You are on page 1of 28

J Intell Robot Syst (2015) 79:549–576

DOI 10.1007/s10846-014-0120-z

A BIM-based Code Compliance Checking Process of Deep


Foundation Construction Plans
Hanbin Luo · Peisong Gong

Received: 29 December 2013 / Accepted: 11 June 2014 / Published online: 4 October 2014
© Springer Science+Business Media Dordrecht 2014

Abstract Deep foundation construction projects are and facilitates the risk identification at the next stage.
quite risky, so it has to be checked by supervisors It ensures the safety of deep foundation and reduces
for code compliance and reviewed by a committee personal injury and property loss.
of experts for risk identification before the construc-
tion begins. The quality of code compliance checking Keywords Code compliance checking · Rule-based
directly affects the safety of deep foundation construc- checking · Deep foundation · BIM
tion. Because of the limitations of manual checking by
even the best supervisors, the goal of this research is to
achieve an approach of BIM-based code compliance 1 Introduction
checking for deep foundation construction presented
in BIM models, as well as a system based on the pro- Deep foundation construction is becoming more and
posed approach. This research abides by the principles more popular owing to the narrow ground space dur-
of four stages in rule-based system, and emphasizes ing the process of urbanization. Deep foundation con-
the development of the library of checking knowledge struction has some unique characteristics including
and the standard of required information which are large investments, long project cycles, and compli-
the key problems in achieving the BIM-based code cated techniques etc. [1]. There are numerous and
compliance checking. Decision tables, production rep- unpredictable risk factors because of the character-
resentation and a developing process similar to IDM istics. As a result, more accidents occur and some
development are used to solve the problems. Further- serious accidents have severe influences on the society
more, a case study is provided to investigate the effect and environment [2–7]. Frijters and Swuste indicated
of the BIM-based code compliance checking process. that the safety aspect of design is an important factor
It is proved that the BIM-based checking improves the which will influence the risk in the next stage [8].
efficiency and precision in code compliance checking Because of the safety risks of deep foundation
construction, national and local governments have
established systems to restrain the activities of deep
H. Luo · P. Gong () foundation design and construction so as to reduce the
School of Civil Engineering and Mechanics,
Huazhong University of Science and Technology,
accident occurrence. A deep foundation construction
Wuhan, 430074, China project, a project with complex geological conditions
e-mail: taller1975@163.com or a project in a complex environment must follow
H. Luo special laws, regulations, codes and procedures. In
e-mail: lhblhb1963@vip.sina.com this sense, deep foundation construction requires that
550 J Intell Robot Syst (2015) 79:549–576

special construction plans be formulated individually An important review of rule-based checking system
by contractors before construction begins. In addition, about building design is provided by Eastman in
the position of design supervision should be held by 2009 [13].The industry is in need of improving the
professional experts, who check and review the spe- efficiencies of the existing manual code checking pro-
cialized construction plans and then sign off, thus cesses, and further improvements can be gained in
completing the process of code compliance checking. code compliance checking through the use of informa-
The control of code compliance at the design stage tion technology.
is an important measure to guarantee safety in the The AEC/FM industry has been changed greatly
construction stage. by the implementation of building information mod-
Seemingly, the risks of deep foundation construc- eling (BIM) [14], and code compliance checking has
tion have been under effective control because of these no exception. The application of information technol-
measures, but it is not the case. For example, despite ogy in code compliance checking has initiated over
the expertise of the individuals involved in a manual two decades. The significant contemporary efforts
review, this type of review has inherent flaws. Accord- appeared in 1995. The real implementing of design
ing to the work rule found in the ‘management method review system on rule checking was initiated in 1995
for high risk division components’ [9], proper review by Singapore which was based on 2D drawings and
of such complex deep foundation projects involves the later efforts led to the CORENET System on
two steps. First, the specialized construction plan is the basis of IFC building models in 1998 [15]. The
reviewed to check its compliance to relevant con- objective of CORENET is to re-engineer the busi-
struction codes by a design supervisor committee, and ness processes of the construction industry to achieve
any exceptions should be noted. Second, the result a quantum leap in turnaround time, productivity and
of step one will be transferred to an expert commit- quality. CORENET system’s users covered more than
tee whose task is risk identification. The risks in the 2500 firms which included architects, engineers, sur-
project are identified and evaluated based on relevant veyors and other professionals [13]. It releases tem-
safety standards. Meanwhile, the hydrogeology and plates and guidelines periodically of different BIM
surrounding environment are also considered by the software such as Revit, Archicad, Bentley, and has
expert committee. So code compliance is the basis for 23 collaboration partners at least until 2014 which
risk identification in the design stage and the quality include professional bodies, service providers and
of code compliance is significant to the expert com- government agencies [16]. Recent applications are
mittee. Nevertheless, code compliance checking is a presented by SMART codes. SMART codes were
time-consuming job because a large number of details supported by the International Code Council(ICC) in
in the design documents require manual calculations 2006 and it can transfer written building codes to
and judgment for an effective review. In fact, code computer-interpretable code sets [17]. ICC SMART
compliance checking has become a mere formality for codes are used as a rule set by the model checking soft-
the reasons mentioned above. The results give less ware(MCS) and the MCS returns a compliance report
support for the next stage of risk identification. The and visual ID of non-compliance to the designer,
human expert does not have enough information to who then revises the design, checking until compli-
evaluate risks properly in such a complex project. ance has been achieved. ICC has initiated work on
The purpose of code compliance checking is to many ICodes, including the international building,
answer whether all the details of the specialized con- Fire, building safety(building safety month 2014) etc..
struction plan of deep foundation construction is in Until now, The ICC has developed and made avail-
accordance with relevant articles in the proper codes. able an impressive inventory of comprehensive and
Because code compliance is a mechanical work with a coordinated International Codes(15 codes),including
large workload, this work is time consuming and error- International Building Code, International Green Con-
prone, which is the main reason for the slack in prac- struction Code, etc. [18]. All these efforts give a solid
tical work. Furthermore, checking in relevant domains foundation to the emergence of larger, more industrial-
such as safety [10], construction quality inspection based efforts in the future. Although there have been
and control [11], building envelope design [12] and so great efforts for code compliance checking, most of
on also have difficulties in efficiency and precision. them are constrained in one or several domains such
J Intell Robot Syst (2015) 79:549–576 551

as estimating spatial validation, fire control, circula- knowledge such as status, attribute and concept,
tion checking, structural design, energy consumption while Ontology-based semantic modeling spe-
simulation, building envelope design and early cost cializes in expressing domain knowledge such as
estimate etc. Deep foundation has been ignored and domain concept, their relationship and classifica-
the possible reason is that it is a temporary product tion. The purpose of code compliance checking
and not a part of final buildings. But because of its is to judge whether a design parameter con-
high risk and the difficulties mentioned in the previ- forms to the relevant code. Its essence is to
ous paragraph, code compliance checking is an urgent say yes or no. So neither predicate logic nor
need for practical work. Ontology-based semantic modeling is a proper
The paper is intended to make use of BIM model way to do rule interpretation in code compliance
and, with the construction plans presented in BIM checking.
model, propose a proper process for efficient and pre- Decision table is a graphic tool suitable for
cise code compliance checking in deep foundation disposing decision problems in the case that
domain. there are many premises and these premises can
In essence, code compliance checking is a rule- combine with each other. So it is a proper tool
based checking. The rule checking process was pro- to structure codes written in natural language.
posed by C.M. Eastman [13] and is composed of four On the other hand, production representation
major stages: is used to express causal relationship between
objects and suitable for rule-based knowledge.
(1) Rule interpretation stage: codes are definewritten Fenves [24] initially structured codes by decision
in natural language. In the domain of civil engi- table, and decision tree were used to structure
neering, including architecture, structure, dec- steel design later [25]. He developed a system
oration, etc., for a long time, the quality of to manage codes [26] and provided an important
a project is controlled by various codes from review of these early efforts [27]. The SASE sys-
different systems, which are sometimes contra- tem‘[28] proposed a comprehensive structure to
dictory and imprecise. The practical application classify relevant codes, and REGNET [29] devel-
of these codes is time-consuming and ineffi- oped by Kerrigan can estimate the applicability
cient. Therefore, it’s necessary to re-structure of various codes for a certain design. The early
these codes in order to improve the efficiency work provided a rule-based approach to structure
of code checking in civil engineering. Thus, the codes. But how to avoid too large rule quantity
first step of BIM-based checking is to translate in the database is a critical difficulty in the use of
code clauses into data representation which can decision table.
be understood by computer. (2) Model preparation stage: A BIM model con-
The current methods of rule interpretation tains a large amount of information, but which
are predicate logic and Ontology-based seman- information is needed for a specific checking
tic modeling [13]. The initial efforts applying purpose? In order to specify the information
predicate logic were undertaken by D. Jain needed to certain code compliance checking,
et al. [19] and W.J. Rasdorf et al. [20] An some model view definitions [30] must be used.
ontology reasoning mechanism was employed Model view definition can derive the needed data
to detect conflicts between diverging partici- required for a specific type of code compliance
pants’ requirements in collaborative design sce- checking and extract subsets of a BIM model to
narios by Kim and Grobler [21] and Pauwels do a more efficient checking [31, 32]. Informa-
et al. [22] researched an approach to establish tion delivery manual is the basis of model view
a semantic rule checking environment for build- definition, so it is a critical factor to the model
ing performance checking. Zhong [23] devel- preparation stage.
oped an Ontology-based semantic modeling of (3) Rule execution stage: This stage is a pro-
regulation constraint for automated construc- cess of carrying out code compliance checking
tion quality compliance checking. But pred- against candidate BIM models. It includes two
icate logic is suitable for expressing factual processes of checking, i.e. preprocessing and
552 J Intell Robot Syst (2015) 79:549–576

post-processing. The key problem at this stage is Rules stored in rule database are constructed
the design of checking algorithm. for any deep foundation. In the case of a cer-
(4) Checking results reporting stage: Some neces- tain deep foundation, only a part of the rules is
sary information must be given at this stage such required to do rule execution. The list of required
as checking points, checking results, correspond- rules and the list of required extracted informa-
ing reference clauses and treatment measures. tion are the necessary premise of certain code
compliance checking. Semantic modeling spe-
A lot of domains have been involved in the
cializes in expressing domain knowledge. So an
BIM-based code compliance checking such as spa-
engineering semantic structure of deep founda-
tial checking [33], circulation and security checking,
tion is proposed to relate rule database and it
energy consumption simulation and early cost esti-
can rapidly define checking objects and check-
mate [34]. But deep foundation is not involved. It is
ing points. On the basis of the semantic structure,
worth noting that the model of four stages proposed
for a specific deep foundation project, once the
by C.M. Eastman is only a framework of the rule-
engineering semantic information is determined,
based system. In fact, some difficulties emerged only
the list of required rules and the list of required
in a certain domain have to be overcome during the
extracted information can be defined. Namely
process of code compliance checking in this domain.
all the information which needs to be extracted
These difficulties can be concluded as a series of
in BIM model has been defined. The engineer-
problems presented in Fig. 1.
ing semantic structure helps to reduce calculation
These problems can be solved differently accord-
times.
ing to the characteristics of different domains. The
(3) Propose a method to ensure the effective exis-
main purpose of this paper is to explore a BIM-based
tence of required information.
code compliance checking process and the domain is
Deep foundation construction plans are made
restricted to deep foundation. A system based on this
for construction guidance, not for code compli-
process will be illustrated to demonstrate its effect.
ance checking. It is possible that some required
The critical contributions of this research are as fol-
information has not been constructed in the pro-
lows:
cess of code compliance checking. So an infor-
(1) Propose a method to do the rule interpretation. mation standard of deep foundation for code
Decision table and production representation compliance checking is a necessary premise to
are introduced to undertake rule interpretation ensure the effective existence of required infor-
and a method based on comprehensive assess- mation. Information delivery manual(IDM) and
ment is used to simplify rules. model view definition(MVD)[35] are effective
(2) Propose a method to rapidly define needed infor- tools to define needed information in BIM model
mation. for a certain stage or application purpose. But the

Fig. 1 The research need of BIM-based code compliance checking


J Intell Robot Syst (2015) 79:549–576 553

development of IDM has not involved deep foun- illustrated in Fig. 2 in BPMN format. The components
dation. In this paper, an information standard of in red rectangle belong to code compliance check-
deep foundation construction for code compli- ing process. This figure contains both traditional and
ance checking is proposed to ensure the effective BIM-based processes. The data object [D.1] is writ-
existence of required information. It removes the ten codes which are used in traditional process and it’s
ambiguous that whether required information is replaced by the data object [D.2] in BIM-based pro-
not extracted properly or has not been estab- cess. The data object [D.3] is the standard of required
lished at all when it cannot be extracted from information which ensures information validity in the
the BIM model. Once the BIM model is struc- BIM model.
tured by the standard, it means that the required On the basis of the information provided in Fig. 1,
information for code compliance checking has there are four objectives which must be accomplished
been established. The development of informa- in BIM-based checking process: (1) Codes in natu-
tion standard of deep foundation construction for ral language must be transformed into digital format.
code compliance checking is beneficial to the (2) The checking list for supervisors must be gener-
IDM development in this field. ated rapidly. (3)The required information in checking
This research presents a BIM-based code list must be established in the candidate BIM model
compliance checking process and a rule-based and extracted effectively. (4) The judgment must be
checking system of deep foundation construction achieved according to the relevant codes which are
plans presented in BIM model. This paper will presented in digital rules. With all of the above-
focus on code compliance checking related to mentioned objectives accomplished, the design super-
deep foundation construction. visors can give an explicit correction proposal while
The framework and methodology of the pro- the experienced experts have enough information to
posed code compliance checking process are analyze the potential risks in the project.
presented in Section 2. The development of Therefore, the three key technologies to be solved
the proposed code compliance checking process in the code compliance checking of deep foundation
addressed in Section 3 includes rule interpre- are as follows:
tation, engineering semantic structure of deep (1) The transformation of the codes in natural lan-
foundation and information standard of deep guage into rules;
foundation for code compliance checking etc.. (2) The classification of these rules by an engi-
A case study is presented in Section 4. The neering semantic structure for the convenience
results demonstrate the benefits of the code com- of quick achievement of the checking list or
pliance checking process, for example, assisting checking points;
experts and designers in checking the efficiency (3) The standard of required information for code
of deep foundation construction plans and reduc- compliance checking ensures the information
ing errors in the design of the plans. A summary used in BIM-based code compliance checking of
of the contributions and discussions about future deep foundation construction is in the candidate
research are in the final section. plan;
Based on the above information, the primary design
of the checking process is presented in Fig. 3.
2 Framework and Methodology
(1) The newly established project for deep founda-
2.1 The Definition of Requirement for Code tion code compliance checking, which includes
Compliance Checking the project name, the owner, the contractor, the
supervisor, the designer, and the project outline,
In this section, the requirements for BIM-based code etc.
compliance checking in deep foundation construction (2) Loading the candidate construction plan and
will be defined based on its features and the anal- storing it for future use.
ysis of the process of manual checking. A typical (3) Operators input the engineering sematic infor-
management of deep foundation construction plan is mation by answering a series of questions(less
554 J Intell Robot Syst (2015) 79:549–576

Fig. 2 Traditional and BIM-based process of code compliance checking

than 10)according to the design content of the putting them into the corresponding positions of
candidate deep foundation construction plan. the checking list in accordance with all required
(4) Based on the engineering sematic information engineering parameters in the checking list.
built in the previous step, the checking list is gen- (7) Inputting manually. The engineering parame-
erated which defines all the checking points of ter information coming from BIM models does
the candidate plan and the corresponding check- not meet all the information requirement of the
ing rules to the checking points. checking list because the candidate plan is made
(5) Entering into the reasoning area and starting to for deep foundation construction guidance, not
undertake the information preparation in accor- for the code compliance checking. In addition,
dance with the checking list. there is still no legal BIM building standard
(6) Extracting all needed engineering parameters for code compliance checking in deep founda-
from the candidate plan presented in BIM and tion field. A research on the standard of BIM
J Intell Robot Syst (2015) 79:549–576 555

2.2 Framework of the Code Compliance Checking


Process

The proposed framework of the code compliance


checking process is illustrated in Fig. 4.
The four stages of the code compliance checking
process are embedded in this framework. The first step
is rule interpretation and the eventual carrier is the
library of checking knowledge. The function of the
library of checking knowledge is to store the check-
ing rules according to a certain classification; a set
of proper rules will be extracted according to the
checking list generated by a definition of engineer-
ing semantic structure mentioned in the following, and
used to check corresponding checking points of the
checking list. Rule interpretation will be illustrated in
the next section.
The second step is model preparation, ensuring
the effective existence of required information which
can be extracted to do rule execution. There are two
approaches to achieve required information for code
compliance checking [36].
One approach is based on a software
application/plug-in, that is to say users can execute
code compliance checking process in the design pro-
cess and can use the same software despite it is a
redeveloped software. Currently available BIM design
tools provide some model checking functions, but
Fig. 3 The primary design of the checking process they do not satisfy most of the domain requirements
in code compliance checking. The advantage of this
model for code compliance will be illustrated approach is that the checking function module can be
in this paper but it is not a legal standard. So developed on the BIM platform and it is quite easy
the engineering parameter information extracted for developers and users to be developed and used
from BIM models can’t meet all the informa- because of the direct invoking of the API of the BIM
tion requirements of the checking list and part platform. On the other hand, the limit of API also
of the information inputting has to be done man- constrains the freedom of checking function design,
ually. The manually inputting interface must be so some complex requirements can’t be satisfied.
designed in this process. The other approach is to use an Industry Foun-
(8) According to the rule coding from the checking dation Class (IFC)-based model viewer or checker
points in the list, the corresponding rules of the system for the implementation. It’s worth noting that
checking knowledge database will be selected a primary obstacle for the code compliance check-
and are used to implement the checking based on ing is that current BIM softwares have their own data
the actual data extracted from the candidate plan format model conventionally and the Application Pro-
(BIM model). gramming Interface(API) has limited the functions
(9) The feedback of the final checking results in the of code compliance checking. Industry Foundation
form of tables includes all the checking points Classes(IFC)[37] published by the building SMART
and the information, whether the checking points is the only neutral model presentation for describing a
meet the checking rules as well as all the corre- building and is an independent neutral data model pre-
sponding articles of the original codes. sentation supported by most BIM design tools [13].
556 J Intell Robot Syst (2015) 79:549–576

Fig. 4 The framework of the code compliance checking process

A Model Checker System to check industry founda- out a method to simplify IFC model for code com-
tion classes (IFC) product models was developed by pliance checking and develop a standard of required
Solibri Inc [38]. It can do a code compliance check- information for code compliance checking so as to
ing on a model presented by IFC. J.-K. Lee [39] ensure the existence of the required information in
researched evaluating building circulation and spatial the candidate IFC model and to reduce information
program on the basis of IFC. P. Pauwels et al[40]. redundancy.
analyzed the interoperability for the design and con- The reasoning area is the kernel of rule execu-
struction industry through IFC approach. So IFC is a tion and need some information from the library of
rich and redundant data-modeling schema and needs checking knowledge, as well as checking lists and the
be simplified for a specific implementation. candidate model in IFC format. Checking lists define
The two approaches have their own advantages and required checking points; the library of checking
defects, but IFC is more likely to be popular in the knowledge provides corresponding rules; and the can-
future with more and more support from the BIM didate deep foundation plan in IFC format gives the
providers. Comparing features of API of application practical engineering parameters. So in the reasoning
software with features of IFC in this research, we area, code compliance checking can be achieved at
use IFC format to display candidate deep foundation every checking point, namely saying yes or no.
plans in order to be unconstrained by certain appli- The reporting of checking results includes infor-
cation software. Therefore, it is necessary to work mation about all the checking points which has been
J Intell Robot Syst (2015) 79:549–576 557

checked, the checking results marked by ‘yes’ or‘no’, words, thus all the rules can be mapped to the cor-
with relevant codes listed for reference. responding components of engineering sematic struc-
ture. The checking rules are structured by means of
2.3 Methodology for the Code Compliance Checking integration of decision table and production represen-
Process tation which will be discussed in the next section.
After the achievement of rule interpretation, we
The proposed methodology of the code compliance face the problem of how to ensure the required infor-
checking is illustrated in Fig. 5. mation for the rules extracted from the candidate BIM
The developed processes of rule interpretation and model. So the standard of the required information for
model preparation are in the subsequent section. code compliance checking must be developed. It is
Their eventual carriers are the library of checking involved in the steps of BIM model building and IFC
knowledge and the candidate plan in IFC format format transformation so as to ensure the existence of
respectively. In rule interpretation, the purpose of the required information in the candidate IFC model
code analysis is to construct engineering sematic and avoid information redundancy. The developing
structure. The first step is to select relevant articles method of the standard is a proper gathering and
and classify them. But what is the principle of the classification of all the rule premises. The eventual
classification? Therefore, it is necessary to analyze carriers of candidate deep foundation plans are IFC
code content which gives the key words (the compo- format model. The invoking of API is an approach to
nents of engineering sematic structure) of classifica- extract required information from BIM models, but it
tion. The selected articles are classified by the key is limited by the API of certain BIM software. The

Fig. 5 The methodology of


code compliance checking
process
558 J Intell Robot Syst (2015) 79:549–576

developed tool to extract required information will be engineering sematic structure and the interpreted
invalid when a model is presented by other BIM soft- rules.
wares. IFC is a neutral model presentation and a key
solution of the problem of API limited by various BIM 3.1.1 The Development of the Engineering Sematic
softwares [37]. Structure
This section illustrates the requirement, framework
and methodology of code compliance checking pro- The first step of building the structure of check-
cess for deep foundation construction. The detailed ing objects is the engineering semantic analysis of
development is revealed in the next section. relevant codes’ content and the collection of rele-
vant articles. Codes content contains information of
the checking objects, and every article has one or
3 The Development of BIM-Based Code
more checking points pertaining to a certain check-
Compliance Checking of Deep Foundation
ing object. Take the cantilevered piles (shown in
Construction
Fig. 7) as an example, the cantilevered pile is a check-
ing object. Under the checking object of cantilevered
The key developing points of BIM-based code com-
piles, the relevant code articles in ‘Hubei technical
pliance checking of deep foundation construction are
specification for engineering of foundation excava-
the library of checking knowledge and the standard of
tion DB42/159-2004’ [41] can be collected to identify
required information. The developing process is pre-
several checking points.
sented in Fig. 6, and their relationship is also revealed
Therefore, at least two checking points, that is, the
in it.
diameter of the piles and the vertical thickness of the
3.1 The Development of the Library of Checking reinforced concrete top beam can be established in the
Knowledge checking object of cantilevered piles. After the engi-
neering semantic analysis on all the relevant articles,
The approach of rule interpretation development is every checking point and checking object are defined.
presented in Fig. 5, and its carrier is the library With the checking objects properly classified, an engi-
of checking knowledge. The library of checking neering sematic structure of deep foundation can be
knowledge is composed by two components i.e., the established as Table 1.

Fig. 6 The key developing points


J Intell Robot Syst (2015) 79:549–576 559

Fig. 7 The building process of checking points and checking objects

The building of the engineering sematic structure Article 10.0.4: the definition of monitoring item
in the system is presented in Fig. 8. A checking must be aligned to the importance level of the deep
list including all checking points can be generated foundation, and some monitoring items are presented
after acquiring a set of checking objects chosen man- in Table 2. The checking result of code compliance
ually according to the candidate construction plan is limited to ‘yes’ or ‘no’ only, so we define items as
because the relationship between checking objects and necessary items, conditionally an applicable items and
checking points has built in the engineering semantic unnecessary items.
structure. Article 4.0.1: deep foundation can be classified in
three importance levels based on Table 3.
3.1.2 The Development of Rule Interpretation After two articles have been analyzed comprehen-
sively, two steps are needed to define the monitoring
Another building work of the library of checking items, and one key step is to define importance level
knowledge is rule interpretation. Every article must be of the deep foundation. So the eventual rules for this
translated into a specific pattern in order to facilitate checking object are presented in a format of rule chain.
the programming. Predicate logic and Ontology-based The premise of Table 2 is relatively simple and directly
semantic modeling are the main methods of rule inter- translated into production representation, and is shown
pretation [13]. But the logic of this code compliance as follows:
checking is not suitable for these two tools, and some Checking point: the monitoring point of stress con-
ordinary tools can be used to accomplish this work. dition of supporting structure
In this paper, decision Table [24] and production rep-
resentation are used to translate rules. Decision tables If: importance level(variable)= 1
are used to structure logic of articles, and the produc- Then: the monitoring of stress condition of sup-
tion representation is employed as the expression form porting structure=necessary
of the rules. If: importance level(variable)= 2 or 3
An example is used to illustrate the developing pro- Then: the monitoring of stress condition of sup-
cess of rule interpretation. The logically related two porting structure=unnecessary
articles compose a checking rule towards the check-
ing point of ‘monitoring item’ in ‘Hubei technical The ‘importance level’ is a variable whose defini-
specification for engineering of foundation excavation tion is dependent on other rules built from Table 3.
DB42/159-2004’ [40]. So these rules are the premises of checking rules from
560 J Intell Robot Syst (2015) 79:549–576

Table 1 The engineering sematic structure of deep foundation

Project namec Checking objects(option) Checking points

Deep Supporting Cantilevered pile Piles diameter


foundation The vertical
project of XX thickness of
building reinforced concrete
top beam
...
Steel sheet pile ...
Underground diaphragm wall ...
Tube-round support ...
Cement barricade ...
Slope Self-stabilization side-slope ...
Combined bolting and shotcrete slope ...
Soil nailing slope ...
Brace Concrete brace ...
Steel brace ...
Management Drain off water of open ...
of ground trench and covered trench
water Well point dewatering ...
Tube well dewatering ...
The vertical impervious
cement injection pile ...
The vertical impervious
jet grouting pile ...
High pressure jet grouting ...
Deep mixing construction ...
Earthwork Earthwork backfill ...
backfill ...
Monitor Monitoring item ...
Laying of monitoring points ...
Monitoring frequency ...
Monitoring warning ...
... ... ...

Table 2, and all of the above rules compose a rule To simplify these original rules, different combi-
chain. nations of value choices of these causes and effect
The logic of Table 3 is relatively more complex are merged. At least nineteen rules are still left,
and it must be analyzed through decision table. There and the logic of each rule is still relatively complex
are three causes and one effect in Article 4.0.1. Two which normally includes three premises in the produc-
of the causes have three value choices and the other tion representation format. Therefore, the simplified
has four. The effect is the importance level of deep result is not beneficial to reducing information redun-
foundation and it has three value choices which are dancy. How to simplify these original rules and greatly
level 1, level 2 and level 3. The above-mentioned value reduce the number of remaining rules? The essence
choices compose thirty-six original rules presented in of this problem is to find a way to map a three-
Table 4. dimensional vector to a one-dimensional vector whose
J Intell Robot Syst (2015) 79:549–576 561

Fig. 8 The building process of engineering sematic structure in the system

value can help to distinguish the three importance At first, we hypothesize all the definitions in
levels. This requirement conforms to the concept of Table 3 are consistent and the weights of the three
comprehensive assessment. causes are equal. The process of the method can be
A method is developed to do the mapping based concluded as follows:
on the process of comprehensive assessment, and a
representative process of comprehensive assessment is (1) Setting score for every value of the three causes
concluded in Fig. 9 [42]. and ensuring that higher score means lower
According to the information of the decision table importance level of deep foundation. The main
presented in Table 4, it is unnecessary to create assess- purpose of this step is to quantify these qual-
ment indexes and their values, since they have been itative indicators in order to make the analysis
determined already. easier.
562 J Intell Robot Syst (2015) 79:549–576

Table 2 The definition of monitoring item

Monitoring item Importance level

Level 1 Level 2 Level 3

Horizontal displacement of the top of supporting and slope N N N


Deformation monitoring of underground facilities N N C
Heaving of the foundation bottom C C U
Monitoring of stress condition of supporting structure N C C
... ... ... ...

Annotation: N–Necessary items


C–Conditionally applicable items
U–Unnecessary items

(2) Building an aggregative function to translate According to the evaluated values of these rules,
three-dimensional assessment indexes into an there is a high degree of satisfaction of consistency
evaluating value. to the importance level of deep foundation. The range
(3) Ranking all the original rules again and checking of the evaluated value at importance level 1 is lim-
all the evaluating values so as to find appropriate ited to less than 500, and importance level 2 is
margins for the three levels. limited to 500 to 1000, and importance level 3 is
above 1000. There are three exceptions emphasized
The LCM(least common multiple)of the value in Table 6, and they are located respectively at the
choices of the three causes are chosen to be the high- edge of all three importance levels in Table 3. These
est score(12) in order to ensure a consistent score and three exceptions is attributed to the inconsistency of
equal weight for the three causes. The score of every importance level definition operated manually namely
value is presented in Table 5. excessively emphasizing a certain cause in these three
Accumulated multiply is chosen to do the aggrega- conditions.
tive function in order to enlarge the differences among Four rules remain by simplifying Tables 3 and
the evaluated values. The computing results have been 4. The eventual rule chain of the checking point
re-ranked as presented in Table 6. ‘the monitoring point of supporting structure stress

Table 3 The definition of importance levels of deep foundation

Environment and geological condition

Excavation deepness H(m) a<H H ≤ a≤2H a > 2H

I II III I II III I II III

H>15 1 1 1 1 1 1 1 1 1
10<H ≤ 15 1 1 1 1 1 2 1 2 2
7<H ≤ 10 1 1 2 1 2 2 1 2 3
H≤7 1 2 2 1 2 3 2 3 3

Annotation: a—the distance of trunk road and near building to the edge of foundation pit
The classification of engineering geological condition:
I complexity—soil of poor quality and abundant groundwater
II less complexity—soil of less poor quality and groundwater
III simpleness—soil of good quality and no groundwater
J Intell Robot Syst (2015) 79:549–576 563

Table 4 The original decision table

SN Causes Effect

Excavation Environment and Excavation The importance level of deep


deepness geological condition deepness H(m) foundation engineering

1 a<H I A 1
2 a<H I B 1
3 a<H I C 1
4 a<H I D 1
5 a<H II A 1
6 a<H II B 1
7 a<H II C 1
8 a<H II D 2
9 a<H III A 1
10 a<H III B 1
11 a<H III C 2
12 a<H III D 3
13 H<a≤2H I A 1
14 H<a≤2H I B 1
15 H<a≤2H I C 1
16 H<a≤2H I D 1
17 H<a≤2H II A 1
18 H<a≤2H II B 1
19 H<a≤2H II C 2
20 H<a≤2H II D 2
21 H<a≤2H III A 1
22 H<a≤2H III B 2
23 H<a≤2H III C 2
24 H<a≤2H III D 3
25 a>2H I A 1
26 a>2H I B 1
27 a>2H I C 1
28 a>2H I D 2
29 a>2H II A 1
30 a>2H II B 2
31 a>2H II C 2
32 a>2H II D 3
33 a>2H III A 1
34 a>2H III B 2
35 a>2H III C 3
36 a>2H III D 3

condition’ can be presented in the format of produc- particularly simplifying the tools because of the low
tion representation as shown in Fig. 10. It is worth logic complexity.
noting that the logic of this instance is relatively com- An instance of premise and rule building process in
plex, and the majority of the rules can be built without the system is presented in Fig. 11. This rule belongs
564 J Intell Robot Syst (2015) 79:549–576

Fig. 9 A representative
process of comprehensive
assessment

to the checking point of ‘the spacing of exploratory the basis of [D.4]. The original information includes
points’. geographical location, site and building geometric out-
lines, building floor identifier, etc., which facilitates
3.2 The Development of the Standard of Required the design of deep foundation. The original and new
Information information built during the design of deep founda-
tion composes the standard of required information.
The main purpose of the standard of required informa- So the development of this standard is related to
tion is to ensure the information required for checking IDM.
is prepared. Deep foundation is a relatively inde- BuildingSMART released roadmap of IDM which
pendent work, which is only a process product in is presented in matrix [43]. ISO 29481-1 applied
the whole building process. The standard of required ‘Generic Design and Construction Process Protocol’
information for its BIM model has not been built, formulated by Salford university to defining IDM
which has become a main obstacle for the popular- roadmap [44]. BuildingSMART lists 44 projects of
ity of BIM-based code compliance of deep founda- IDM in building life cycle and 9 of them are on
tion. The development of the BIM model standard of the list of priority. Twenty projects of them have
deep foundation ensures that the required information made some progress already. BuildingSMART wants
for code compliance has been built in the candidate to confirm some extremely important data deliv-
model. ery scenes supported by IDM in building life cycle,
The design of deep foundation follows the design and facilitate the development of preferential projects
of architecture and structure. The building of BIM so as to attract sponsors’ support for the execution
model for deep foundation needs some original infor- of IDM. The development of the required informa-
mation obtained from the architecture and structure tion standard of deep foundation is beneficial to the
model which is in [D.4] of Fig. 2, and a lot of new development of IDM in the area of deep foundation
information about deep foundation must be built on construction.

Table 5 The score of every value choice

Causes

Excavation deepness Environment and geological condition Excavation deepness H(m)

a<H=4 I=4 A=3


H < a ≤ 2H = 8 II = 8 B=6
a > 2H = 12 III = 12 C=9
D = 12

Annotation: A = H > 15; B = 10 < H ≤ 15; C = 7 < H ≤ 10; D = H ≤ 7


J Intell Robot Syst (2015) 79:549–576 565

Table 6 The final re-ranked decision table

Causes Effect

SN Excavation Environment and Excavation The importance level of deep Evaluated


deepness geological condition deepness H(m) foundation engineering value

1 4 4 3 1 48
2 4 4 6 1 96
3 4 4 9 1 144
4 4 4 12 1 192
5 4 8 3 1 96
6 4 8 6 1 192
7 4 8 9 1 288
9 4 12 3 1 144
10 4 12 6 1 288
13 8 4 3 1 96
14 8 4 6 1 192
15 8 4 9 1 288
16 8 4 12 1 384
17 8 8 3 1 192
18 8 8 6 1 384
21 8 12 3 1 288
25 12 4 3 1 144
26 12 4 6 1 288
27 12 4 9 1 432
29 12 8 3 1 288
33 12 12 3 1 432
8 4 8 12 2 384
11 4 12 9 2 432
19 8 8 9 2 576
20 8 8 12 2 768
22 8 12 6 2 576
23 8 12 9 2 864
28 12 4 12 2 576
30 12 8 6 2 576
31 12 8 9 2 864
34 12 12 6 2 864
12 4 12 12 3 576
24 8 12 12 3 1152
32 12 8 12 3 1152
35 12 12 9 3 1296
36 12 12 12 3 1728

Actually, all the premises of rules contain all the properly according to the components of deep founda-
required information, so we can define all the required tion. The essence of this structuring is to find proper
information from the premises of rules. Meanwhile, carriers for this required information, and it must be
another key work is to organize this information beneficial to the work of model building. Most carriers
566 J Intell Robot Syst (2015) 79:549–576

Fig. 10 The rule chain of


If a H Data
the checking point ‘the Rule 4 Value assignment according to
and prepareation Tab.5
monitoring point of Environment geological condition No
If evaluated value 500
supporting structure stress and
Then Grade=1
H 7
condition’ Rule 1 Rule 5
If 500 evaluated value 1000
Then grade 2
Then Grade=2
Yes If evaluated value 1000
Then Grade=3
Premise
Output
If a H checking
If grade 1
and No Then checking point=need
Environment geological condition
If grade 1
and
Then checking point no need
7 H 10 Rule 2
Then grade 2

Premise
If a H and Environment checking Yes
and
No
Environment geological condition
and
H 7 Rule 3
Then grade 3

Premise
checking Yes
No

are built in the design stage of deep foundation, while is presented in Tables 8 and 9. Delivery requirement
the rest of the carriers are in the design of architecture of the structure design is similar to that of the architec-
and structure. ture design. It is worth noting that the BIM model is
The development process of the required infor- delivered from architecture design, structure design
mation standard of deep foundation is presented as to deep foundation design, and finally code com-
follows: pliance checking on deep foundation design is con-
ducted by the supervisors. This process is presented in
(1) Collecting all the required information from the
Fig. 1.
premises of rules.
The information displaced in Table 8 is necessary
(2) Defining the required basic information in BIM
for the design of deep foundation. Other information
model of architecture and structure for the design
such as building components is not needed.
of deep foundation.
The designer of deep foundation adopts some basic
(3) Defining the required information for new build-
information obtained from the design stages of archi-
ing components at the design stage of deep
tecture and structure according to Table 8, and then
foundation.
builds the BIM model under the guidance of required
The collection process from premises of rules is information presented in Table 9, which ensures that
presented in Table 7. the required information for code compliance is built
The essence of the work in step 2 and step 3 in the candidate BIM model. The BIM model should
is to define the requirement for information deliv- be transformed to IFC format, so the developed
ery at different design stage, and label the origin of extracting information module of the system is valid
information. Cheng Zhou [45] proposed a template for models built by different BIM application soft-
of delivery requirement definition for IDM develop- ware. The standard of required information for deep
ment. It is improved, in this paper, to present the foundation proposed in this section can be a foun-
required information. Some required information from dation for the IDM and MVD development of deep
the design stages of architecture and deep foundation foundation.
J Intell Robot Syst (2015) 79:549–576 567

The building of rule premises

Name of premises Name of checking points values operators


Requirements of spacing of exploratory points 1 spacing of exploratory points 12.0, 14.0 [ ]
Requirements of spacing of exploratory points 2 spacing of exploratory points 14.0

The building of checking rules

Name of premises Name of checking rules Rule ID Checking results


Requirements of spacing of exploratory points 1 spacing of exploratory points 00001 pass
Requirements of spacing of exploratory points 2 spacing of exploratory points 00002 deny

The building of checking rules

If spacing of exploratory points [12.0 , 14.0]


The rule of spacing of exploratory
point presented in production Then pass
representation If spacing of exploratory points 12.0 or 14.0
Then deny

Fig. 11 The building process of premises and rules in the system

3.3 The Development of Rule Execution The preliminary result of code checking in the sys-
and Reporting tem page is presented in Fig. 12. The candidate model
is displayed at the upper part of the page, and the
The rule execution is to decide whether the rele- relevant code articles are displaced at the bottom when
vant checking rules are executed at a certain checking users click a certain component on the top.
point. A lot of checking points have one-step rules and A final result of checking will be generated
some have rule chain as the instance presented in 3.2. in a table format which includes checking points,
568 J Intell Robot Syst (2015) 79:549–576

Table 7 The collection process from premises of rules

Checking Checking Rule No premise Required Building


objects points information stage

Cantilevered Pile diameter Rule1 Pile diamete ∈ Pile diameter Deep


pile [400,1200] foundation
Rule2 Pile diamete engineering
< 400mm
Rule3 Pile diamete
> 1200mm
... ... ... ... ...
monitoring the existence Rule1 The classification The classification Structure
items of monitoring of engineering of engineering
point for geological geological
stress condition condition=II condition
of supporting ... ...
structure
... ... ... ... ... ...

engineering parameters (extracted from model), name underground diaphragm wall and Cantilevered pile
of checking rules, corresponding code articles, and as the foundation excavation support. Moreover, the
checking conclusions. underground diaphragm can be taken as the outer
wall of the basement structure. The underground
diaphragm wall is divided into 165 sections with a
total length of 1067m. The brace system of Section 1
4 Case Study
adopts a horizontal brace system with five layers of
reinforced concrete, and in the horizontal brace system
4.1 The Project Outline
in both Section 2 and 3, there are only four layers of
reinforced concrete. Meanwhile, all sections use steel
The International Finance Center is located in the core
upright and cast-in-place concrete piles as the vertical
area of Binjiang central business district in Wuchang
brace system.
alongside Yangtze River, with a distance of 250m from
the flood bank. The site of the center is considered
4.2 The Code Compliance Checking
as a key area for the new-round urban development
of the Construction Plan of Deep Foundation
in Wuhan. The project is a towering building with an
overall floorage of three million square meters and
The construction plan of the deep foundation is pre-
a total investment of more than thirty billion RMB
sented in two formats, one is the traditional two-
(Fig. 13).
dimensional drawings and written words, and the
The overall floorage of A01 region of the Inter-
other is the BIM model. The manual code checking
national Finance Center is 663159m2 , which is made
was carried out by the supervisor committee orga-
up by one main super high-rise building, one auxil-
nized by the quality supervision department of Wuhan.
iary building of office, and one auxiliary building of
Meanwhile the quality supervision department of
apartment and one podium building. The basement of
Wuhan also conducted a code compliance checking
the main building has six floors with a total floorage
by using the BIM-based checking system we devel-
of 70171m2 , and there are 119 floors above the base-
oped. This arrangement aims to investigate the effect
ment with a height of 606m and a total floorage of
of the BIM-based code compliance checking for deep
302399m2 .
foundation.
The deep foundation is a rectangle with the length
of 304m and width of 121m, around which are the (1) The BIM-based checking process (Fig. 14)
J Intell Robot Syst (2015) 79:549–576 569

Table 8 Delivery requirement of architecture design

Object/information Attribute Description Model builder Model receiver


group information

Architecture Deep foundation


engineering
R/O R/O
Project/building information
project Project ID R R
Project name R R
Staff Designer R R
information
Major code R R
Site Site ID R R
Meta data Submitter, R R
versions,date
texture R R
Geographical R R
position R R
Geometry Site geometry R R
outline
Area R R
... ... ... ...
Whole building Building ID R R
position Contrast site R R
Geometry Building R R
geometry outline
... ... ... ...
Building floor Floor ID R R
Floor height R R
Floor elevation R R
... ... ... ...

Annotation: R=require; O=option

The supervisors of the quality supervision after setting up the checking objects. Click
department of Wuhan logged in the system, and the button of ENTERING REASONING
set up a new project in the module of BASIC WORKSPACE, then the checking list could
INFORMATION OF PROJECT, and input basic be built which contains all checking points
information relevant to the project. The sys- and construction parameters of the plan based
tem introduced the checking objects page after on the checking object information of the
the project had been established, then the main project.
checking objects could be ticked manually in The system could directly extract the infor-
accordance with the actual content of the special mation needed for the checking list from the
construction plan, thus completing the structur- BIM model in the page. Click the button
ing work. of EXTRACT CONTRUCTION PARAMETER,
The operating personnel started checking in and then storage location of the BIM model
the module of project checking management could be found in IFC format. Once the model
570 J Intell Robot Syst (2015) 79:549–576

Table 9 Delivery requirement of deep foundation design

Object/information Attribute information Description Model builder Model receiver


group

Deep foundation Supervisor for


engineering code compliance
R/O R/O
The basic information of deep foundation engineering
project Project ID R R
Project name R R
Staff Designer information R R
The classification of Inherited from R R
engineering geological structure
condition
Geological Inherited from R R
exploration structure
data
Major code R R
... ... ... ...
Site Site ID Inherited from architecture R R
Meta data Inherited from architecture R O
texture Inherited from architecture R R
Geographical Inherited from architecture R R
position R R
Geometry Inherited from architecture R R
Area Inherited from architecture R O
... ... ... ...
Supporting
Cantilevered pile Component ID R R
position Contrast site R R
Geometry component R R
geometry outline
Piles diameter R R
Texture R R
Steel level R R
Concrete level R R
... ... ... ...
Underground Component ID R R
diaphragm wall
... ... ... ...
Slope
... ... ... ... ...
Brace
... ... ... ... ...
J Intell Robot Syst (2015) 79:549–576 571

Fig. 12 The preliminary


result of code checking in
system page
Cantilevered pile

Attributes of
Cantilevered pile

Relevant code articles


about cantilevered piles

1 ,
400mm 1200mm ,

Fig. 13 The layout of deep foundation in the International Finance Center


572 J Intell Robot Syst (2015) 79:549–576

Building engineering semantic


structure of this project

The checking list including all


checking points and required
information of this project

Extracting required information


from BIM model

The checking result

Engineering Checking Code Code Checking


Checking points parameters rules name articles results
Diameter of piles 1200 Diameter of piles pass
vertical thickness of the vertical thickness of the
800 pass
reinforced concrete top beam reinforced concrete top beam

Fig. 14 The process of code compliance checking

was chosen, all the needed construction parame- The final checking result was presented in a table
ters of the checking list would be extracted. including checking points, checking results, rel-
Click the button of NEXT STEP, the check- evant code articles etc..
ing data was prepared to do checking calculation. (2) The manual checking process
J Intell Robot Syst (2015) 79:549–576 573

The supervisor team conducted the manual conform to the checking rules. While in manual check-
checking based on paper-written codes, two- ing, the two checking points conformed to the codes.
dimensional drawings and working plan; five Because the length and thickness of the soil body
supervisors spent two hours in inquiring rele- are not divided according to the practical construction
vant codes, then the code compliance checking plan when establishing the soil body in the BIM model
of the project was achieved. The first page of the they differed greatly from the data of the length of
checking results and suggestions are presented in section-divided excavation and thickness of layering
Fig.15. excavation in the written construction plan.
Another conflict is about the checking point of
strength level of concrete of the cast-in-site bored
4.3 The Comparative Analysis on the Checking
pile. The manual checking result conformed to the
Results Made by the System and the Manual Work
code requirement, while the system checking result
did not conform to the code requirement. This is
The system operated at 41 checking points, and made
because the supervisors adopted the national code
the conclusion on whether the practical engineering
‘Technical specification for Retaining and Protection
parameters at each checking point followed the codes.
of Building Foundation excavations JGJ120-99’ [46],
Meanwhile, the original text of relevant code articles
whereas the system applied the local code ‘Hubei
was attached for reference.
Technical Specification for Engineering of Founda-
Manual checking had been finished at 25 check-
tion ExcavationDB42/159-2004’ [41]. For the national
ing points, a smaller number compared with system
code, the strength level of concrete is required to
checking. In the checking results of all checking
be over C20, but the local code requirement is over
points, only code name and the article number were
C30. That’s why two different checking results were
noted because of limited time.
obtained (Fig. 16).
The four conclusions achieved by the system con-
To sum up, the advantage of the checking system
tradict the manual checking results. This may result
for deep foundation is that the system can undertake
from the problems occurring in the information build-
the code compliance checking quickly and compre-
ing of the BIM model and the inconformity of the
hensively.
codes.
The BIM-based checking of the whole project
One conflict occurred in earth excavation, in sys-
needs no more than 10 minutes. Moreover, the check-
tem checking, the thickness of layering excavation
ing points cover all the aspects. Meanwhile, it can’t
and the length of section-divided excavation didn’t
be denied that the precision of the checking will be
influenced by the building quality of the BIM model.
Unfortunately, currently there is still no legal standard
of BIM model for code compliance checking in deep
foundation. As a result, the information building can-
not fully satisfy the need of the checking lists, for
A Technological Demonstration Suggestion for The Deep which manual input is needed and a small amount of
Foundation Project of The International Finance Center
false information may exist due to auto extraction.
The advantage of manual checking is that it ensures
the validity of checking in relatively complex con-
ditions, but the shortage lies in its low efficiency
and failure to ensure the integrity of the checking.
Although the five supervisors had spent two hours in
the checking, but part of the checking points were still
omitted. In addition, its comprehensive cost is high.
It’s worth noting that, due to high time con-
sumption, owners of some projects are reluctant to
conduct code compliance checking. Therefore, BIM-
Fig. 15 The first page of the checking results and suggestions
by the supervisor committee based code compliance checking becomes a new and
574 J Intell Robot Syst (2015) 79:549–576

Fig. 16 The comparison of


the two checking results

Conflicting
points

Manual BIM-based
Checking points
checking checking
strength level of concrete
deny pass
of cast-in-site bored pile

Conflicting
points
Manual BIM-based
Checking points
checking checking
thickness of layering
pass deny
excavation
length of section-divided deny
pass
excavation

efficient way to promote the application of code com- of the checking knowledge library and the standard
pliance checking to avoid potential quality problems of required information. They help achieve the gen-
in deep foundation construction. eration of the checking list, selection of rules and
extraction of required information for code compli-
ance checking. This BIM-based checking solves the
5 Conclusion and Discussion actual difficulties of manual checking in practice,
improves the checking efficiency and precision, which
The research realizes a BIM-based code compliance is hard to be realized through manual checking, and
checking of deep foundation. The main achievement ensures the quality and efficiency of the special con-
of the research is the suggested process of a BIM- struction plan of deep foundation.
based code compliance checking. Currently, the sys- The future research will focus on continuous estab-
tem based on the process has established 23 checking lishment of the library of checking knowledge with
objects and 122 checking points, 305 checking rules more relevant codes, the solution of different stan-
and 528 premises of rules. dards and contradiction of the codes at the same
The achievements of BIM-based code compliance checking point and the development of IDM and MVD
checking process of deep foundation are the building for deep foundation.
J Intell Robot Syst (2015) 79:549–576 575

Acknowledgments The writers are grateful for the financial 17. Conover, D.: Development and Implementationof Auto-
support from the National Natural Science Foundation of China mated Code Compliance Checking in the U.S., Interna-
(Grants. 71301059), and the Fundamental Research Founds for tional Code Council (2007)
National University, China University of Geosciences (Wuhan, 18. International Code Council, About ICC (2014). http://www.
CUGW130210). iccsafe.org/AboutICC/Pages/default.aspx
19. Jain, D., Law, K.H., Karwinkler, H.: On processing stan-
dards with predicate calculus. In: Proceedings of the 16th
Conference on Computing in Civil Engineering, ASCE,
References Atlanta (1989)
20. Rasdorf, W.J., Lakmazaheri, S.: Logic-based approach for
modeling organization of design standards. J. Comput. Civ.
1. HongboZhou, Hui Zhang: Risk assessment methodology Eng. 4, 102–123 (1990)
for a deep foundation pit construction project in Shanghai, 21. Kim, H., Grobler, F.: Design coordination in building infor-
China. J. Constr. Eng. Manag. 137, 1185–1194 (2011) mation modeling using ontological consistency checking.
2. Chiew, S.P.: Behaviour of strut-waler connections with In: Proceedings of the ASCE International Workshop on
different stiffening details. Struct. Eng. 84, 28–31 (2006) Computing in Civil Engineering, vol. 1, pp. 410–421 (2009)
3. Zhao, H., Li, Y.J.: Study on engineering project risk man- 22. Pauwels, P., Van Deursen, D., Verstraeten, R., De Roo,
agement based on Bayesian network. Shenyang Univ. of J.: A semantic rule checking environment for building
Technology 1, 239–243 (2008) performance checking. Autom. Constr. 20, 506–518 (2011)
4. Wang, X.M.: Fifty millions RMB will be compensated to 23. Zhong, B.T., Ding, L.Y., Luo, H.B., Zhou, Y., Hu, H.M.:
Hangzhou Metro for the accident by six insurance agents Ontology-based semantic modeling of regulation constraint
(2008). http://news.21cn.com/domestic/difang/2008/11/19/ for automated construction quality compliance checking.
5491008 1.shtml Autom. Constr. 28, 58–70 (2012)
5. T&TI, Sao Paulo metro shaft collapse. Accessed 17 Feb 24. Fenves, S.J.: Tabular decision logic for structural design. J.
2007 (2007). www.tunnelsonline.info Struct. Eng. 92, 473–490 (1966)
6. Shiwei, Y., Zhu, K.: A dynamic all parameters adaptive BP 25. Nyman, D.J., Fenves, S.J., Wright, R.N.: Restructuring
neural networks model and its application on oil reservoir study of the AISC specification, Civil Engineering Stan-
prediction. Appl. Math. Comput. 195, 66–75 (2008) dards, SRS 393, Department of Civil Engineering, Univer-
7. Song Yi, Liangbo Quan: Ten months 13 death by safety sity of Illinois at Urbana-Champaign, IL (1973)
accident in Shenzhen subway construction site. Accessed 26. Fenves, S.J., Wright, R.N.: The representation and use of
18 Oct 2009 (2009). http://news.ycwb.com/2009-10/28/ design specifications, Technical Note 940, NBS, Washing-
content 2309047 2.htm ton, DC (1977)
8. Frijters, A.C.P., Swuste, P.H.J.J.: Safety Assessment in 27. Fenves, S.J., Garrett, J.H., Kiliccote, H., Law, K.H., Reed,
Design and Preparation Phase. Saf. Sci. 46, 272–281 K.A.: Computer representations of design standards and
(2008) building codes: U.S. perspective, The International Jour-
9. Ministry of Housing and Urban-Rural Development of The nal of Construction Information Technology, University of
People’s Republic of China, Management Method for High Salford,U.K (1995)
Risk Division Components in Chinese (2009) 28. Fenves, S.J., Wright, R.N., Stahl, F.I., Reed, K.A.: Introduc-
10. Zhang, S., Teizer, J., Lee, J.-K.: Building information mod- tion to SASE: Standards Analysis, Synthesis, and Expres-
eling(BIM) and safety: automatic safety checking of con- sion, Report NBSIR 87-3513 U.S, Department of Com-
struction models and schedules. Autom. in Constr. 29, 183– merce, National Bureau of Standards (1987)
195 (2013) 29. Kerrigan, S., Law, K.: Logic-based regulation compliance-
11. Boukamp, F., Akinci, B.: Automated processing of con- assistance. In: Proceedings of The 9th International Confer-
struction specifications to support inspection and quality ence on Artificial Intelligence and Law, Edinburgh (2003)
control. Autom. in Constr. 17, 90–106 (2007) 30. MVD, The Model View Definition, IFC Solutions Factory
12. X.Tan, Hammad, A., Fazio, P.: Automated code compliance (2011). http://www.blisproject.org/IAI-MVD
checking for building envelope design. J. Comput. Civ. Eng. 31. Han, C., Kunz, J., Law, K.: Client/server framework for on-
24, 203–211 (2010) line building code checking. J. Comput. Civ. Eng. 12, 181–
13. Eastman, C., Lee, J., Jeong, Y., Lee, J.: Automatic rule- 194 (1998)
based checking of building designs. Autom. Constr 18, 32. Borrmann, A., Rank, E.: Specification and implementation
1011–1033 (2009) of directional operatorsin a 3D spatial query language for
14. Eastman, C., Teicholz, P., Sacks, R., Liston, K.: BIM building information models. Adv. Eng. Inform. 23, 32–44
Handbook: a Guide to Building Information Modeling for (2009)
Owners, Managers, Architects, Engineers, Contractors and 33. Lee, J.-K., Lee, J., Jeong, Y.-S., Sheward, H.: Development
Fabricators. Wiley, Hoboken (2011) of space database for automated building design review
15. Khemlani, K.: CORENET e-PlanCheck: Singapore’s Auto- systems. Autom. Constr. 24, 203–212 (2012)
mated Code Checking System (2005). http://www.aecbytes. 34. Eastman, C.M., Teicholz, P., Sacks, R., Liston, K.: BIM
com/buildingthefuture/2005/CORENETePlanCheck.html Handbook-A guide to Building Information Modeling for
16. Construction and Real Estate NETwork, Collaboration Part- Owners, Managers, Designers, engineers, and Contractors.
ners (2014). http://www.corenet.gov.sg/ Wiley (2008)
576 J Intell Robot Syst (2015) 79:549–576

35. MVD, The Model View Definition, IFC Solutions Factory 41. Hubei Department of Construction, Hubei technical specifi-
(2011). http://www.blisproject.org/IAI-MVD cation for engineering of foundation excavation DB42/159-
36. Zhang, S., Teizer, J., Lee, J.-K., Eastman, C.M., Venugopal, 2004, Hubei Department of Construction (in Chinese)
M.: Building information modeling (BIM) and safety: (2004)
Automatic safety checking of construction models and 42. Guo, Y.: Comprehensive Evaluation Theory, Methods
schedules. Autom. Constr. 29, 183–195 (2013) and Extensions, (in Chinese). Science Press, Beijing
37. Liebich, T., Adachi, Y., Forester, J., Hyvarinen, J.: Indus- (2004)
try Foundation Classes IFC23, International Alliance for 43. Karlshoj, J.: Information Delivery Manual: Roadmap
Interoperability (2006) (2011). http://iug.buildingsmart.com/idms/roadmap/IDM
38. Solibri Inc, The world leader in design spell checking Roadmap 1 1.pdf/view
(2009). http://www.solibri.com/index.php?option=com 44. Cooper, R., Kagioglou, M., Aouad, G.: Development of
content&amp;task=view&amp;id=14&amp;Itemid=31 a Generic Design and Construction Process, European
39. Lee, J.-K.: Building environment rule and analysis (BERA) Conference on Product Data Technology, pp. 205–214
language and its application for evaluating building cir- (1998)
culation and spatial program, Ph.D. Dissertation, Georgia 45. Zhou, C.: Research on IDM-based Building Information
Institute of Technology (2011) Delivery Standard, M.D. dissertation, Shanghai Jiao Tong
40. Pauwels, P., De Meyer, R., Van Campenhout, R.: Interop- University (in Chinese) (2013)
erability for The Design and Construction Industry through 46. China Academy of Building Research, Technical Specifi-
Semantic Web Technology, 5th International Conference cation for Retaining and Protection of Building Founda-
on Semantic and Digital Media Technologies (SAMT), tion Excavations JGJ 120-99, China Academy of Building
Saarbrücken, Germany (2010) Research (in Chinese) (1998)

You might also like