Professional Documents
Culture Documents
Automation in Construction
journal home page : www . e lsevier. com / locate/autc o
n
a r t i c l e i n f o a b s t r a c t
Article history: Decisions for building maintenance require integration of various types of information and knowledge created
Accepted 19 September 2012
by different members of construction teams such as: maintenance records, work orders, causes and knock-
on effects of failures, etc. Failing to capture and use this information/knowledge results in significant costs
Keywords:
due to ineffective decisions. Majority of the current building maintenance systems mainly focus on
Case-based reasoning
BIM capturing either information or knowledge. This research aims to develop an integrated system to capture
Building maintenance information and knowledge of building maintenance operations when/after maintenance is carried out to
Building knowledge modelling understand how a building is deteriorating and to support preventive/corrective maintenance decisions. To
develop the system, a number of case studies were investigated and interviews were conducted with
professionals from different building maintenance departments in public organisations. This methodology
helped identify the building main- tenance process and the opportunities for knowledge capture and
exchange. A taxonomy for building mainte- nance was then identified which enabled a formal approach
for knowledge capture. The proposed system utilises the functions of information modelling techniques
and knowledge systems to facilitate full retrieval of information and knowledge for maintenance work. The
system consists of two modules; BIM module to capture relevant information and Case-Based Reasoning
(CBR) module to capture knowledge. The system can help main- tenance teams learn from previous
experience and trace the full history of a building element and all affected elements by previous
maintenance operations. It is concluded that the integrated knowledge-based BIM systems can provide
advanced useful functions for construction operations. On the other hand, incorporating Knowledge
Management principles embedded in CBR systems with Information Management principles embedded in
BIM
systems is a way forward for the transformation from ‘Building Information Modelling’ to ‘Building
Knowledge
Modelling’.
© 2012 Elsevier B.V. All rights reserved.
0926-5805/$ – see front matter © 2012 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.autcon.2012.09.008
I. Motawa,
I. Motawa,
A. Almarshad
A. Almarshad
/ Automation
/ Automation
in Construction
in Construction
29 29
(2013)
(2013)
173–182
173–182
17 17
and Alshawi [11] were adopted in a way to overcome these problems. channel for information. BIM comprises ICT frameworks and tools
However, these integrated IT solutions were mainly for information that can support stakeholders’ collaboration over projects life-cycle.
sharing with little emphasis on knowledge capture and sharing. In While BIM is thought to transform the way the built environment is
another approach to improve the capability of these IT solutions in working [14], studies related to BIM have often been focusing on
terms of data management, several technologies were integrated storing and sharing technical information. BIM capabilities can be
with these solutions such as Radio Frequency Identification (RFID) enhanced by knowledge-based techniques, such as Case-Based
by Legner and Thiesse [12] and Chien-Ho Ko [13] to collect and Reasoning (CBR), to
share maintenance data with the most minimal manual data entry. enable both information and knowledge-sharing that will benefit
These efforts, in addition to others, drove to the development of the stake- holders. By incorporating Knowledge Management principles
generic IT solution provided by the principles of Building Information embedded in CBR systems with Information Management principles
Modelling (BIM). embedded in BIM systems, the transformation from ‘Building
BIM can provide the facilities of these integrated solutions Information Modelling’ to ‘Building Knowledge Modelling’ can be
and overcome such problems in a comprehensive manner. BIM can better recognised.
enhance collaboration between team members and facilitate further While many of the current FM applications adopt BIM technology
the mutual and provide functions for BM (such as: assisting in helpdesk call
centres and managing assets, spaces, costs and utilities), Knowledge
Management
Start A
Problem reported from unit
1. Verbal
2. Written
1A
Administration general Approval for payment
manager 12
Asse
ssm Job description with
ent 4 estimated quotes based
E Apprrequ 2A on MTC contract pricings
(for Approval).
m ov est
Approval from er al
branch manager ge for Design/Planning divisions
(within authorised nc pa 3.2A
budget), Otherwise
4
top management 5 2A 11 1. Visiting site and
approval
Asse evaluating problem.
Ap
pr ssm 2. Preparing job description
ov ent 3.1A with estimated quotes based
al requ on MTC contract pricings.
est
Start B
Submitting a finalised
detailed quote 10
Civil Team 1
Problem reported From
Assessment
maintenance team
1. Scheduled 1B 2 A/B Mech. Team 1
2. Observed
Region A branch manager request
Outsourced contractor
In-house team
8 8 Sub
mitti
Supplier ng a
9 finali
Carryout and sed
detai
completing work led
quot
e
Legend
Regular process line Limited process line Urgent process line
General Contract
Conditions
Particular Contract
Conditions
Legal Action/
Bidding law and recommendation
Regulation
Knowledge Subjects
Product Knowledge
BM technical work
Technical Specifications/
packages
Warranties &
Insurance
Internal
Process
External
Administrative
User
Category
BIM Module
Case Problem
attributes description
CBR Library
Attribute weights K-nearest neighbour in a field
of new case Fine-ranking of a field
Knowledge case
weighed list
Solved cases
Filter by
Yes
Refining
Fields
No
Knowledge case
Retaining new case details
Fig. 6. Data for maintenance cases in BIM application (output from Autodesk Revit [32]).
I. Motawa,
I. Motawa,
A. Almarshad
A. Almarshad
/ Automation
/ Automation
in Construction
in Construction
29 29
(2013)
(2013)
173–182
173–182
17 17
of avoiding same mistakes and adopting successful solutions from the specific knowledge of previously experienced problem cases
previous cases. There are also cases when dispute arises between dif- in order to identify a solution for new problems. CBR is also an
ferent stakeholders regarding the kind of remedy actions for certain approach to incremental, sustained learning, since a new experience
maintenance operations. Therefore, detailed analysis and investiga- is retained each time a problem has been solved making it
tion of the causes and effects in addition to detailed knowledge of immediately available for future problems. A typical CBR system
building defects are required. operates according to the fol- lowing steps: 1) identify the current
These reasons make approaches such as ES and ANN problem situation, 2) search past cases similar to the new one, 3)
inappropriate for modelling knowledge of BM. ES need to represent retrieve the most similar case to solve the current problem, 4)
knowledge in the form of well-defined rules with limited domains evaluate the proposed solution, and 5) update the system by learning
of knowledge. This is not always appropriate in BM because the from this experience. The structure of the CBR module developed for
available knowledge on the recorded maintenance events cannot be the proposed BM system is shown in Fig. 5. The following sections
encapsulated in rules explicitly and it is difficult to formulate a discuss the main components of this structure.
complete and accurate set of rules to capture the expertise required
for a generic KM system for BM. In addition, ES cannot normally
learn over time which is a key aspect of BM as new solutions 4.2.1. Case representation
always come to the industry and ES is based on prior knowledge of BM knowledge cases can be represented in a variety of forms using
a few known alternatives. On the other hand, ANN systems were a range of AI representational formalisms such as: frames, objects,
found inappropriate, as they neither provide details of variable predi- cates, semantic nets and rules [31]. For the proposed system,
manipulations nor the rationale for the conclusions drawn, which is the infor- mation of projects and elements already stored by the BIM
needed in cases of dispute. While ANN can learn from new cases, it module will be retrieved to identify all technical and maintenance
needs huge data to train and develop. From this brief discussion, information of the BM case, including any information of the related
CBR systems were found the most suitable approach for modelling building elements to the maintained element. Common attributes of
the knowledge domain of BM. the cases should be identified in a way to simplify entering case
The psychological plausibility of CBR makes it fundamentally description and to enable the retrieval of the most suitable cases.
differ- ent from other major AI approaches as a problem solving From the interviews conducted to develop the adopted BM process, a
paradigm in the form of intra-domain analogy [30]. Cognitive number of case attributes for BM were defined as shown in Table 2.
psychology research considers reasoning by reusing past cases as a These attributes have been ranked and given weights to distinguish
powerful and frequently applied way to solve problems for human. the importance of each attribute to a
Instead of relying solely on the general knowledge of a problem knowledge case. These attributes in addition to “Problem
domain or making relationships between problem descriptors and description”
conclusions, CBR is able to utilise will discriminate BM cases, so useful cases can be retrieved using
the adopted evaluation functions in the “Similar case retrieval” step.