INTEROPERABILITY BETWEEN GIS AND BIM

A Semantic-based Multi-representation Approach

Clément Mignard

, Gilles Gesquière

and Christophe Nicolle

Active3D, 2 Rue René Char, BP 66 606 21066 Dijon Cedex, France

LSIS - UMR CNRS 6168, IUT de l'Université de Provence, 13200 ARLES Aix-Marseille, France

LE2I – UMR CNRS 5158, IUT Dijon-Auxerre, Université de Bourgogne, BP 47870, 21078 Dijon Cedex, France

Keywords: Ontology, Knowledge Acquisition, Geographical Information System (GIS), Building Information Systems

(BIM), Industry Foundation Classes (IFC).

Abstract: Interoperability of information systems is partially resolved due to many standards such as networks

protocols, XML derived languages and object oriented programming. Nevertheless, semantic heterogeneity

limits collaborative works and interoperability. Despite ontology and other semantic technics, the binding of

heterogeneous information systems requires new technics of managing and displaying information

according to the semantic representation of each stakeholder of the collaboration. In this paper we addressed

the problem of merging geographical information systems and building information model. The way to

achieve this goal must solve several heterogeneity problems due to the data life cycle, the data temporality,

the binding between 2D geo-referenced modelling and 3D geometric models or problem of scalability for

real-time 3D display from remote server for managing a real environment of several million m2. To bridge

this gap, we present a new architecture based on a semantic multi-representation of heterogeneous

information.

1 INTRODUCTION

Today, at a time when environmental issues are

becoming more insistent, ways to control costs in the

management and development of a territory are

increasingly sought. This may involve the facility

management of a set of building block, that one

wishes to identify and observe to limit the costs of

maintenance or the creation of new entities in order

to anticipate the ecological impacts and economic,

and at different levels. These goals require to have a

lot of heterogeneous information on assets to

manage, at several moments of their life cycle and at

different levels. This unification is an expensive

process which is not always adapted to the trends of

the trade or the market. The global information

system becomes quickly obsolete and unsuited

regarding the data model evolutions and

improvements. In order to unify and centralize the

management of real estate, urban and extra urban, it

is necessary to develop a new form of collaborative

architecture (Döllner et al., 2007). This architecture

will allow to combine in a homogeneous

environment a set of heterogeneous information

from diverse information systems such as those from

the BIM domain and the GIS domain.

The term BIM (Building Information Modeling) has

been coined recently to demarcate the next

generation of Information Technologies (IT) and

Computer-Aided Design (CAD) for buildings which

focus on drawing production. BIM is the process of

generating, storing, managing, exchanging and

sharing building information in an interoperable and

reusable way (Vanlande et al., 2008). A BIM system

is a tool that enables users to integrate and reuse

building information and domain knowledge

throughout the building lifecycle (Campbell, 2007).

The Geographic Information Systems (GIS) are

becoming a part of mainstream business and

management operations around the world in

organizations, both in public and private sectors, as

diverse as cities, state government, civil engineering,

telecommunications, urban planning, petroleum

exploration, land surveying, etc... The term GIS

refers to any system that captures, stores, analyzes,

manages, and presents data that are linked to at least

one location. BIM and GIS need to be coupled in a

common environment in an interoperable way.

359

Gesquière G., Mignard C. and Nicolle C..

INTEROPERABILITY BETWEEN GIS AND BIM - A Semantic-based Multi-representation Approach.

DOI: 10.5220/0003664403590362

In Proceedings of the International Conference on Knowledge Management and Information Sharing (KMIS-2011), pages 359-362

ISBN: 978-989-8425-81-2

 2011 SCITEPRESS (Science and Technology Publications, Lda.)

Since 2008, as part of a European project, we

develop a collaborative web platform dedicated to

urban facility management. Our Urban Information

Model (UIM) approach combines both BIM and GIS

using semantic modelling to access global

knowledge of a complete urban environment,

including sets of buildings and urban objects that

compose this environment. This approach is based

on a semantic architecture using ontology evolution

mechanisms (Gruber, 1993). We have developed a

specific 3D-viewer making possible semantic

management of Level of Details (LoDs) according to

user profile and context. The multi-representation

introduced in our architecture adds to the traditional

LoD the notion of Contextual LoDs (C-LoDs). A C-

LoD is not only displayed depending on the distance

between the view point and the object as it is usually

the case. The representation is chosen according to

other criteria that depend on user (like the business

process to which he is attached), external criteria as

day/night or weather, or even of the object itself.

The semantic management drives streaming

processes which extract the semantic and 3D

representations of urban objects from a relational

database.

2 SIGA3D OVERVIEW

Our proposal is based on a semantic architecture

articulated in 6 levels (Figure 1). The import/export

level is dedicated to the parsing of various file

formats required to model the UIM from different

sources (GIS/BIM). The data model level makes it

possible the combination of geometrical data and

semantics. The level "contextual view" associates

user profiles and business rules to build contextual

LoDs. The connection level is mainly dedicated to

the streaming process between the databases and the

interface. The interface level displays the urban

environment into a 3D digital mock-up coupled with

a semantic tree of urban elements.

The innovative part of this architecture is mainly

contains in the data model level and contextual

views level. This part is the base of our semantic

LoD proposal. The first data model level part is

architecture of graphs representing the ontology,

allowing the context management and versioning of

the data (CMF for Contextual Model Framework).

Graphs operators are also defining to facilitate the

implementation of changes in conceptualization.

Figure 1: SIGA3D Architecture.

Information about reference systems for space

and time (CRS and TimeZone) are also managed in

this part. The other part defines a unified syntax-

based knowledge representation based on the

languages OWL, RDFS, and rules RuleML, SWRL

and described in this document in an expressive way

with description logic. It is called DMF (Data Model

Framework). DMF also contains operators for the

management of space and time and the definition of

local contexts that allow us to conduct a multi-

representation of data. The goal of this part is to

provide models used in inference engine to infer and

to check the data modelled by the CDMF modelling

operators.

3 DATA MODEL FRAMEWORK

The Data Model Framework is made of operators to

construct urban data models. These operators allow

the description of classes and properties that can be

used to define complex concepts using operators of

intersection, union, involvement, etc ...

The spatial data and especially georeferenced

coordinates do not make sense without the

knowledge of the coordinate reference system. This

information will appear in the next layer of our

architecture that manages the context of model

graph, to unify the management of coordinates. The

same kind of information is provided for time, with

the management of Time zones (Gutierrez et al.,

2007).

The management of local contexts, which allows

multi-representation, is done in this part by defining

new stamped operators, corresponding to the dmf

operators defined above. For example, script 1

defines three local contexts, designer,

structureEngineer and achievementDate. Script 2

3DEngine

DataLayer

Semantic LayerGeometric Layer

Semantic Viewer

SIGA3DInterface

SIGA3DConnexion

Streaming

SIGA3D

Contextual Views

View ManagementTool

ProfilManagementTool

SIGA3D Data

ModelFramework

SIGA3D

Import/Export

IFCDWG/DXF/DGN/GML

KMIS 2011 - International Conference on Knowledge Management and Information Sharing

360

defines several properties and a spatial

representation for a class ‘buildingPlan’ which

depends of the user. The contextual operators

dmf:[c

, …,c

]Class, dmf:[c

, …,c

]property and

dmf:[ c

, …,c

] spatialEntity are used.

<dmf:Class rdf:ID=’Profession’/>

<dmf:temporalEntity rdf:ID=’Day’/>

<dmf:property rdf:ID=’unitType’/>

<Day rdf:ID=’March’><unitType

rdf:resource=’#unitMonth/></Day>

Script 1: Definition of three local contexts.

<dmf:Class rdf:ID=’BuildingPlan’/>

<dmf:[designer]property

rdf:ID=’line_thick’/>

<dmf:[structureEngineer]property

rdf:ID=’wall_material’/>

<dmf:[designer]property

rdf:ID=’contains_plan’/>

<dmf:[designer,structureEngineer]property

rdf:ID=’contains_plan’/>

<dmf:spatialEntity rdf:ID=’the_plan’/>

<dmf:[designer]property

rdf:ID=’3D_plan’/>

<dmf:[designer,structureEngineer]property

rdf:ID=’2D_plan’/>

<the_plan rdf:ID=’plan_of_building_1’>

<url_2D_plan

rdf:resource=’/building/1/plan/plan2D.dwg’/>

<url_3D_plan

rdf:resource=’/building/1/plan/plan3D.ifc’/>

</the_plan>

<dmf:[designer,March]Class

rdf:ID=’Plan_availability’/>

<line_thick

rdf:dataType=’&xsd;float’>10</line_thick>

<wall_material

rdf:dataType=’&xsd;float’>wood</wall_materia

<contains_plan

rdf:resource=’the_plan’/>

</BuildingPlan>

Script 2: Example of contextual operators.

This example describes an object, BuildingPlan,

which has several properties. For a designer, the

BuildingPlan is defined with a line_thick, a plan

containing two representations. The same object is

defined differently for a structure engineer, with the

material of walls, wall_meterial, and an attached

plan with only one 2D representation.

4 CONTEXT MODEL

FRAMEWORK

This part of our architecture is composed of three

main blocks. The first block sets the context for each

graph of DMF, the second block defines a set of

graph operators to facilitate the writing and limit the

redundancy of data in the context management and

the third block defines a set of operators on graphs

to describe more accurately the geographical

information by defining relations between the

spatio-temporal data models of DMF. Context

management in this architecture is done by defining

a special graph called SystemGraph. A

SystemGraph is a graph or a set of graphs using

operators. These operators are graphs of the second

block of the CMF. The use of these operators can

simplify the management of the evolution of

knowledge of the model. So, rather than storing for

each modification of the model a new version of the

information, the CMF layer store the modification as

operations on graphs. The SystemGraph can be

describing using the following operators:

cdmf:graph connects graph and data. These data are

described according to the data model. They can be

a combination between other graphs using the CMF

graph operators AddGraph (union of graphs),

RemoveGraph, InterGraph, CompInterGraph and

MapGraph

cdmf:of represents the context. This property

defines a list of resources representing the access

context.

cdmf:model defines for a system graph the data

model which is used. This data model defines

elements which will appear in the graph.

cdmf:action defines user’s rights to access the data

(read/write/remove). If no action is defined in the

system Graph, which means that only the

visualization of the data is allowed.

cdmf :synchronizationGraph defines a list of Graph

depending of a special model where we define all

kind of spatial and temporal relationship between

data models.

cdmf:reference_frame defines the timezone and the

Coordinates Reference System (CRS) used for the

data model associated to the SystemGraph. These

values are valid for all data of the associated graph,

even if data sources are defined in another reference

system (in which case it is needed to make

transformations during the displaying phase of data).

The spatio-temporal synchronization is not a

common graph operator and is very specific to the

description of geographical information. It allows

INTEROPERABILITY BETWEEN GIS AND BIM - A Semantic-based Multi-representation Approach

361

defining the validity of a model by describing

relationships with other models. It can be used in

case of model evolution to assure the consistency of

the global model. For example, if we define a

building model and an electric power network

model, it is possible to describe a topological

relation between the two models to say they are

spatially connected. Then, when one of the models is

modified, for example to reposition the building

because of a bad georeferencing, we know we have

to modify the other model to keep the spatial

connection relation consistent.

5 CONCLUSIONS

This paper presents an ongoing research on the

definition of an Urban Modelling Architecture. This

paper focus on a new mechanism of LoDs called

contextual LoDs. It is the merge of classical

geometric approach to define LoDs and two

semantic multi-representations formalisms: the first

part is based on contextual trees to define user

profiles and business rules at the data model level.

The second part defines local contexts to allow

multi-representation at a lower level, i.e. for each

objects of the model. The concept of contextual

LoDs is designed to be integrated in an Urban

Facilities Management (UFM) platform. It is an

extension of the BIM concept for the management of

urban objects. Our framework facilitates data

maintenance (data migration, model evolution)

during the lifecycle of an urban environment and

reduces the volume of data with specific graph

operators. The urban approach also implies to

manage precisely the spatial and temporal

dimensions that have been considered in the

definition of the contextual LoDs part. This

approach is based on the CityGML 1.0 (Kolbe et al.,

2009) and IFC 2x3 standards.

In the figure 2, a result of the integration of IFC

and GIS data (the IFC building is in red, and a

couple of buildings coming from GML file are in

blue) into the urban ontology can be seen.

Our future works will be to achieve the

implementation of our framework for the UFM

platform, including the contextual LoDs

management. These works are based on our previous

works on Active3d (Vanlande et al., 2008) and

designed to be fully compatible with both standards:

the one for geographic information (e.g. ISO/TC

211) and the second for the construction world (e.g.

ISO/PAS 16739).

Figure 2: Example of a 3D georeferenced scene with

multiple data sources.

REFERENCES

Campbell D. A., (2007), Building information modeling:

the Web3D application for AEC, Published in

Proceeding Web3D '07 Proceedings of the twelfth

international conference on 3D web technology, ACM

New York, NY

Döllner J., Hagedorn B (2007), Integrating Urban GIS,

CAD, and BIM Data By Service-Based Virtual 3D

City Models. 26th Urban Data Management Symp.,

Stuttgart, Germany

Gruber, T., R. (1993), A translation approach to portable

ontologies- Knowledge Acquis., 1993., pp 1-27

Gutierrez, C., Hurtado, C. A., and Vaisman A. (2007),

Introducing Time into RDF. IEEE Transactions on

Knowledge and Data Engineering, 19(2):207-218

Kolbe, T. H. (2009) Representing and Exchanging 3D

City Models with CityGML, Lee, Jiyeong / Zlatanova,

Sisi (Eds.), Proceedings of the 3rd Int.Workshop on

3D Geo-Information, Seoul, Korea. Lecture Notes in

Geoinformation & Cartography, Springer Verlag,

2009

Vanlande, R., Cruz C., Nicolle C. (2008) IFC and

Buildings Lifecycle Management", Journal of

Automation in Construction, Volume 18, Issue 1,

Elsevier, 2008, pp 70-78

KMIS 2011 - International Conference on Knowledge Management and Information Sharing

362