Ltd. (2005). "Navisworks Timeliner & Clash Detective."
www.navisworks.com.
Olofsson, T., Stehn, L., and Lagerqvist, O. (2004). "Industriellt byggande -
Byggbranschens nya patentlösning?" Väg- och Vattenbyggaren, 19-24.
Sjögren, J. (2005). "SBUF projekt no. 11598 - Förstudie produktion
tillämpning av 4D." Skanska Teknik AB, Malmö, Sweden.
Solibri, Inc. (2005). "Solibri Model Checker." www.solibri.com.
Strusoft. (2005). "Structural Design Software." www.strusoft.com.
Tekla. (2005). "Tekla Structures." www.tekla.com.

66
 Appendix A

Appendix A PROJECT MATRIX

Project Components

Drawings Multiple
Production Cost Material
Project Components & design
simulation estimation specifications
Documents models

Dimension 2D + 3D 3D + 3D 4D 5D nD

Expected Results

Specification of
method

Activities / phase:

1. prestudy
2. pilot
implementation
3. system
developments
4. full scale tests
5. reporting and
evaluation

Participants

67
 Appendix B

Appendix B 4D CASE STUDY

B.1 Introduction
This appendix to the ITstomme final report describes a case example that
includes 4D simulations of two construction alternatives. Both alternatives
illustrate construction methods that are typical to cast in place concrete
structures. The aim of the simulations is to provide a discussion basis for
comparison of two construction approaches.

Figure 10 provides an overview of the main layout of the building that is used
in the case study. The building measures 60 meters in length and is 12 meters
wide. All displayed slabs and walls are cast in place concrete components,
except for balconies that are prefabricated components.

69
SBUF project 11333 ITstomme - Final Report

Figure 10: Generic 3D CAD model of case example.

B.2 Set up
The 3D model is created by using AutoCAD Architectural Desktop (ADT) as
client software to an Internet based database system, developed by Enterprixe
Ltd. An object hierarchy is created for different CAD object groups by using
the database system.

Two alternative practices are modelled for the same building model in two
parallel object hierarchies.

- 0-Reference object hierarchy containing components related to the use

of traditional formwork. The objective of this scenario is to represent
typical sequenced and concurrent activities on a construction site that
are related to casting walls and slabs;
- An industrialized object hierarchy aimed at representing the potential
for permanent formwork systems in combination with the use of
prefabricated reinforcement mats and Self Compacting Concrete.

70
 Appendix B

Both scenarios contain the following main object groups, Figure 11:

- Building Model, including the geometry of building components. This

model is the same for both alternatives;
- Reinforcement, containing reinforcement bars for wall and slab
components;
- Formwork containing formwork elements for wall and slab
components;
- Casting sequences including sections of walls and slabs grouped as
individual casting sequences.

Figure 11: Main object hierarchy in AutoCAD ADT Enterprixe client.

Sub-object groups are created for the main object groups. An example of this
decomposition for formwork is included in Figure 11. Gantt bar charts, created
in MS Project, are imported to the project database and manually linked to the
respective objects and object-groups. Colour settings for the visualization of
activities are made after the linking of activities to the 3D CAD models is
completed.

In the next section the modelled construction processes are outlined in detail.

B.3 Alternative I: 0 Reference

To illustrate today’s common construction practice for cast in place concrete
structures, a 0 Reference scenario is used. This scenario is planned and
modelled by using existing and commonly used planning methods by
participants in the project team. The 3D CAD building model forms a basis for
planning casting sequences, formwork and reinforcement.

71
SBUF project 11333 ITstomme - Final Report

The objective of the visualization is to show typical sequenced and concurrent

activities on a construction site that are related to casting concrete walls and
slabs. In the following sections the planning and modelling processes are
described that form the basis for the 4D simulations.

B.3.1 Concrete casting sequences

Walls

Sequences for the casting process of walls are planned by using 2D drawings
and a set of coloured markers. Coloured lines are drawn over walls of the
structure; each line representing casting work for one day. The length and
distribution of lines are manually determined by a number of iterations in
which a set of decision criteria was used. Main decision criteria are:

- Required formwork;
- Volume of concrete;
- Work flow direction; and
- Work space planning.

Figure 12 shows the final plan for wall casting sequences, in which each
number represents a work day; in total eight days. These lines are the key for
sequencing casting work and determine formwork planning. Casting sequences
are represented by eight folders in the 3D CAD model hierarchy, under which
the individual wall CAD objects are grouped.

Figure 12: Eight casting sequences for concrete walls.

Slabs

Casting of slabs is planned by dividing slabs into sections of approximate equal

size. These sections are modelled as 3D CAD slab objects. The size of sections

72
 Appendix B

is determined by standard concrete delivery batches; i.e. m3 of concrete in a

concrete truck. All casting sections are grouped under two nodes in the 3D
CAD model hierarchy, representing casting work for two days work.

B.3.2 Formwork

Walls

Formwork for walls is modelled and represented by using standard 3D CAD

objects, such as walls, beams and columns. Formwork elements in the model
purely have a visual purpose and are abstract representations of actual
formwork elements. Brochure material from formwork suppliers is used to
select and model typical formwork. Form elements are grouped in the 3D CAD
model by sections of standard length.

Slabs

Formwork for slabs contains shoring and plywood plates. This formwork is
modelled in standard sections, partly according to product specifications from
formwork suppliers. An example of the representation of formwork for slabs is
included in Figure 13.

73
SBUF project 11333 ITstomme - Final Report

Plywood on shoring

Figure 13: 3D CAD representation of formwork for concrete slabs and

reinforcement bars in walls.

B.3.3 Reinforcement

Walls

Reinforcement bars in walls are modelled by using 3D CAD steel section

objects, Figure 13. Dimensions, locations, and number of reinforcement bars
are determined by visual analyses of the 3D CAD model. The reinforcement
bars serve a visual purpose and are not modelled for structural analyses.

Slabs

Reinforcement bars for slabs are modelled in one direction and distributed over
equal distances. The reason to model bars in one direction is to minimize 3D
modelling work. The purpose of these bars is to show the reinforcement
activity work zones rather than the actual reinforcement components.

74
 Appendix B

B.3.4 Planning

Construction operations for the 0 Reference alternative are planned by using

Gantt schemas, Figure 14. A work day for casting concrete walls and slabs is
typically planned in four periods of two hours.

II

III

IV
I

II

2H 2H 2H 2H

Figure 14: 0 Reference Gantt schema for casting concrete walls.

Work carried out during the four successive two-hour phases of a work day
includes the following activities:

I. During the first two hours formwork is installed on one side of walls
that are cast later during the day, Figure 15-I. Formwork for this
activity is taken down from the casting sequence of the previous day;
II. Reinforcement activities start after the first two hours. During this
phase the crew that worked on formwork during the first two hours,
continues the work on the formwork for the casting sequence of the
next day, Figure 15-II;
III. Side two of the formwork is installed during the third two-hour
sequence of the day, Figure 15-III. Reinforcement activities continue
during this phase; and
IV. During the fourth two-hour sequence of the day concrete is poured,
Figure 15-IV.

75
SBUF project 11333 ITstomme - Final Report

I: Install form I day n, remove n-1 II: Install reinforcement n, form I n+1

III: Install form II n, reinforcement n IV: Casting concrete walls

Figure 15: Four successive two-hour construction phases on day n, related to

casting of walls visualized in a 4D CAD model. Red components are in
activity, yellow components represent installed formwork, brown components
represent installed reinforcement bars and grey components represent finished
concrete walls and slabs.

In the next section the 4D modelling process for alternative II is outlined in

which an attempt is made to industrialize certain aspects of alternative I.

76
 Appendix B

B.4 Alternative II: Industrialized

The objective of the industrialized scenario is to visualize the potential of
innovative production technologies. Three production technologies are
combined in 4D simulations:

- Permanent formwork systems;

- Prefabricated reinforcement mats; and
- Self Compacting Concrete.

This scenario is modelled in 3D CAD by using similar approaches as used for

Alternative I. Because of the use of different production systems the planning
process has a different character. In the following sections the planning process
and representation of activities are outlined in further detail.

B.4.1 Formwork

As opposed to Alternative I, where casting sequences determine formwork

activities; the permanent formwork objects determined the planning of
Alternative II. A product specification is included in Figure 16.

1. Cast concrete

2. Permanent formwork:

- Wall element

Figure 16: Permanent formwork system.

2D drawings were sent to a supplier of permanent formwork systems, who

manually planned wall and slab form elements in the drawings, Figure 17.
Geometric data was manually extracted by the supplier from drawings and
used in spread sheets to calculate required resources and element costs. Hand

77
SBUF project 11333 ITstomme - Final Report

written element ID-numbers are the key in linking 2D drawings and spread
sheets together.

All formwork elements for walls and slabs are modelled in a 3D CAD model
that is organized according to element ID-numbers resulting from 2D
drawings.

Figure 17: Wall and slab formwork specifications in 2D paper based

environment.

B.4.2 Reinforcement

Walls

The permanent formwork system applied to the building used in the case study
does not require installation of reinforcement bars in walls on site.
Reinforcement bars for walls are therefore not modelled in 3D CAD.

Slabs

Part of the reinforcement needed for slabs is pre-installed on the applied

permanent formwork system, Figure 16. These reinforcement bars are not
modelled in a 3D CAD model. A prefabricated reinforcement system is chosen
for additional reinforcement bars that are needed, Figure 17. This
reinforcement comes to site on prefabricated mats, rolled out on site. For the
3D CAD model of this reinforcement system a similar approach is adapted as

78
 Appendix B

used for Alternative I. The purpose of these bars is to show the reinforcement
activity work zones rather than the actual reinforcement components.

Figure 18: (Left) 3D CAD systems can provide input for reinforcement
prefabrication. (Right) Prefabricated reinforcement mats rolled out, decreasing
activities on site compared to traditional reinforcement.

B.4.3 Concrete casting sequences

By using the permanent formwork system there is no clear distinction needed

between casting concrete walls and slabs. Walls and objects are cast in
successive sequences. The 4D model objective is to show the flow of Self
Compacting Concrete, Figure 19. Casting sequence objects as used in
Alternative I are split in smaller 3D CAD objects (Figure 19) to better
represent the flow of concrete in 4D CAD simulations.

79
SBUF project 11333 ITstomme - Final Report

Figure 19: (Left) Casting process of Self Compacting Concrete (SCC)

decreases activities on site. Less labour is needed during casting processes
compared to traditional concrete. (Right) 4D CAD model representation of
SCC-flow (arrows not included in 4D CAD model and later added to clarify
image). Red components are in activity, brown components represent installed
reinforcement bars and grey components represent permanent formwork,
concrete walls and slabs.

B.4.4 Planning

Construction operations are planned by using Gantt schemas. All formwork

elements, including activity length are imported from spread sheets to Gantt
schemas. Installation sequencing is done in Gantt schemas by linking activities,
i.e. elements. Main scheduling criteria are: work flow direction and work space
conflicts.

80
 Appendix B

Figure 20: Gantt schema of Alternative II. Activity name and duration are
imported from spread sheets. The order of activities is determined by relations
in the Gantt schema.

B.5 4D results
Both construction alternatives can now be simulated in 4D CAD. The two
processes show to some extent similar work flows. Activities for formwork,
reinforcement, and concrete are carried out concurrently, enabling and
constraining the execution of other activities. The main differences between
both alternatives are the dependencies between work flows and the production
rate.

In Alternative I these flows are interwoven in a four times two-hour cycle,

Figure 15. Delay in one of these flows has almost directly impact on other
flows, making at it a vulnerable process. For example, a delay in installation of
formwork for walls constrains installation of reinforcement and casting of
concrete. One construction day is lost when concrete cannot be cast at the end
of the day. Construction operations for alternative II are determined by the
installation of form elements that do not follow a strict cycle. A delay in the
installation of one element does often not have a direct impact on activity flows
for reinforcement and concreting.

81
SBUF project 11333 ITstomme - Final Report

Simulation of the various work flows in 4D CAD provides users with a visual
impression of the differences between construction alternatives and their work
flows. However, these visualizations do not directly provide users with data to
compare alternatives. As an example the 4D comparison included in Figure 21
can be given. What is compared in the two alternatives? What performance
does one compare? Work spaces, resources, ect.? These observations and
questions serve as a point of departure for research carried out at CIFE in 2004,
which is reported in (Jongeling 2005b; Jongeling 2004a).

Figure 21: Comparison of construction alternatives in a 4D CAD environment.

What is compared? (Left) Alternative I: 0 Reference process. (Right)
Alternative II: Industrialized approach. Red components are in activity, yellow
components represent installed traditional formwork, brown components
represent installed reinforcement bars and grey components represent
permanent formwork, finished concrete walls and concrete slabs.

82
 Appendix C

Appendix C 5D COST ESTIMATION

C.1 Introduction
This appendix to the ITstomme report describes the background and
developments regarding the 5D cost estimation process.

Today’s modern object oriented 3D CAD systems can in most cases calculate
and make reports of quantity of material needed in a building project, so-called
BoM, ‘Bills Of Material’. This is because the CAD objects are parametric and
built up by different property values that generate the graphics like ‘width’,
‘height’, ‘length’, etc. The CAD objects can also have non-graphical properties
like ‘material’, ‘classification code’ etc.

The current cost estimation process, are still based on ‘non-intelligent’ 2D

drawings. All quantity take-offs are still calculated manually, by measuring the
drawings. It is not a big difference if the drawing is created with a CAD
system, or with a pen. This process is very time consuming and has to redone,
each time a revision of the project is made.

The goal of the 5D developments in the ITstomme project is to automate the

process of calculating bills of material, and also sort it in a way, so that it better
suits the needs for the cost calculating process, and cost calculating software
tools. The proposed method is to let the cost calculator organize the product
model in a ‘cost hierarchy’, by grouping all CAD objects of same type, to a
‘5D cost object’. These objects are then mapped to a cost ‘recipe’ and
transferred to the cost estimating system.

83
SBUF project 11333 ITstomme - Final Report

C.2 Current praxis

Cost estimation is an important and time consuming process. It is done many
times during a project lifecycle. At JM, three types of cost calculations are
made in a typical project:

1. Pre cost estimate, which is based on experience ‘key values’

(Förhandskalkyl)
2. Base cost estimate, which is based on early stage drawings
(Grundkalkyl)
3. Production cost estimate, which is based on building drawings.

In this project we will include number 2 and 3.

C.2.1 Quantity take-off process

To be able to make a cost estimate, the amount of needed material (the volume
of different building parts), needs to be calculated. The common praxis is to
measure from drawings and store this data in a spreadsheet. Quantity take-offs
are typically measured in units like perimeter (m), area (m2) or volume (m3).

Different quantity items are calculated with different accuracies. In some cases
more generic (non exact) quantities are calculated such as ‘facade area’, but in
other cases more exact quantities are used e.g. ‘number of columns’. Some
quantities are calculated from drawings produced by the architect and some
from the structural engineer’s drawings.

When JM makes quantity take-offs, they are typically grouped by the location
in the building, or by the different contract boundaries. The quantity take-off
can also be sorted by phases in a project. In a cost calculating system this is
called (in Swedish) an ‘Upost’. The quantity calculation is either made by
specialised ‘quantity take-off’ companies such as (in Sweden), ‘Mängda’ and
‘Bygganalys’, or by the contractor / project management themselves. Building
parts of same type is summarized, as one quantity item, e.g. all inner walls at a
certain location. One big drawback in current process is that the whole
calculation has to be redone, every time the drawings are changed.

To define and transfer the semantics of the information, different classification

systems are developed e.g. (in Sweden) SBEF, and BSAB. The SBEF system is
quite old, and sorts the building parts in a logical system with codes from 0 to
99, where 0 means ground works and 99 means roof. The BSAB system is

84
 Appendix C

newer and more comprehensive, and will substitute the SBEF system in due
time.

C.2.2 Transfer quantities to cost estimation systems

To transfer the bills of quantities, to cost estimating software tools there are
standardised file formats available. The two most common are the SBEF file
and the newer sbXML file.

The cost estimation process at JM is currently based on the SBEF system and
for this reason this system is applied within the ITstomme project. However,
the developed concepts in the project are fully applicable to the BSAB system
and can be developed to support the sbXML file format.

C.2.3 Cost calculation process

When the quantities are imported to a cost calculating system, it has to be

mapped with a ‘recipe’ code, which is the way how a cost calculating system
works: Each building part, e.g. a wall, can be constructed in many ways - by
in-situ cast concrete, by wood and panels, by bricks etc. Each type of
construction is called a ‘building part type’. When later the building part types
are constructed, these types are used as ‘activities’ or ‘methods’ such as
building the form, pouring the concrete etc, and ‘resources’ such as concrete,
reinforcement, labour etc. The result of these activities and resources is called a
‘production result’. The geographical position of a building part inside a
building is called a ‘location’.

A recipe in the cost calculating system, knows the needed material and labour
(resources) and also activities (methods) needed to build one unit of a building
part type (typically 1 square meter). Using this principle it is not needed to
manually calculate each nail in a building. The recipe knows how many there
are needed to construct 1 m2. This process is called Activity-Based Costing
(ABC).

C.3 Proposed method

In product models from different design disciplines it is important to define
which CAD objects (from a specific application and discipline) are the
‘information carriers’ to generate quantities for different building parts. This is
due to the fact that there might be some overlapping between the design
disciplines. In this project we made a study of the division between architects’
CAD objects, and structural engineers’ CAD objects.

85
SBUF project 11333 ITstomme - Final Report

Product models are typically sorted in a hierarchy that reflects the logical
location of a building part within the building. This can also be expressed as a
path in the model e.g.
BUILDING/STOREYS/GROUNDFLOOR/LOAD BEARING
STRUCTURES/COLUMNS/COLUMNOBJECT

The traditional view of a product model is a reflection of the designers’ view,

resulting in a kind of logical sorted model, based on building parts or material.
This is very obvious and explainable, if we analyse how drawing layers in the
past have been defined. The common (in Sweden) ‘Point’ layer standard
consists of a mixture of building parts (e.g. columns, beams, windows) and
materials (concrete, steel, wood).

We have found out that the cost calculator’s view differs completely from the
designer’s view. As mentioned earlier, their view is based on phases, contracts,
or construction methods. Therefore there is a need for more than one static
hierarchy. Actually, each actor in the building process might need to organize
the data differently.

During 2002, a similar integration project between product models and

estimating tools was performed (Edgar 2002). One of the conclusions of that
project was that each separate CAD object cannot generate a separate cost
item. There will be too many items and this is not the way a cost calculator
works. Instead the CAD objects must be grouped by the same type; exactly in
the same way as the quantity take-offs of today are grouped by building part
types. As a first practical step, we propose that the cost calculator ‘maps’
(groups) the CAD model to different building part types. We call each such
group a ‘5D object’, and the quantity is the derived sum of all CAD objects that
are mapped to it. After this the 5D object must be classified. To do so the cost
calculator must define which classification code to use (or in other words: the
recipe code that the cost estimating system uses).

The optimal case would be product models delivered by consultants that

already have the recipe codes defined. There are a few pioneering companies
that provide their consultants with standardized libraries that have to be used
when the consultants created their product models. However this is not the
main practise in the building sector today. Theoretically this is of course the
optimal solution, but the building process and most of today’s companies are
not organised in this way today. To succeed with this method, the designers
must be much more integrated and steered by their clients, such as JM.

86
 Appendix C

To sum up: to be able to automate the quantity take-off process, and cost
calculation process, the CAD objects needs to be :

- Sorted in a 5D cost hierarchy

- Grouped as 5D Cost Items
- Marked with a recipe code

Either this task is done by the cost calculator, or preferably by the designers,
during the design process.

C.4 Implemented method

A product model is typically sorted in one hierarchy. The hierarchy reflects the
different locations in a building (based on the designers view). This makes it
easy to generate bills of quantities for a certain location. Queries can be asked
to the database, such as ‘how many concrete walls of 160mm width exist on
floor 2?’

In the hierarchy of a product model, objects are also sorted by building parts on
a lower level in the hierarchy. These are called classes and typically define the
characteristics or type of objects, e.g. columns, beams, walls etc. The classes
can be compared with the way a 2D drawing is sorted by layer names. The
class could in the future be marked with a classification code such as the (in
Sweden) BSAB standard.

A model hierarchy can consist of many subclasses and at the lowest level in the
hierarchy the CAD objects are stored. Each object is owned by another object.
This is typically called a ‘master-slave’ relation. If a user makes a copy of floor
1 as floor 2 then all sublevels (called nodes) are also copied and e.g. moved.

C.4.1 Mapping and grouping to 5D objects

In the receiving cost calculating system MAP, the hierarchy (the locations), are
expressed in maximum 5 levels. The locations are defined by a code in the
format ‘1.1.1.1.1’. Instead of classes, MAP has codes such as the SBEF system
mentioned above. E.g. ‘63’ means ‘Inner walls’. To be able to reflect the
product model from a cost calculator’s view, a new 5D hierarchy must be
created. In this hierarchy we could add classes, so that they reflect the SBEF
codes, such as 63_Innerwalls.

87
SBUF project 11333 ITstomme - Final Report

All standard recipes that JM uses must be stored in a library that accessible by
all designers over Internet. For this purpose an XML file is developed and
placed at a server. In the model viewer software developed in the ITstomme
project, a new 5D user interface has been developed where the cost calculator
can create a 5D hierarchy, and can perform the grouping and mapping work.
The CAD objects that belong (that is owned) by a 5D node, is linked to the 5D
hierarchy.

The final work order for the cost calculator is:

- Create a suitable 5D hierarchy, reflecting the calculation needs (e.g.

based on a phase or contract)
- Choose a specific type of building part (e.g. walls). This will limit the
amount of available SBEF codes, meaning choosing the right class (e.g.
63_Innerwalls)
- Choose a recipe code, from the available recipe list for a class.
- Select from the screen, or by the tree view, the CAD objects that should
be mapped with the 5D Object.
- Select ‘Link’, and the volume or another parameter of selected objects
is summarized, and 5D-links are created to the objects.

88
 Appendix C

Figure 22: Screen capture from the 5D cost estimation process by using a
model viewer developed in the ITstomme project. 3D objects are selected by
an estimator in the 3D scene (left) or via the PE (Project Explorer) (right) after
which a cost estimation post is created in a so-called 5D hierarchy (on the left
of the PE). A cost estimation file is created from the 5D hierarchy and used in
cost estimation software. The 5D and Property window are enlarged in Figure
23.

89
SBUF project 11333 ITstomme - Final Report

Figure 23: (Left) The 5D hierarchy

is created by the cost estimator,
representing and linking his/her
view to the product models in the
central database. (Right) Every 5D
item (in this case the post ‘bärande
väggar’) has properties, such as
derived and summed parameters
from CAD components (Area,
Volume) and a Recipe obtained
from a recipe database.

90
 Appendix C

C.4.2 Reports of quantities and 5D cost items

A special report definition to generate so-called 5D reports has been developed

in the ITstomme project. A report can be generated via the Internet, by using a
web portal to access the central database, Figure 24. From the reporting tool a
XML file is exported and converted to a SBEF file by a program developed in
this project, Figure 25.

Figure 24: An example of a 5D report, which is saved as a XML file and

converted to a SBEF file that can be imported to the MAP cost estimation
system in use at JM

91
SBUF project 11333 ITstomme - Final Report

Figure 25: Example of a SBEF file that can be imported to by the cost
estimator

C.4.3 Importing the quantities to MAP

MAP has an import function to read SBEF files. When the file is imported, all
cost items have full the intelligence. Different cost calculations can be made
for different ‘Cost locations’ in the building.

MAP will calculate the cost, by using the attached recipe and quantity
information. The whole process to quantify, to sort and to create 5D items in
MAP is now much more automated and efficient compared to current praxis.

92