Semantic enrichment of Building and City Information Models: a ten-year review
Fan Xue, Liupengfei Wu*, and Weisheng Lu
Department of Real Estate and Construction Management, The University of Hong Kong, Hong Kong, China

This is the peer-reviewed post-print version of the paper:
Xue, F., Wu, L., & Lu, W. (2021). Semantic enrichment of Building and City
Information Models: a ten-year review. Advanced Engineering Informatics, 47,
101245. Doi: 10.1016/j.aei.2020.101245
The final version of this paper is available at https://doi.org/10.1016/j.aei.2020.101245.
The use of this file must follow the Creative Commons Attribution Non-Commercial No
Derivatives License, as required by Elsevier’s policy.

Abstract:
Building Information Models (BIMs) and City Information Models (CIMs) have flourished in
building and urban studies independently over the past decade. Semantic enrichment is an
indispensable process that adds new semantics such as geometric, non-geometric, and
topological information into existing BIMs or CIMs to enable multidisciplinary applications in
fields such as construction management, geoinformatics, and urban planning. These two paths
are now coming to a juncture for integration and juxtaposition. However, a critical review of
the semantic enrichment of BIM and CIM is missing in the literature. This research aims to
probe into semantic enrichment by comparing its similarities and differences between BIM and
CIM over a ten-year time span. The research methods include establishing a uniform
conceptual model, and sourcing and analyzing 44 pertinent cases in the literature. The findings
plot the terminologies, methods, scopes, and trends for the semantic enrichment approaches in
the two domains. With the increasing availability of data sources, algorithms, and computing
power, they cross the border to enter each other’s domain. Future research will likely gain new
momentums from the demands of value-added applications, development of remote sensing
devices, intelligent data processing algorithms, interoperability between BIM and CIM
software platforms, and emerging technologies such as big data analytics.
Keywords: Building Information Model; City Information Model; Geographic Information
System; Semantic enrichment; Smart city; Intelligent building
1
2
3
4
5
6

1. Background
In the architecture, engineering, and construction (AEC) industry, information is the common
keyword of the Building Information Model (BIM) and City Information Model (CIM). The
concept of BIM is coined to refer to a digital representation of a facility’s physical and
functional characteristics; it is a shared knowledge resource for information about a facility,
forming a reliable information foundation for decisions throughout its life cycle (NIBS, 2015).
1

7
8
9
10
11
12

Building information modeling is a nomenclatural term used to refer to a family of technologies
and related practices used to represent and manage information used and created for the process
of designing, constructing, and operating buildings (Davies and Harty, 2011). BIM has roots
that can be traced back to the production and manufacturing industry, where designers tend to
develop a digital model of the product for optimization and prototyping before it is massproduced (Eastman et al., 2011).

13
14
15
16
17
18
19
20
21
22
23

The concept of CIM has emerged in parallel to refer to a system of urban elements and
environments represented in 2D and 3D symbols (Stojanovski, 2013). The gerund, city
information modeling, is used to represent the technologies and practices used to develop a city
information model and harness its power for various smart applications, e.g., navigations,
transportation, urban climate, and urban morphology (Xu et al., 2014), which are normally
placed under the umbrella of a smart city. 3D city models as the baseline of CIM has a much
longer history if one considers its roots in Geographic Information System (GIS), a computer
system for capturing, storing, checking, and displaying data related to positions on the Earth’s
surface (NGS, 2012). According to Julin et al. (2018), such 3D city models are typically built
by merging photogrammetry and laser scanning data with GIS data.

24
25
26
27
28
29
30
31
32
33
34
35
36

The essential difference between CIM and GIS is the scope and form of the “I,” i.e., information.
The information in GIS is explicitly referenced according to the scope of the Earth’s surface
and often managed in forms of ‘layers’ (Goodchild, 1991; Lu et al., 2018). In contrast, a CIM’s
scope focuses on urban areas but includes more types of non-GIS information, e.g., BIM, Light
Detection and Ranging (LiDAR), inhabitant behaviors, and city energy simulation, and
managed in the form of cross-reference relation ‘networks’—like a BIM (Xu et al., 2014; Liu
et al., 2017). Thus, in this paper, at the risk of oversimplification, a 3D city model in GIS is
treated as a specific type of CIM. In recent years, other buzzwords such as the ‘Cyber-Physical
System’ and ‘digital twin’ are emerging (Xue et al., 2020) from agriculture, defense, energy,
healthcare, transportation, and manufacturing systems to represent the virtual, digital models
of these systems. In non-technical language, one can perceive the ‘real-time’ BIM and CIM as
the cyber replica or digital twin of a building and a city, respectively.

37
38
39
40
41
42
43
44
45
46

Characteristics of BIM and CIM information are essential for differentiating and understanding
these buzzwords. Information is the shared key to enable their designated applications, though
the scope and organization are different. As listed in Table 1, BIM represents building elements
(e.g., beams and columns) with building-level information (e.g., geometric details, energy
usage data, and life cycle costing information) to meet AEC industrial needs (Karan et al.,
2016). The taxonomy of BIM information distinguishes geometric, semantic, and topological
types (Pratt-Hartmann, 2004; Schlueter and Thesseling, 2009). Geometric information directly
relates to the shapes and forms of facilities, whereas semantic information captures their
intrinsic properties (e.g., functionality), and topological information gives the relationships
2

47
48
49

among these objects. In comparison, CIM involves a three-dimensional model that connects
buildings and other urban information sources (e.g., roads, public spaces, street lights, and even
people on the street) in a city (Xu et al., 2014).

50
51

Table 1. Scope and organization of information in BIM, and CIM

BIM

Dimension of
geometry
3D

CIM

3D

Scope of information
Up to building level (e.g., rooms, buildings, and
bridges)
Up to city level (e.g., metro shop, park bench, city
hall, street greenery, and landscape)

Organization of relational
information
Explicit via relations between
3D components (e.g., joints)
Geo-referenced 3D City
elements

52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76

The term ‘semantic enrichment’ has been defined in several ways. For example, Clarke and
Harley (2014) defined semantic enrichment as the process of adding a layer of topic metadata
to content so that the machine can understand and connect with it. Later, Sacks et al. (2017)
pointed out that semantic enrichment aims to add absent or new information, compile as-is or
as-built conditions, and extend the schemas to existing BIMs/CIMs. Bloch and Sacks (2018)
further define semantic enrichment as an indispensable process that encompasses the
classification of objects, aggregation and grouping, unique identification, completion of
missing objects, and reconstruction of occluded objects in the case of application to models.
This study adopts Bloch and Sacks’ (2018) definition given its comprehensiveness. The
exterior geometry of a building or a city might be fixed from the outset; semantic information,
however, will update along the way when they are developed, utilized, and updated. These
updated semantics, e.g., as-built or as-is conditions, should be properly captured in BIM or
CIM. Over the past decade, semantics enrichment has been developed in BIM and CIM
independently, and it remains so. Nevertheless, scholars are increasingly interested in seeing
more details of buildings in a CIM. Many review papers have reflected the up-to-date
processing of building and city data to create new BIMs/CIMs, e.g., for buildings (Volk et al.,
2014), indoor environments (Kang et al., 2020), roads (Wang et al., 2016), and applications
such as building energy (Reinhart and Davila, 2016) and urban ecology (Berling-Wolff and
Wu, 2004). However, a study to juxtapose the research niche of enriching existing models is
absent and thus much desired. This study is particularly opportune because BIM and CIM,
previously dominated by giant software vendors, are starting to join hands with each other (Esri,
2019). Therefore, it would be intriguing to review what has been researched in BIM and CIM
semantic enrichment and to extrapolate what will likely happen under the general trend of BIM
and CIM integration.

77
78
79
80
81

The primary aim of this research is to compare the semantic enrichment studies in BIM and
CIM in the past 10 years for outlining the research trends and prospects. The goal is achieved
by deploying an analytical framework that includes a conceptual model, a set of comparative
studies, and future direction predictions. The rest of this article consists of four sections.
3

82
83
84

Section 2 describes the research methodology. Section 3 compares the semantic enrichment of
BIM and CIM. Section 4 gives the trends and prospects of semantic enrichment of BIM and
CIM. Findings are drawn in Section 5.

85
86
87
88
89
90

2. Research methods
The primary research method adopted in this paper is a comparative archival study. It follows
a four-step methodology to conduct the study: (1) Developing a guiding conceptual model, (2)
Data collection by published case selection and decoding, (3) Comparative study of semantics
enrichment for BIM and CIM, and (4) Extrapolation of the future development.

91
92
93
94
95
96
97
98

2.1 Guiding conceptual model
A guiding conceptual model was developed to facilitate data collection and analysis. As shown
in Figure 1, the conceptual model includes six interrelated components, namely, (1) The subject
facility, (2) Baseline model, (3) Sources of new semantics, (4) Semantic enrichment method,
(5) Enriched model, and (6) Application. For a given facility, semantic enrichment is the
process that extracts new semantics from the new semantics sources and adds them to the
baseline BIM or CIM so that the updated model has a richer digital representation of the facility.

99

100
101

Figure 1. A guiding conceptual model of semantic enrichment in BIM and CIM

102
103
104
105
106
107
108
109

The matrices for measuring the new semantic richness of BIM and CIM are the Level of
Development (LOD) and the Level of Detail (LoD), respectively, as shown in Figure 1. LOD
for BIM is defined on the temporal development of buildings, where LOD 100 is the conceptual
design without shape information or geometric representation and LOD 500 is the as-is BIM
with verified representation in terms of size, shape, location, quantity, and orientation (Boton
et al., 2015; Lu et al., 2018). There are intermediate levels, such as LOD 350 and LOD 450
(Chen et al., 2015; Abualdenien and Borrmann, 2019). So, the semantics in a BIM can be
4

110
111
112
113
114
115
116
117
118
119

enriched by designs, plans, surveying, documents, inspections, and operations along the
building life cycle. In this view, BIM with replenishing semantics can be a potential baseline
model for enriching CIM. In contrast, LoD, which originated in the CityGML (City Geography
Markup Language) standard (Chen et al., 2020), emphasizes the spatial detailing in CIM. For
buildings, LoD 0 denotes 2.5D block models, and LoD 4 means a ‘walkable’ model with
building interiors. For example, a well-designed BIM in the tendering phase can be LoD 4.3
for CityGML, but only LOD 200 for BIM. In non-building facilities such as vegetation, LoD
0 represents crown projection, whereas LOD 4 gives third-level branches (Ortega-Córdova,
2018). There also can be intermediate LoD levels, such as LoD 1.1 and LoD 3.4 in Biljecki et
al. (2016b).

120
121
122
123
124
125
126
127
128
129

The six components in the conceptual model can be further described with categorical values,
as listed in Table 2. The subject facility can be indoor space and furniture, buildings,
infrastructure, city objects, or even the environment. Traditionally, there are two silos, i.e.,
building level and city level, which do not talk to each other. A baseline model is a given,
normally preliminary, digital representation of the facility. It can be a BIM or CIM. Here, a
CIM is a fully 3D city model or a 2D GIS model with certain city objects linked to their 3D
models. Semantic enrichment in the BIM and CIM integration scenario focuses on the
interoperability problems between the two. The minor descriptor in Table 2 shows that the
baseline model could be in an open or commercially protected format.

130
131

Table 2. Descriptors of the six components in the conceptual model of semantic enrichment
Component
Subject
facility

Baseline
model
Source of
new
semantics

Major
Indoor
Building
Infrastructure
City objects
Environment
BIM
CIM
Data

Existing components
Ontology
Automated
method
Enriched
CIM/BIM

Semantic reasoning
Semantic registration
Semantic segmentation
Format

Descriptor
Minor

Open
Commercial
Open
Commercial
Raster (2D/3D)

Code
IN
BLD
INFR
CO
RE
BIM-O
BIM-C
CIM-O
CIM-C
D-R

Vector (2D/3D)
Tabular

D-V
D-T

BIM
CIM
Formal
Ad hoc rules
Spatial relations

C-BIM
C-CIM
O-F
O-R
O-S
Rsn
Rgst
Sgmt

BIM
CIM

5

Examples
Furniture, room space
Facades, structural elements
Bridge, tunnel
Lamppost, park
Vegetation
IFC
Autodesk Revit
CityGML
Esri ArcGIS
Digital pictures, radar images,
voxels
LiDAR points, lines, polygons
Descriptive data stored in rows
and columns
BIM components
3D city objects
Complete ontology definition
Case-specific rules
Geological relationship
Rule-based reasoning
ICP algorithm
Supervised

New semantics
Application

Pre-operation
Operation &
Maintenance
Urban environment
Energy
Unspecified case

Geometric
Non-geometric
Both

NS-G
NS-N
NS-B
PO
OM

New openings on a wall
A group of a furniture

UB
EN
UNS

Urban planning
Energy performance simulation
No application mentioned

Construction progress control
Facility maintenance

132
133
134
135
136
137
138
139
140
141
142
143

The source of new semantic can be classified into data, existing components from BIM/CIM,
and ontology. The data can be further divided into raster data, vector data, and tabular data.
Raster data (GISGeography, 2015) could be 2D or 3D, as represented in satellite photos, radar
images, and voxels. Vector data consists of LiDAR points, lines, and polygons (Koch and
Heipke, 2006). Tabular data is descriptive data stored in rows and columns in the database and
can be linked to spatial data. The second category of new semantics sources is the components
stored in various BIM or CIM libraries. They can be produced by manufacturers, software
vendors, or other interested modelers for private or open access. The ontology sources contain
formal ontology, ad hoc rules, and geological relationships. Unlike data and existing BIM/CIM
components, ontology is an often-neglected source of new semantics, particularly position and
typology information.

144
145
146
147
148

149
150
151
152
153
154
155
156
157
158

In general, there are three types of semantic enrichment methods:
(i)
Semantic reasoning based on pre-determined rules or ontology,
(ii)
Semantic registration based on many-round trial-and-error searches, and
(iii) Semantic segmentation based on an annotated training data set.
It needs to be pointed out that we excluded manual semantic annotation methods in this study
in that they are labor-intensive, time-consuming, and error-prone (Tang et al., 2010; Brilakis,
2010; Xiong et al., 2013). This is particularly true when it comes to larger sites or city level.
The literature search later in this paper also suggests that pure manual semantic methods are
rarely reported, although they are indispensable to fine-tune or correct the models generated
from automated methods. Some research created new geometric and non-geometric semantics
with a higher LOD/LoD. The output-enriched models could appear as BIM or CIM. Various
applications can be realized through semantic enrichment. As shown in Figure 1, in a sense,
semantic enrichment is the process of inferring and enriching the semantics to a given model
based on domain knowledge (Bloch and Sacks, 2018).

159
160
161
162
163
164

2.2 Data collection
The data collection process encompassed collecting, identifying, scoping, synthesizing,
screening, and reducing following the Preferred Reporting Items for Systematic Reviews and
Meta-Analyses (PRISMA) standard. The first step encompasses the collection of related
publication by applying the combinations of query strings of “"Semantic enrichment"
6

165
166
167
168
169
170
171

("building information model*" OR BIM OR "geographical information system*" OR GIS OR
"city information model*" OR CIM)” and “((BIM GIS) OR CIM) (semantic* OR LOD OR
interoperability)” in the full-text search engines of Google Scholar. The first query string
focused on the papers explicitly declares the keyword semantic enrichment in related fields,
while the second one aims to include more papers on semantic enrichment in a BIM and CIM
integration scenario. Similar query terms were applied with the ‘snowball’ method in Scopus
and the university library search engine to ensure the completeness of collected data.

172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188

Then, in collecting, identifying, and scoping the literature, a quick screening was conducted to
identify relevant articles regarding the titles, abstracts, and keywords. The priority was given
to the journals in the building and urban domains. The reason is that journal papers offer more
reliable and accurate materials due to rigorous review procedures (Akram et al., 2019). Based
on the proposed conceptual model, non-empirical studies such as commentary and position
papers were excluded. Empirical research can provide ‘verifiable’ evidence, which is
rigorously resultant from specific empirical studies. A snowballing technique (Oraee et al.,
2017) was employed to ensure the scope of the review is complete. The full-texts of the
screened publications were reviewed by all of the authors to confirm the coverage’s
completeness. Finally, the process resulted in 44 publications, including 36 journal papers, 4
conference papers, 2 reports, 1 Ph.D. thesis, and 1 book chapter. We extended the scope to
include the early years, but most of the pertinent publications fell between 2010 and 2019 (both
inclusive). The reason is plausible because the National BIM Standard was published in the
United States in 2007, and Open Geospatial Consortium (OGC) members adopted CityGML
as the official standard almost simultaneously in 2008. Thus, we crossed out a few cases before
2010 to reflect the most recent advancements in the fields.

189
190
191
192
193
194
195
196
197
198
199
200
201

In the third step of synthesizing, screening, and reducing, we examined the semantic
enrichment methodologies in the 44 publications guided by the conceptual model, as shown in
Figure 1. With the descriptors defined in Table 2, the basic information of the selected
empirical studies is extracted and tabulated in Table 3. It can be seen from Table 3 that the
studies are grouped into clusters. The first 22 publications were enrichment cases for BIM,
while the remaining 22 were for CIM. The publication year, subject facility, baseline model,
source of new semantics, semantic enrichment methods, new semantics, and output enriched
model were coded for each paper. For example, the fourth item Kim et al. (2013) started from
a commercial BIM baseline model (BIM-C), added a new source of vector data (D-V),
employed the Iterative Closest Point (ICP) – a registration method (Rgst), enriched building’s
BIM format (BLD) with new geometric and non-geometric semantics (NS-B), and
implemented an application in the pre-operation phase (PO).

202
203

Table 3. Basic information of the empirical studies on semantic enrichment for BIM and CIM
No.

Reference

Subject
facility

Baseline
model

Source

7

Method

New
semantics

Enriched
model

Application

1
2
3
4
5
6
7
8
9
10
11
12

BIM semantic enrichment
BIM-O
D-T
BIM-O
D-R
BIM-C
D-R
BIM-C
D-V
BIM-C
D-V
BIM-C
C-CIM
BIM-O
D-T
BIM-O
O-R
BIM-C
D-R
BIM-C
D-R
BIM-O
O-R
BIM-C
D-V

BLD
BLD
BLD
BLD
BLD
BLD
BLD
INFR
IN
BLD
BLD
BLD

13

Eastman et al. (2010)
Golparvar-Fard (2011)
Bosché (2012)
Kim et al. (2013)
Xiong et al. (2013)
Irizarry et al. (2013)
Zhang et al. (2014)
Lee et al. (2014)
Hong et al. (2015)
Wang et al. (2015)
Belsky et al. (2016)
Zeibak-Shini et al.
(2016)
Bassier et al. (2016)

BLD

BIM-O

D-V

14
15

Sacks et al. (2017)
Bloch et al. (2018)

INFR
IN

BIM-O
BIM-C

D-T
O-R, D-T

16

Xue et al. (2018)

BIM-C

17

Hamid et al. (2018)

18
19
20
21
22

Bienvenido-Huertas et
al. (2019)
Koo et al. (2019)
Simeone et al. (2019)
Xue et al. (2019a)
Xue et al. (2019b)

BLD,
IN
BLD,
IN
BLD

23

Wittner (2010)

24

El-Mekawy et al.
(2011)
He et al. (2012)
Irizarry et al. (2012)
Löwner et al. (2013)
Xu et al. (2014)
Mignard et al. (2014)

25
26
27
28
29
30
31

Rsn
Rgst
Rgst
Rgst
Rgst
Rsn
Rsn
Rsn
Sgmt
Sgmt
Rsn
Rsn

NS-G
NS-G
NS-G
NS-B
NS-G
NS-N
NS-B
NS-B
NS-G
NS-G
NS-G
NS-B

BIM
BIM
BIM
BIM
BIM
BIM
BIM
BIM
BIM
BIM
BIM
BIM

PO
PO
UNS
PO
UNS
PO
PO
OM
PO
EN
PO
OM

NS-G

BIM

PO

NS-N
NS-B

BIM
BIM

PO
PO

D-R

Sgmt,
Rsn
Rsn
Sgmt,
Rsn
Rgst

NS-B

BIM

PO

BIM-C

D-T

Rsn

NS-B

BIM

PO

BIM-C

D-R

Sgmt

NS-B

BIM

OM

BIM-O
C-BIM
BIM-C
O-R
BIM-C
D-V
BIM-C
D-V
CIM semantic enrichment
BLD,
CIM-C
D-V, D-R
IN
BLD
CIM-O
C-BIM,
O-F
CO
CIM-C
O-R, D-R
BLD
BIM-C
C-CIM
CO
CIM-O
O-R
CO
CIM-O
C-BIM
CO,
CIM-O
O-F
BLD
INFR
BIM-O
C-CIM
BLD
BIM-O
D-V

Sgmt
Rsn
Rgst
Rgst

NS-B
NS-B
NS-B
NS-B

BIM
BIM
BIM
BIM

PO
OM
PO
PO

Rsn

NS-B

CIM

UB

Rsn

NS-G

CIM

PO

Rsn
Rgst
Rsn
Rsn
Rsn

NS-G
NS-N
NS-G
NS-N
NS-N

CIM
CIM
CIM
CIM
CIM

UB
PO
PO
UB
UB

Rsn
Rsn

NS-B
NS-B

CIM
CIM

PO
UB

CIM-O
CIM-O
CIM-O
CIM-O
CIM-C

C-BIM
O-R
O-F
O-R
O-F

Rsn
Rsn
Rsn
Rsn
Rsn

NS-B
NS-G
NS-G
NS-G
NS-B

CIM
CIM
CIM
CIM
CIM

OM
PO
EN
UNS
UB

CIM-O

O-F

Rsn

NS-B

CIM

PO

CIM-C

C-BIM,
D-V, D-T
D-T
O-S
C-BIM,
D-R, D-T
C-CIM
C-BIM
D-V, D-R

Rsn

NS-B

CIM

PO

Rsn
Rsn
Rsn

NS-G
NS-B
NS-B

CIM
CIM
CIM

OM
UB
OM

Rsn
Rsn
Rsn

NS-B
NS-G
NS-G

CIM
CIM
CIM

OM
UB
UB

BLD
BLD
IN
IN

32
33
34
35
36

Borrmann et al. (2015)
Amirebrahimi et al.
(2015)
Kang et al. (2015)
Biljecki et al. (2016a)
Biljecki et al. (2016b)
Deng et al. (2016)
Hor et al. (2016)

37

Karan et al. (2016)

38

Peng et al. (2016)

BLD
BLD
BLD
CO
CO,
BLD
BLD,
RE
BLD

39
40
41

Howell et al. (2017)
He et al. (2018)
Lee et al. (2018)

RE
RE
INFR

BIM-O
CIM-C
CIM-C

42
43
44

Wang et al. (2019)
Zhao et al. (2019)
Aleksandrov (2019)

CO
INFR
BLD,
RE

BIM-O
CIM-C
CIM-O

204

8

205
206
207
208
209
210
211
212

3. A review of the semantic enrichment of BIM and CIM
An overview of the 44 semantic enrichment studies is visualized as a Sankey chart in Figure 2.
In Figure 2, the six major components in the conceptual model were shown as columns from
left to right, while two columns, ‘model format’ and ‘new semantics,’ were added to explain
the formats of baseline models and the results of methods, respectively. The size of a
rectangular component ‘stock’ indicates the number of cases associated with the label in Table
3. A curved ‘flow’ between two components stands for the frequency of co-occurrence in the
44 cases.

213

214
215

Figure 2. Sankey chart of the semantic enrichment studies for BIM and CIM (2010 – 2019)

216
217
218
219
220
221
222
223
224
225

3.1 The subject facilities
As shown in Figure 2, the distributions of subject facilities were considerably uneven, where
28 out of 44 studies focused on buildings. In BIM semantic enrichment, seventeen studies
targeted buildings, six on indoor areas, and two on infrastructures. Note that some papers, e.g.,
Xue et al. (2018) and Hamid et al. (2018), involved both building and indoor furniture in their
validations. In contrast, the subjects were more balanced for CIM. Twelve CIM cases targeted
buildings, one on indoor areas, seven on city objects, four on environments, and three on
infrastructure. The building, which is vital to the daily work and life of humankind in cities, is
a critical meeting point between BIM and CIM.

226
227
228
229
230
231

3.2 The baseline models and formats
The baseline models, by definition, can be divided into two types: BIM and CIM. Each type
had two formats, i.e., commercial and open, as shown in Figure 2. Overall, 27 cases were based
on BIMs, while 17 on CIMs. For BIM semantic enrichment, all 22 cases used BIMs as the
baseline models, 14 of which were commercial BIM formats and solutions (e.g., Autodesk
9

232
233
234
235
236

Revit). In comparison, open BIM formats and solutions, such as IFC and ISO 16739-1:2018,
were adopted in 8 studies, slightly less than the commercial ones. Powerful and open—to a
certain extent—APIs (Application Programming Interfaces) and online free development
documents offered by commercial BIM software vendors can be good explanations (Xue et al.,
2019a).

237
238
239
240
241
242
243
244
245
246
247

Unlike the BIM ones, only 5 out of 22 CIM cases were enriched above baseline BIMs (Irizarry
et al. 2012; Borrmann et al., 2015; Amirebrahimi et al., 2015; Howell et al., 2017; Wang et al.,
2019), while the majority was based on CIMs. Over half (14 out of 22) of baseline models had
open formats such as CityGML, while the remaining employed commercial formats and
platforms such as ESRI ArcGIS. The open formats, such as CityGML or IFC, are known to
have better transparency, operability, or scalability. For example, Deng et al. (2016) applied an
instance-based method to generate mapping rules between CityGML and IFC to enable
bidirectional semantic conversions. The commercial solutions to CIM, similar to those for BIM,
were good at handling large-scale models and data. For instance, He et al. (2018) enriched
geological semantics on top of a fully 3D city model in ArcGIS.

248
249
250
251
252
253
254
255
256
257
258
259
260
261

3.3 The sources of new semantics
The sources of new semantics consist of three major types: data, ontology, and existing
BIM/CIM components. As shown in Table 3 and Figure 2, 17 out of 22 semantic enrichment
cases for BIM extracted new semantics from various types of data. Among these, 6 studies
involved vector data. For instance, Kim et al. (2012) and Xue et al. (2019a; 2019b) processed
LiDAR points to produce semantically enriched BIMs. Also, 6 studies employed raster data.
For example, Xue et al. (2018) used photos. Eastman et al. (2010), Zhang et al. (2014), Sacks
et al. (2017), and Hamid et al. (2018) exploited tabular data. A few studies looked into the
ontology sources (Lee et al., 2014; Belsky et al., 2016). Simeone et al. (2019) deployed an
ontology-based system together with ad hoc rules as the source for enriching BIM; one study
used both tabular data and ad hoc rules in semantic enrichment (Bloch et al., 2013). For the
existing components, Irizarry et al. (2013) and Koo et al. (2019) reused BIM and CIM
components.

262
263
264
265
266
267
268
269
270
271

In 11 out of 22 CIM studies, data was the semantic source. For instance, Howell et al. (2017)
applied tabular data, including water metering as the source; Wittner (2010) and Aleksandrov
(2019) triangulated both raster and vector data. Ten studies applied ontologies to inferring new
semantics (e.g., geological relationships utilized in He et al. [2018]). Various rules, as implicit
or incomplete ontologies, were also popular as the semantic sources. For example, Löwner et
al. (2013) studied 44 combinations of Geometric and Semantics Level of Detail
(GLOD/SLOD). BIM components were adopted as the semantic source in six cases, where the
most frequent research design trend was the open BIM to CIM conversion, such as from IFC
to CityGML. For instance, Xu et al. (2014) and Kang et al. (2015) tested different ways for
10

272
273
274
275
276
277

importing IFC components into CIMs. A few cases channeled semantic elements from
reference CIMs to generate new CIMs (Irizarry et al., 2012; Borrmann et al., 2015; Wang et
al., 2019). Some studies integrated multiple types of semantic resources. For example, He et al.
(2012) exploited both raster data (e.g., 2D cadastral maps) and ad hoc rules; EI-Mekawy (2011)
used BIM components and formal ontology; Lee et al. (2018) involved raster data, tabular data,
and BIM components simultaneously.

278
279
280
281
282
283
284
285
286
287
288
289

3.4 The enrichment methods
3.4.1 Building information model
The 22 BIM cases covered all three types of enrichment methods, namely semantic reasoning,
semantic registration, and semantic segmentation. A semantic reasoning method holds a set of
fixed, usually predetermined processes without computational evolution or adaptations. More
specifically, semantic reasoning methods included rule-based, formal ontology-based, and
simulation ones. As exhibited in Table 3, semantic reasoning was the method for 11 of the BIM
cases. An example was Belsky et al.’s (2016) SeeBIM engine which reasoned IFC models with
five parts: parser, database, three-tiered structure inference rules, IFC writer, and ruleprocessing engine; through the reasoning engine, missing precast joints, connections, and slab
aggregations of a concrete parking garage were enriched automatically.

290
291
292
293
294
295
296
297

The semantic segmentation consists of supervised machine-learning methods, which concluded
hyper-models of classification from annotated training data sets. Six cases relied on semantic
segmentation. For example, Bienvenido-Huertas et al. (2019) applied the J48 algorithm; Koo
et al. (2019) applied support vector machines for semantic enrichment. Some segmentation
studies involved multiple methods. For instance, Bloch (2018) compared supervised machine
learning to reasoning methods in terms of feature collection and datasets; Bassier et al. (2016)
applied both segmentation and reasoning methods.

298
299
300
301
302
303
304
305
306
307

In total, 7 out of 22 BIM cases adopted semantic registration methods, which were evolutionary
computations through multiple rounds (iterations) of searching in a trial-and-error fashion. Kim
et al. (2012) employed the ICP method for matching 3D measurement data with existing BIM.
Advanced evolutionary computation (EC) methods, such as the covariance matrix adaptation
evolution strategy (CMA-ES) and Niching Migratory Multi-Swarm Optimization (NMMSO),
were applied to process 2D and 3D raster data to semantically enriched as-built BIMs for
complex indoor and outdoor scenes (Xue et al., 2018; 2019b). The key to semantic registration
is a reward (or penalty) function, which can be the photo’s structural similarity (SSIM) metrics
or root-mean-square errors (RMSE) between 3D data and models (Xue et al. 2019a).

308
309
310
311

3.4.2 City Information Model
The variety of enrichment methods for CIM was much lower. Semantic reasoning dominated
the CIM studies. Many reasoning approaches were applied for mapping and converting IFC,
11

312
313
314
315
316
317
318
319
320
321
322
323
324

an open BIM standard, to CityGML. Example methods include semantic web technology (Zhao
et al., 2019) and data interface modeling (Amirebrahimi et al., 2015). Enrichment to
CityGML’s LoD standard was a topical research direction in CIM. Examples included Biljecki
et al.’s (2016a; 2016b) 16 levels of LoDs and Löwner et al.’s (2013) significant extensions like
GLOD, SLOD, and ALOD (appearance LoD). With the help of proprietary open APIs and free
development documents available online, commercial formats and models were also used for
semantic reasoning. For instance, He et al. (2018) enriched the expressions in geological
relationships via a Diagrammer Geodata baseline model. Whether open or commercial, the key
to enabling semantic reasoning was the interoperability issues between different formats and
standards. Once the interoperability issues were solved, e.g., by mapping the ontologies or
functional APIs, semantic reasoning would become a shortcut to reuse existing BIM and CIM
resources to create CIMs at high LoD levels. Only one study (Irizarry et al., 2012) adopted the
registration method.

325
326
327
328
329
330
331
332
333
334

3.5 The enriched models and semantics
In the 22 cases of enriched BIMs, twelve received both geometric and non-geometric new
semantics; eight had new geometric semantics only, and two gained non-geometric semantics
only. In contrast, ten out of 22 CIM cases had both types of new semantics; nine cases aimed
at geometric semantics only; while only three cases, Irizarry et al. (2012), Mignard et al. (2014),
and Xu et al. (2014) enriched non-geometric semantics to its CIM. To sum up, semantic
enrichment studies for BIM weighed the non-geometric semantics more (14 out of 22) apart
from the geometry (20 out of 22); CIM focused on the geometry (19 out of 22). The result was
consistent with the guiding scales, i.e., LOD and LoD, to BIM and CIM.

335
336
337
338
339
340
341
342
343
344
345
346
347

More interesting was about the subtypes of the physical facilities and their new semantics. As
listed in Table 4, there were over 20 types of objects (e.g., furniture and room space) under the
five major types of subject facilities. Their geometric attributes (e.g., area), non-geometric
types (e.g., room type), identifiers (e.g., manufacturer), and relations (e.g., spatial relationship)
were concluded from the literature. First, the objects listed under the building and city objects
were the most diversified; meanwhile, these objects also had the most diversified geometric
attributes, e.g., a building object can have a long list of attributes including height, distance,
size, area, material, shape, dimension, volume, thickness, length, and proximity. The reason
for the most listed objects under the building and city objects may be the diversification of
related University majors and faculties. For example, the faculty of Architecture, the faculty of
Civil Engineering, and the faculty of Geographic Information Technology have various studies
on the semantics of buildings and cities in BIM and CIM.

348
349

Table 4. List of geometric and non-geometric semantics for BIM and CIM in the literature
Enriched
model

Subject
facility

Geometric

Non-geometric

12

BIM

BIM &
CIM

Indoor
facility

Building

Infrastructure

CIM

City
objects

Environment

Objects: furniture, room space,
kitchen cabinetry

Type: Room type, OmniClass number,
model, style number, identity data

Attribute: size, area, material, height

Identifier: manufacturer, material
product line, phasing, release date, cost,
assembly code, assembly description
Type: functional classification

Objects: interior entities (e.g., door, wall,
window, column, beam, covering, ceiling,
footing, pile, plate, roof, slab, railing,
ramp, stair, floor, precast concrete
component are BIM objects defined inside
building or as building parts), balconies,
exterior entities (e.g., door, wall, window,
roof, facade are CIM objects because they
are part of the external environment of the
building), building, parking garage

Identifier: construction progress,
degree of intervention, knowledge
representation and management, cost
Relation: spatial relation, logical
relation

Attributes: height, distance, size, area,
material, shape, dimension, volume,
thickness, length, proximity
Objects: bridge, highway, tunnel

Type: design options

Attribute: quantity, shape

Identifier: cost, environmental impact

Objects: exterior entities (e.g., in CIMs,
the outer surfaces of windows, doors,
roofs, and facades are regarded as part of
the exterior of buildings), building,
outbuilding shell, road, public garden,
drainage system, structure elements,
ground, streetlight, pedestrian path

Relation: spatial relation
Type: place type, component type, land
title number
Identifier: land use, land cover,
building facilities information, campus
water supply pipeline information,
component information, coordinate

Attribute: length, shape, geometrical
resolution, semantic depth, texture,
distance
Objects: water utility, vegetation,
geological object

Relation: spatial relation

Attribute: size, shape, texture

Identifier: pipe description, household
description

Type: telemetry data, valve and pump
states, smart metering data

Relation: spatial relation, geological
structural relation, geological
phenomenon relation, geological object
relation
350
351
352
353
354
355
356

The most diversified identifiers were found in indoor facilities and city objects. Most subtypes
of relations were found from the enrichment cases on the environment semantics. For example,
the spatial relation in the geological environment gives the position of the rock in space relative
to a reference object (e.g., the position of a tree). As another example, the logical relation in
the geological environment gives the dependence between two rock layers. The items in Table
4 reveal each subject facility’s potential in semantic enrichment for BIM and CIM, which can
13

357
358
359
360

partially explain why buildings were the most targeted facilities in literature. In the future, the
indoor objects’ identification can become an emerging trend for BIM, while city objects’
potentials on the object types and the environment’s potential on the relations can lead to new
explorations in the semantic enrichment for CIM.

361
362
363
364
365
366
367
368
369
370
371
372

3.6 The applications
Applications and the values embedded specify the target semantics and initiate enrichment
processes. In BIM semantic enrichment studies, 15 were applied in pre-operation phases (e.g.,
planning and construction), four in operation and maintenance, one in energy, while two were
unspecified. Most novel value-added applications of BIMs lie in the pre-operation phases of
buildings. The distribution reflects the scope of BIM studies and the interest of BIM researchers
and practitioners. In comparison, applications of CIM semantic enrichment showed more
balanced and diverse fields. Nine CIM cases were applied in the urban environment (e.g., urban
planning), seven in pre-operation, four in operation and maintenance, one in energy, while one
paper did not specify its application. The diverse applications echo the diverse sources of CIM
semantics, as shown in Table 3.

373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391

4. The trends and prospects of semantic enrichment of BIM and CIM
Figure 3 shows the stacked bar charts of the yearly semantic enrichment cases for BIM and
CIM from different perspectives. First, Figure 3a shows that in BIM studies, there has been a
surge of indoor facility studies since 2018, while the study of buildings has increased relatively
steadily since 2010. In comparison, the CIM cases had more diversified and balanced subject
facilities. For example, in 2019, the four cases covered four different types of subject facilities,
while the environment was the latest emerged subject (in 2016). Virtual reality and augmented
reality (VR/AR), which had recently become a household term, was one enabler of indoor
facility studies (Chen and Xue, 2011). For instance, Google’s Project Tango was a pioneer
project for household VR/AR, from which two smartphone models were mass-produced. One
was Lenovo Phab 2 Pro, released in December 2016, and the other was Asus Zenfone AR,
announced in 2017. Although Tango was not a successful project in terms of business, the new
VR/AR sensor data directly inspired a few indoor facility studies as semantic sources: e.g., Xue
(2019a). In 2020, new mainstream smartphone models have Apple’s AR-Kit or Google’s
ARCore installed; furthermore, VR/AR is growing extensively. For example, IDC (2019)
predicted that global VR/AR spending would have a five-year compound annual growth rate
(CAGR) of 77.0%. Thus, semantic enrichment of indoor facilities should gain new momentum
in the future.

14

392
393

Figure 3. Stacked bar charts of yearly published semantic enrichment cases for BIM and CIM

394
395
396
397

Figure 3b shows that since 2018, commercial baseline models have been increasingly used in
BIM cases. During the same period, only a few CIM cases adopted commercial baseline models.
For the BIM cases, one reason was the poor efficacy of the open BIM standard— IFC. For
15

398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413

example, Zhiliang et al. (2020) pointed out that the information corresponding to the
established property sets can only be read, but its semantics cannot be automatically identified
in the current IFC. De Gaetani (2020) recognized faults when exchanging sample IFC BIM
models and related Gantt charts. Besides, information redundancy, such as the randomized
sequential identifiers (#-Ids) and complicated cross-referencing relation systems, makes it
difficult to extract and, particularly, to enrich BIM information efficiently using IFC
(Borrmann et al. 2018; Xue and Lu, 2020). In comparison, CIM semantic enrichment studies
used to be benefited from OGC’s long list of open standards covering a wide spectrum from
data to model to tool. However, a few recent semantic enrichment studies for CIM were
benefited from commercial but powerful APIs, e.g., He et al. (2018), or the CIM-BIM
integration, e.g., Lee et al. (2018) and Zhao et al. (2019). Considering the development of IFC
and recent collaborations between BIM and CIM software giants, we would likely see more
semantic enrichment studies utilizing commercial standards and solutions in the next five years
for BIM and CIM. In the longer run, open standards have the ground to continue its prevalence
for CIM, as well as for BIM in case the APIs to IFC such as the IfcOpenShell project (Krijnen,
2011) are competitive against commercial platforms.

414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437

For both BIM and CIM, all three types, i.e., data, model, and ontology, have been involved as
sources of new semantics, as shown in Figure 3c. In the BIM cases, data was the most
frequently used, while existing components were the least. In contrast, situations were the
opposite for CIM – data was the least frequent while existing components and ontologies were
the most. Availability of city-scale measurement data should not be among the reasons because
accurate data sets such as satellite photos, radar images, and 3D LiDAR point clouds became
much available and affordable in the recent decade. However, the availability of effective
algorithms and methods to pre-process and process the city-scale data sets can be a cause. For
example, the 2015 Dublin LiDAR was an open-source data set consisting of 1.4 billion 3D
points; it wasn’t until 2019 that Zolanvari et al. published a pre-processed data set with a sheer
18.6% point cloud with semantic annotations. Nevertheless, researchers on CIM can refer to
up-to-date data processing methods for BIM (e.g., deep learning, statistical modeling, modelbased down-sampling, and optimization algorithms) to handle city-scale data. Therefore, we
are open to seeing more data-driven semantic enrichment studies for CIM in the future. For
BIM, components and model is a promising source of new semantics to take advantage of the
CIM-BIM integration.
The three types of methods were roughly equally employed for BIM semantic enrichment in
terms of frequency, while semantic segmentation methods appeared in the second half of the
review period. The situation for CIM was different, where most cases were conducted through
semantic reasoning methods. The major reason should be that the rule of ontology-based
reasoning is the most efficient (i.e., least time cost) when data grows to the city scale. In
comparison, semantic registration methods for BIM applied a considerable number of trial16

438
439
440
441
442
443
444
445

and-error tests to match the indoor facility to correct 3D models; and semantic segmentation
methods require a large number of training examples (e.g., Zolanvari et al. 2019). However,
we are optimistic to see that the research communities for BIM and CIM can learn effective
and efficient technologies from each other, along with the emerging trend of crossing the
disciplinary boundary. Besides, new technologies, such as big data analytics, which can
discover new relationships from complex urban big data, are also potential game-changers to
process city-scale data sets (Lu et al., 2018). Therefore, the semantic enrichment methods for
CIM are very likely to be more efficient and diversified.

446
447
448
449
450

A shared value-added application can be a strong driver for promoting and integrating the
complementary semantics in BIM and CIM. For example, Figure 3e shows that both started to
impact the operation and maintenance phase in 2014. However, there still are some obstacles:
•

Different standard formats (Liu et al., 2017), for example, the IFC schemas are not fully
compatible with the definitions in CityGML;

•

Lack of legal protection for intellectual property rights (Ku and Taiebat, 2011), for
example, authorship, readership, and ownership of semantics within the model;

•

Funds and time costs required for integration (Suermann and Issa, 2009), for example,
employing modelers and consultants;

•

Unsuitable external environmental conditions (Azhar et al., 2012), for example, lack of
professional interactivity and external motivations;
Insufficient management (Eastman et al., 2011), for example, lack of a well-established
workflow and immature dispute resolution;

451
452
453
454
455
456
457
458

•

459
460

•

Different spatial scale (Song et al., 2017), for example, data and information with
multiple spatial scales; and

•

Different user focus (Liu et al., 2017), for example, BIM users used to focus on
construction projects, while CIM users may focus on spatial analysis of outdoor
environments.

461
462
463
464

465
466
467
468
469
470
471
472
473
474

In the past, much focus was on resolving technical difficulties, while management,
environmental and legal obstacles received less attention. Therefore, researchers should realize
that openness and collaboration are the keys to favorably integrating BIM and CIM.
Application-driven semantic enrichment, frequent communication, and government policies
are also essential for successful integrations. For example, recent smart buildings and city
initiatives require people in these two fields to collaborate to develop more semantic
enrichment applications (e.g., energy). It would be a misconception to assume that enriched
models will be automatically implemented in real life. Instead, realizing semantic enrichment
applications requires positive attitudes from people in different sectors. This is the only way
the nexus between the research and practice of semantic enrichment will spiral upward.

475

17

476
477
478
479
480
481
482
483
484
485
486
487
488

In short, this research developed a conceptual model consisting of six components to revisit the
existing semantic enrichment studies in BIM and CIM for outlining the trends and prospects.
Value-added applications and research gaps brought opportunities for future investigations.
Firstly, BIM semantic enrichment starting from CIM was rare in the past, and future research
breakthroughs can be extended from this perspective. Secondly, compared with other subject
facilities, researchers can also pay attention to indoor facilities. In the era of big data, people
can generate and access more and more indoor data, which can involve humans in the loop
through the semantics of indoor behaviors and facility usages. Thirdly, with the rapid
development of data processing algorithms, future research should attempt to use city-scale
data sets as a source of new semantics to help develop smart buildings and cities and manage
the life cycle of infrastructures. Last, but not least, research using semantic reasoning is
relatively mature and saturated, and future research can tackle more semantic registration or
segmentation methods.

489
490
491
492
493
494
495
496
497
498
499
500
501
502

The findings of this study also have practical implications. Although semantic enrichment is
relatively successful in the pre-operation phase, there is still a lack of systematic considerations
to consolidate isolated applications in different phases and departments. One reason is that user
interfaces are isolated at different phases of the construction life cycle (e.g., design,
construction, and post-maintenance). Another possible reason is that different government
departments hold CIM datasets. For example, in Hong Kong, the Land Department holds 3D
digital maps, while building semantics are separately managed by the Building Department,
Housing Department, Planning Department, and Home Affairs Department. Therefore, in
addition to technical obstacles, there are also institutional barriers. Furthermore, neither private
companies nor the public can access the datasets. Therefore, common urban semantic platforms,
such as Queriocity (Lopez et al., 2012) and Smart City Platform (ASTRI, 2019), should aim to
solve transparency and accessibility issues. Researchers interested in semantic enrichment can
also pay attention to emerging theoretical challenges from common urban semantic platforms.

503
504
505
506
507
508
509
510
511
512
513
514

5. Conclusions
Semantic enrichment is a process that adds semantic information to a building information
model (BIM) or a city information model (CIM) to enable applications pertinent to construction
management, geoinformatics, and urban planning. However, the process of harnessing
semantic enrichment to improve the existing BIM and CIM seems a conundrum. This paper
proposes a conceptual model representing the series of generic semantic enrichment processes
for both BIM and CIM. In the model, a semantic enrichment study first targets a subject facility,
then adopts a baseline BIM or CIM, locates sources of new semantics, distills the new
semantics using various methods, which results in an enriched BIM or CIM by appending the
newly distilled semantics, and finally realizes applications. The proposed model was applied
to 44 empirical studies in the literature.

515

18

516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534

A comparative analysis of semantic enrichment in BIM and CIM showed considerable
differences in terms of scopes, terminologies, methodologies, and scales. For example, a recent
study on BIM semantics focused more on buildings and indoors, while CIM research covered
buildings, infrastructure, urban objects, and environments. The future research will probably
gain opportunities from the penetrations of remote sensing devices (such as VR/AR
smartphone), intelligent data processing algorithms, BIM and CIM software platforms, and
other emerging technologies such as big data analytics. Since 2018, commercial platforms that
serve as baseline models for semantic enrichment in BIM and CIM have become common in
literature. There are BIM cases since open BIM standard IFC is not effective. CIM semantic
enrichment studies used to benefit from OGC’s long list of open standards; however, a few
recent CIM studies benefited from commercial but powerful APIs. When it comes to the source
of new semantics for BIM cases, data was the most frequently used, while existing components
were the least. In contrast, situations were the opposite for CIM. Notably, the enrichment
methods were rather different, too. Semantic reasoning, semantic registration, and semantic
segmentation were roughly equally employed for BIM semantic enrichment in terms of
frequency, while most CIM cases were conducted through semantic reasoning methods. BIM
semantic enrichment was mainly applied in pre-operation phases, while CIM semantic
enrichment was mainly applied in urban environments. However, recent trends indicated that
BIM and CIM have similar applications in the operation and maintenance phase.

535
536
537
538
539
540
541
542
543

The study is the first-ever one of its kind to examine the niche area of semantic enrichment.
Moreover, the study combs the development in fields and helps readers understand the changes
in semantic enrichment in the past ten years and its prospects or future research. Another
contribution of this research is the conceptual model of semantic enrichment comprising six
components that explain the rationale of semantic enrichment from a data processing
perspective. By triangulating the conceptual model with the historical data from the niche of
semantic enrichment research, this study summarized the research findings, based on which
future development could proceed on a more streamlined footing.

544
545
546
547
548
549
550
551
552
553
554

Although the study in this paper has many merits, it also has the following limitations. Firstly,
manual screening and coding work could inevitably present subjectivity. The accuracy can be
improved by discovering more related publications in future work. Secondly, the selection and
discussion of the semantic enrichment studies were specified to the empirical city and building
cases in the scholarly publications in the urban-related fields. Practitioners might consider reallife demands and conditions as well. A systematic recommendation of semantic enrichment
methods distilled from the best practices in the literature should be included in future work.
Amidst the general trend of enhancing semantic enrichment, researchers and practitioners alike
are encouraged to think out of the box, e.g., where are the optimal ‘sweet points’ of BIM and
CIM between the high LOD/LoD semantics and various costs (e.g., funds, privacy, social

19

555
556

inequality, and human-machine conflict), as it is certainly not the higher the LOD/LoD, the
better.

557
558
559
560
561
562

Acknowledgements
The work described in this paper was fully supported by a grant from the Research Grants
Council of the Hong Kong SAR (Project No. 27200520). We would like to express our
gratitude to the anonymous referees for the constructive comments on the clarity and quality
of this paper.

563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592

References
* Indicates references for empirical studies analyzed by the authors
Akram, R., Thaheem, M. J., Nasir, A. R., Ali, T. H. & Khan, S., 2019. Exploring the role of
building information modeling in construction safety through science mapping. Safety
Science, 120, 456-470. https://doi.org/10.1016/j.ssci.2019.07.036
*Aleksandrov, M., Diakité, A., Yan, J., Li, W., Barton, J., Doig, J. & Zlatanova, Z., 2019.
Progressing precinct modelling on the UNSW campus and beyond: BIM/PIM and 3DGIS.
Project Report. Australia Government. Sydney: CRC for Low Carbon Living.
https://www.be.unsw.edu.au/sites/default/files/upload/grid/LCL_CRC_3DPIM2_Data_M
odelling.pdf.
*Amirebrahimi, S., Rajabifard, A., Mendis, P. & Ngo, T., 2015. A Data Model for Integrating
GIS and BIM for Assessment and 3D Visualisation of Flood Damage to Building. CEUR
Workshop Proceedings. 1323, pp. 78-79. Brisbane: Locate. http://ceur-ws.org/Vol1323/paper27.pdf.
Abualdenien, J. & Borrmann, A., 2019. A meta-model approach for formal specification and
consistent management of multi-LOD building models. Advanced Engineering
Informatics, 40, 135-153. https://doi.org/10.1016/j.aei.2019.04.003
Azhar, S., Khalfan, M. & Maqsood, T., 2012. Building information modelling (BIM): now and
beyond.
Construction
Economics
and
Building,
12(4),
15-28.
https://doi.org/10.5130/AJCEB.v12i4.3032
Borrmann, A., Beetz, J., Koch, C., Liebich, T. & Muhic, S., 2018. Industry Foundation Classes:
A standardized data model for the vendor-neutral exchange of digital building models. In
B. A., K. M., K. C. & B. J., Building Information Modeling (pp. 81-126). Cham,
Switzerland: Springer. https://doi.org/10.1007/978-3-319-92862-3_5
Belsky, M., Sacks, R. & Brilakis, I., 2015. A framework for semantic enrichment of IFC
building models. Doctoral dissertation, Technion-Israel Institute of Technology, Israel,
viewed 15 December 2019, https://bit.ly/2RRPyg6
Berling-Wolff, S. & Wu, J., 2004. Modeling urban landscape dynamics: A review. Ecological
Research, 19(1), 119-129. https://doi.org/10.1111/j.1440-1703.2003.00611.x

20

593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632

*Belsky, M., Sacks, R. & Brilakis, I., 2016. Semantic enrichment for building information
modeling. Computer‐Aided Civil and Infrastructure Engineering, 31(4), 261–274.
https://doi.org/10.1111/mice.12128
*Biljecki, F., Ledoux, H. & Stoter, J., 2016a. An improved LOD specification for 3D building
models.
Computers,
Environment
and
Urban
Systems,
59,
25–37.
https://doi.org/10.1016/j.compenvurbsys.2016.04.005
*Biljecki, F., Ledoux, H., Stoter, J. & Vosselman, G., 2016b. The variants of an LOD of a 3D
building model and their influence on spatial analyses. ISPRS Journal of Photogrammetry
and Remote Sensing, 116(C), 42–54. https://doi.org/10.1016/j.isprsjprs.2016.03.003
*Bloch, T. & Sacks, R., 2018. Comparing machine learning and rule-based inferencing for
semantic enrichment of BIM models. Automation in Construction, 91, 256–272.
https://doi.org/10.1016/j.autcon.2018.03.018
*Bienvenido-Huertas, D. E., Nieto-Julián, J. J., Moyano, J. M., Macías-Bernal, J. & Castro, J.,
2019. Implementing Artificial Intelligence in H-BIM Using the J48 Algorithm to Manage
Historic
Buildings. International
Journal
of
Architectural
Heritage, 1-13.
https://doi.org/10.1080/15583058.2019.1589602
*Bassier, M., Vergauwen, M. & Van Genechten, B., 2016. Automated semantic labelling of
3D vector models for scan-to-BIM. Paper presented at the 4th Annual International
Conference on Architecture and Civil Engineering, 25–26 April, België, viewed 1 May
2020. https://doi.org/10.5176/2301-394X_ACE16.83.
*Bosché, F., 2012. Plane-based registration of construction laser scans with 3D/4D building
models.
Advanced
Engineering
Informatics,
26(1),
90-102.
https://doi.org/10.1016/j.aei.2011.08.009
*Borrmann, A., Kolbe, T. H., Donaubauer, A., Steuer, H., Jubierre, J. R. & Flurl, M., 2015.
Multi‐scale geometric‐semantic modeling of shield tunnels for GIS and BIM
applications. Computer‐Aided Civil and Infrastructure Engineering, 30(4), 263-28.
https://doi.org/10.1111/mice.12090
Brilakis, I., Lourakis, M., Sacks, R., Savarese, S., Christodoulou, S., Teizer, J. & Makhmalbaf,
A., 2010. Toward automated generation of parametric BIMs based on hybrid video and
laser
scanning
data. Advanced
Engineering
Informatics, 24(4),
456-465.
https://doi.org/10.1016/j.aei.2010.06.006
Boton, C., Kubicki, S. & Halin, G., 2015. The challenge of level of development in 4D/BIM
simulation across AEC project lifecycle. A case study. Procedia Engineering, 123, 59-67.
https://doi.org/10.1016/j.proeng.2015.10.058
Clarke, M., & Harley, P. 2014. How smart is your content? Using semantic enrichment to
improve your user experience and your bottom line. Science, 37(2), 41.
http://www.councilscienceeditors.org/wp-content/uploads/v37n2p40-44.pdf
Chen, K., Johan, H., & Erdt, M., 2020. Appearance-driven conversion of polygon soup building
models with level of detail control for 3D geospatial applications. Advanced Engineering
Informatics, 44, 101049. https://doi.org/10.1016/j.aei.2020.101049
21

633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672

Chen, K., & Xue, F. 2021. The renaissance of Augmented Reality in construction: history,
present status and future directions. Smart and Sustainable Built Environment, in press.
Doi: https://doi.org/10.1108/SASBE-08-2020-0124
Davies, R. & Harty, C., 2011. Building Information Modelling as innovation journey: BIM
experiences on a major UK healthcare infrastructure project. Paper presented at the 6th
Nordic conference on construction economics and organization–shaping the
construction/society nexus, 13–15 April, Copenhagen, viewed 1 January 2020.
https://www.semanticscholar.org/paper/Building-Information-Modelling-as-innovationBIM-on-Davies-Harty/1c26f66825994ca920f452151a30093b234af677
*Deng, Y., Cheng, J.C.P. & Anumba, C., 2016. Mapping between BIM and 3D GIS in different
levels of detail using schema mediation and instance comparison. Automation in
Construction, 67, 1–21. https://doi.org/10.1016/j.autcon.2016.03.006
De Gaetani, C. I., Mert, M., & Migliaccio, F. (2020). Interoperability Analyses of BIM
Platforms for Construction Management. Applied Sciences, 10(13), 4437.
https://doi.org/10.3390/app10134437
Eastman, C., Teicholz, P., Sacks, R. & Liston, K., 2011. A guide to building information
modeling for owners, managers, designers, engineers and contractors: Handbook, B. I. M,
viewed 1 January 2020. 10.1002/9781119287568.
*Eastman, C., Sacks, R., Panushev, I., Venugopal, M. & Aram, S., 2010. Precast/Prestressed
Concrete National BIM Standard Documents: Model View Definitions for Precast
Concrete. Technical Report, BIM Advisory Committee, Precast BIM Standard Project,
Precast/Prestressed Concrete Institute. https://itc.scix.net/pdfs/w78-2010-18.pdf
*El-Mekawy, M., Östman, A. & Shahzad, K., 2011. Towards interoperating CityGML and IFC
building models: a unified model based approach. In Advances in 3D geo-information
sciences (pp. 73-93). Berlin, Heidelberg: Springer. https://doi.org/10.1007/978-3-64212670-3_5
ESRI, 2019. ESRI-Autodesk Partnership. Retrieved December 8, 2019, from
https://bit.ly/38hz7Q7
*Golparvar-Fard, M., Peña-Mora, F. & Savarese, S., 2011. Integrated sequential as-built and
as-planned representation with D 4 AR tools in support of decision-making tasks in the
AEC/FM industry. Journal of Construction Engineering and Management, 137(12), 10991116. https://doi.org/10.1061/(ASCE)CO.1943-7862.0000371
GISGeography., 2015. Vector vs Raster: What’s the Difference Between GIS Spatial Data
Types? Retrieved May 8, 2020, from GISGeography: https://gisgeography.com/spatialdata-types-vector-raster/
Goodchild, M. F., 1991. Geographic information systems. Progress in Human
geography, 15(2), 194-200. https://doi.org/10.1177%2F030913259101500205
*He, H., Hu, D. & Lü, G., 2018. GIS Application to Regional Geological Structure
Relationship Modelling Considering Semantics. ISPRS International Journal of GeoInformation, 7, 267. https://doi.org/10.3390/ijgi7070267
22

673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710

*He, S., Besuievsky, G., Tourre, V., Patow, G. & Moreau, G., 2012. All range and
heterogeneous multi-scale 3D city models. Usage, Usability, and Utility of 3D City
Models–European
COST
Action
TU0801,
02006.
https://doi.org/10.1051/3u3d/201202006
*Howell, S., Rezgui, Y. & Beach, T., 2017. Integrating building and urban semantics to
empower smart water solutions. Automation in Construction, 81, 434.
https://doi.org/10.1016/j.autcon.2017.02.004
*Hong, S., Jung, J., Kim, S., Cho, H., Lee, J. & Heo, J., 2015. Semi-automated approach to
indoor mapping for 3D as-built building information modeling. Computers, Environment
and Urban Systems, 51, 34-46. https://doi.org/10.1016/j.compenvurbsys.2015.01.005
*Hamid, M., Tolba, O. & El Antably, A., 2018. BIM semantics for digital fabrication: A
knowledge-based
approach.
Automation
in
Construction,
91,
62-82.
https://doi.org/10.1016/j.autcon.2018.02.031
*Hor, A. H., Jadidi, A. & Sohn, G., 2016. BIM-GIS integrated geospatial information model
using semantic web and RDF graphs. ISPRS Annals of Photogrammetry, Remote Sensing
and Spatial Information Sciences, 3(4), 73-79. 10.5194/isprs-annals-III-4-73-2016
Hong Kong Applied Science and Technology Research Institute (ASTRI). 2019. Smart City
Platform, Retrieved October 27, 2020, from https://www.astri.org/tdprojects/smart-cityplatform/.
IDC. 2019. Worldwide Augmented and Virtual Reality Spending Guide. Framingham, MA,
USA:
International
Data
Corporation.
https://www.idc.com/getdoc.jsp?containerId=IDC_P34919
*Irizarry, J., Karan, E. P. & Jalaei, F., 2013. Integrating BIM and GIS to improve the visual
monitoring of construction supply chain management. Automation in construction, 31,
241-254. https://doi.org/10.1016/j.autcon.2012.12.005
*Irizarry, J. & Karan, E. P., 2012. Optimizing location of tower cranes on construction sites
through GIS and BIM integration. Journal of information technology in construction,
17(23), 351-366. https://www.itcon.org/2012/23
Julin, A., Jaalama, K., Virtanen, J. P., Pouke, M., Ylipulli, J., Vaaja, M. ... & Hyyppä, H., 2018.
Characterizing 3D city modeling projects: towards a harmonized interoperable system.
ISPRS
International
Journal
of
Geo-Information,
7(2),
55.
https://doi.org/10.3390/ijgi7020055
*Kim, C., Son, H. & Kim, C., 2013. Automated construction progress measurement using a 4D
building information model and 3D data. Automation in Construction, 31, 75–82.
https://doi.org/10.1016/j.autcon.2012.11.041
*Koo, B., La, S., Cho, N. & Yu, Y., 2019. Using support vector machines to classify building
elements for checking the semantic integrity of building information models. Automation
in Construction, 98, 183-194. https://doi.org/10.1016/j.autcon.2018.11.015

23

711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749

*Karan, E., Irizarry, J. & Haymaker, J., 2016. BIM and GIS Integration and Interoperability
Based on Semantic Web Technology. Journal Of Computing In Civil Engineering, 30(3),
04015043. https://doi.org/10.1061/(ASCE)CP.1943-5487.0000519
Koch, A. & Heipke, C., 2006. Semantically correct 2.5D GIS data — The integration of a DTM
and topographic vector data. ISPRS Journal of Photogrammetry and Remote Sensing,
61(1), 23-32. https://doi.org/10.1007/3-540-26772-7_39
*Kang, T. W. & Hong, C. H., 2015. A study on software architecture for effective BIM/GISbased facility management data integration. Automation in construction, 54, 25-38.
https://doi.org/10.1016/j.autcon.2015.03.019
Kang, Z., Yang, J., Yang, Z. & Cheng, S., 2020. A review of techniques for 3D reconstruction
of indoor environments. ISPRS International Journal of Geo-Information, 9(5), 330.
https://doi.org/10.3390/ijgi9050330
Krijnen, T., 2011. IfcOpenShell. http://ifcopenshell. org
Ku, K. & Taiebat, M., 2011. BIM experiences and expectations: the constructors’ perspective.
International Journal of Construction Education and Research, 7(3), 175-197.
https://doi.org/10.1080/15578771.2010.544155
*Löwner, M., Benner, J., Gröger, G. & Häfele, K.H., 2013. New concepts for structuring 3D
city models–an extended level of detail concept for CityGML buildings, Paper presented
at International Conference on Computational Science and Its Applications, 24-27 June,
Berlin, viewed 27 October 2020. https://doi.org/10.1007/978-3-642-39646-5_34
*Lee, S., Kim, K. & Yu, J., 2014. BIM and ontology-based approach for building cost
estimation.
Automation
in
Construction,
41,
96-105.
https://doi.org/10.1016/j.autcon.2013.10.020
Lu, W., Lai, C. & Tse, T., 2018. BIM and Big Data for Construction Cost Management.
HongKong, China: Routledge. 10.1201/9781351172325
Lu, W., Peng, Y., Xue, F., Chen, K., Niu, Y. & Chen, X., 2018. The fusion of GIS and Building
Information Modeling for big data analytics in managing development sites. Reference
Module in Earth Systems and Environmental Sciences. https://doi.org/10.1016/B978-0-12409548-9.09677-9
*Lee, P., Wang, Y., Lo, T. & Long, D., 2018. An integrated system framework of building
information modelling and geographical information system for utility tunnel maintenance
management. Tunnelling and Underground Space Technology Incorporating Trenchless
Technology Research, 79, 263-273. https://doi.org/10.1016/j.tust.2018.05.010
Lopez, V., Kotoulas, S., Sbodio, M. L., Stephenson, M., Gkoulalas-Divanis, A. & Mac
Aonghusa, P. 2012. Queriocity: A linked data platform for urban information management.
Paper presented at the International Semantic Web Conference, 11-15 Novermber, Berlin,
viewed 27 October 2020. https://doi.org/10.1007/978-3-642-35173-0_10
Liu, X., Wang, X., Wright, G., Cheng, J. C., Li, X. & Liu, R., 2017. A state-of-the-art review
on the integration of Building Information Modeling (BIM) and Geographic Information

24

750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789

System (GIS). ISPRS International Journal of Geo-Information, 6(2), 53.
https://doi.org/10.3390/ijgi6020053
*Mignard, C. & Nicolle, C., 2014. Merging BIM and GIS using ontologies application to urban
facility management in ACTIVe3D. Computers in Industry, 65(9), 1276-1290.
https://doi.org/10.1016/j.compind.2014.07.008
National Institute of Building Sciences, 2015. National Building Information Modeling
Standard. Version 3, https://www.nationalbimstandard.org
National Geographic Society, 2012. GIS (geographic information system).
http://www.nationalgeographic.org/encyclopedia/geographic-information-system-gis/
Oraee, M., Hosseini, M. R., Papadonikolaki, E., Palliyaguru, R. & Arashpour, M., 2017.
Collaboration in BIM-based construction networks: A bibliometric-qualitative literature
review. International Journal of Project Management, 35(7), 1288-1301.
https://doi.org/10.1016/j.ijproman.2017.07.001
Ortega-Córdova, L., 2018. Urban Vegetation Modeling 3D Levels of Detail.
http://resolver.tudelft.nl/uuid:8b8967a8-0a0f-498f-9d37-71c6c3e532af
Pratt-Hartmann, I., 2004. Fragments of lLanguage. Journal of Logic, Language and
Information, 13(1), 207–223. https://doi.org/10.1023/B:JLLI.0000024735.97006.5a
Ping, J., Changxiu, C., Jiong, X. & Rongguo, C., 2013. Design and Implementation of Manage
Model for Raster Metadata in Spatial Database. Remote Sensing Technology and
Application, 27(2), 315-321. CNKI:SUN:YGJS.0.2012-02-024
*Peng, L., Tang, S., Liu, M., Li, P., Zhu, Q. & Guo, H., 2016. Building Interval Planning
Approval Approach of BIM and GIS Integration. Geomatics World, 23(2), 32-37.
http://en.cnki.com.cn/Article_en/CJFDTotal-CHRK201602007.htm
Reinhart, C. F. & Davila, C. C., 2016. Urban building energy modeling–A review of a nascent
field.
Building
and
Environment,
97,
196-202.
https://doi.org/10.1016/j.buildenv.2015.12.001
*Sacks, R., Ma, L., Yosef, R. & Borrmann, A., 2017. Semantic enrichment for Building
Information Modeling: Procedure for compiling inference rules and operators for complex
geometry. Journal of Computing in Civil Engineering, 31(6), 04017062.
https://doi.org/10.1061/(ASCE)CP.1943-5487.0000705
Schlueter, A. & Thesseling, F., 2009. Building information model based energy/exergy
performance assessment in early design stages. Automation in Construction, 18(2), 153–
163. https://doi.org/10.1016/j.autcon.2008.07.003
*Simeone, D., Cursi, S. & Acierno, M., 2019. BIM semantic-enrichment for built heritage
representation.
Automation
in
Construction,
97,
122–137.
https://doi.org/10.1016/j.autcon.2018.11.004
Stojanovski, T., 2013. City Information Modeling (CIM) and urbanism: blocks, connections,
territories, people and situations. Paper presented at the Symposium on Simulation for
Architecture & Urban Design, 7–10 April, California, viewed 1 January 2020,
http://orcid.org/0000-0002-1790-0254
25

790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828

Suermann, P. C. & Issa, R. R., 2009. Evaluating industry perceptions of building information
modelling (BIM) impact on construction. Journal of Information Technology in
Construction, 14(37), 574-594. http://www.itcon.org/2009/37
Song, Y., Wang, X., Tan, Y., Wu, P., Sutrisna, M., Cheng, J. C. & Hampson, K., 2017. Trends
and opportunities of BIM-GIS integration in the architecture, engineering and construction
industry: a review from a spatio-temporal statistical perspective. ISPRS International
Journal of Geo-Information, 6(12), 397. https://doi.org/10.3390/ijgi6120397
Tang, P., Huber, D., Akinci, B., Lipman, R. & Lytle, A., 2010. Automatic reconstruction of asbuilt building information models from laser-scanned point clouds: A review of related
techniques. Automation
in
construction, 19(7),
829-843.
https://doi.org/10.1016/j.autcon.2010.06.007
Volk, R., Stengel, J. & Schultmann, F. 2014. Building Information Modeling (BIM) for
existing buildings—Literature review and future needs. Automation in construction, 38,
109-127. https://doi.org/10.1016/j.autcon.2013.10.023
*Wang, C., Cho, Y. K. & Kim, C. 2015. Automatic BIM component extraction from point
clouds of existing buildings for sustainability applications. Automation in Construction,
56, 1-13. https://doi.org/10.1016/j.autcon.2015.04.001
*Wittner, Y., 2010. A GIS-Based Building Data Model Storing, Visualizing, and Analyzing
Building Information for United States Coast Guard and Environment Systems Research
Institute. PhD Thesis, University of Redlands, University of Redlands, Redlands.
10.26716/redlands/master/2010.23
*Wang, M., Deng, Y., Won, J. & Cheng, J. C., 2019. An integrated underground utility
management and decision support based on BIM and GIS. Automation in Construction,
107, 102931. https://doi.org/10.1016/j.autcon.2019.102931
Wang, W., Yang, N., Zhang, Y., Wang, F., Cao, T. & Eklund, P., 2016. A review of road
extraction from remote sensing images. Journal of traffic and transportation engineering
(english edition), 3(3), 271-282. 10.1016/j.jtte.2016.05.005
*Xu, X., Ding, L., Luo, H. & Ma, L., 2014. From building information modeling to city
information modelling. Journal of information technology in construction, 19, 292–307.
https://www.itcon.org/2014/17
*Xue, F., Chen, K., Lu, W. & Zetkulic, A., 2019a. From Semantic Segmentation to Semantic
Registration: Derivative-Free Optimization–Based Approach for Automatic Generation of
Semantically Rich As-Built Building Information Models from 3D Point Clouds. Journal
of
Computing
in
Civil
Engineering,
33(4),
04019024.
https://doi.org/10.1061/(ASCE)CP.1943-5487.0000839
*Xue, F., Lu, W. & Chen, K., 2018. Automatic generation of semantically rich as‐built
Building Information Models using 2D images: A Derivative‐Free Optimization approach.
Computer‐Aided Civil and Infrastructure Engineering, 33(11), 926–942.
https://doi.org/10.1111/mice.12378

26

829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859

*Xue, F., Lu, W., Chen, K. & Webster, C.J., 2019b. BIM reconstruction from 3D point clouds:
A semantic registration approach based on multimodal optimization and architectural
design knowledge. Advanced Engineering Informatics, 42, Advanced Engineering
Informatics, 42, 100965. https://doi.org/10.1016/j.aei.2019.100965
Xue, F. & Lu, W., 2020. A semantic differential transaction approach to minimizing
information redundancy for BIM and blockchain integration, Automation in Construction,
in press. https://doi.org/10.1016/j.autcon.2020.103270
*Xiong, X., Adan, A., Akinci, B. & Huber, D., 2013. Automatic creation of semantically rich
3D building models from laser scanner data. Automation in construction, 31, 325-337.
https://doi.org/10.1016/j.autcon.2012.10.006
Xue, F., Lu, W., Chen, Z. & Webster, C. J., 2020. From LiDAR point cloud towards digital
twin city: Clustering city objects based on Gestalt principles. ISPRS Journal of
Photogrammetry
and
Remote
Sensing,
167,
418-431.
https://doi.org/10.1016/j.isprsjprs.2020.07.020
*Zeibak-Shini, R., Sacks, R., Ma, L. & Filin, S., 2016. Towards generation of as-damaged BIM
models using laser-scanning and as-built BIM: First estimate of as-damaged locations of
reinforced concrete frame members in masonry infill structures. Advanced Engineering
Informatics, 30(3), 312-326. https://doi.org/10.1016/j.aei.2016.04.001
*Zhao, L., Liu, Z. & Mbachu, J., 2019. Highway alignment optimization: An integrated BIM
and GIS approach. ISPRS International Journal of Geo-Information, 8(4), 172.
https://doi.org/10.3390/ijgi8040172
*Zhang, J., Yu, F., Li, D. & Hu, Z., 2014. Development and Implementation of an Industry
Foundation Classes‐Based Graphic Information Model for Virtual Construction.
Computer‐Aided
Civil
and
Infrastructure
Engineering,
29(1),
60-74.
https://doi.org/10.1111/j.1467-8667.2012.00800.x
Zolanvari, S. M., Ruano, S., Rana, A., Cummins, A., da Silva, R. E., Rahbar, M. & Smolic, A.,
2019. DublinCity: Annotated LiDAR Point Cloud and its Applications. arXiv preprint
arXiv, 1909.03613. https://arxiv.org/abs/1909.03613
Zhiliang, M., Zhenhua, W., Wu, S. & Zhe, L., 2011. Application and extension of the IFC
standard in construction cost estimating for tendering in China. Automation in
Construction, 20(2), 196-204. https://doi.org/10.1016/j.autcon.2010.09.017

27