Professional Documents
Culture Documents
1-s2.0-S0926580521000042-main
1-s2.0-S0926580521000042-main
1-s2.0-S0926580521000042-main
Automation in Construction
journal homepage: www.elsevier.com/locate/autcon
A R T I C L E I N F O A B S T R A C T
Keywords: The construction industry has a higher occupational casualty rate than other industries. As a proactive approach
Construction safety to safety management, Construction Hazard Prevention through Design (CHPtD) can significantly eliminate or
Safety risk assessment reduce the construction safety risk. However, this concept is not implemented effectively in practice because the
Building information modeling (BIM)
technical issues that underlie CHPtD have not been addressed. This paper proposes a novel method of quanti
Design stage
tative construction safety risk assessment for building projects at the design stage. This method consists of three
indexes: Likelihood, consequence and exposure. These indexes are calculated using occupational injury, fatality
and specific construction planning data which are accurate and objective. A plug-in that links building infor
mation modeling (BIM) with safety risk data is developed in Autodesk Revit, which can automatically calculate
construction safety risk to help architects and structural designers quickly select design alternatives. A case study
is presented to demonstrate the feasibility and effectiveness of the proposed method.
* Corresponding author.
E-mail address: luying_happy@126.com (Y. Lu).
https://doi.org/10.1016/j.autcon.2021.103553
Received 7 July 2020; Received in revised form 23 October 2020; Accepted 5 January 2021
Available online 23 January 2021
0926-5805/© 2021 Elsevier B.V. All rights reserved.
Y. Lu et al. Automation in Construction 124 (2021) 103553
construction processes and safety risk which is necessary to perform safety risk. Esmaeili et al. [29] proposed an attribute-based risk assess
CHPtD. Only 3.2% of architects and 9.8% of structural designers stated ment approach to quantify the risk of construction activities.
that their firm has guidelines for reviewing CHPtD [14]. The main The third approach is occupation-oriented, which concentrates on
reason for this is that the technical issues underlying CHPtD in order to analyzing injury and fatality data of different occupations from the
help architects and structure designers better perform CHPtD have not construction industry viewpoint. Baradan and Usmen [30] analyzed the
yet been addressed [15]. Safety risk assessment in the design stage is nonfatal injury and fatality risks of sixteen construction occupations.
necessary to perform CHPtD. Traditional approaches to assessing con The results indicated that ironworkers and roofers have the highest
struction safety risk are implemented after design, such as risk checklists safety risk. Choe and Leite [31] demonstrated that each occupation has a
and risk management systems, which rely on manual operations and are unique risk data structure with regard to injury sources. Occupation-
time-consuming and inefficient in practical construction projects [9]. oriented approaches are unable to identify further hazards generated
Therefore, as noted by Hardison and Hallowell [16], the key points of from the subsequent construction process. However, occupation-
technical issues include how to use risk-based methods to select design oriented approaches have several advantages over project-based ap
alternatives that optimize construction safety, and how to use novel proaches and activity-based approaches in the design stage. First, the
tools to improve the selection of an optimal design solution. The avail data collection of occupation-oriented approaches is more accurate and
ability of CHPtD computer tools can be a favorable stimulus for CHPtD objective. The previous two approaches are mainly based on historical
implementation [4]. Building information modeling (BIM) provides a data or survey-based data [32–34], while occupation-oriented ap
platform to enable other BIM-based development to integrate external proaches can extract real statistical data from existing occupational
data with the parametric BIM environment and perform further auto databases. Second, different occupations have different levels of casu
matic risk analysis in the design of buildings, which can help architects alty frequency and severity, even in the same construction project or
and designers quickly select or improve design alternatives [17]. The use construction activity. Therefore, safety risks should be quantified ac
of BIM-related technologies in building design and safety management cording to the unique nature of different occupations.
are growing rapidly. However, the majority of previous research
remained in the hazard identification stage [18,19]. Few studies have 2.2. Quantitative model for construction safety risk assessment
focused on the use of the BIM platform to quantify and calculate
comprehensive safety risk. Therefore, the BIM-integrated construction Different scholars have different interpretations of the concept of
safety risk assessment at the design stage was proposed in this study to safety risk. Esmaeili and Hallowell [35] proposed that safety risk in
further implement the concept of CHPtD. cludes likelihood and consequence. Likelihood represents the number of
This paper proposes a quantitative construction safety risk assess injuries or fatalities per hour. Consequence describes the severity of
ment model that considers three indexes: likelihood, consequence and injuries or fatalities. Some scholars use likelihood and consequence to
exposure. These indexes are calculated using occupational injury, fa establish a quantitative model and evaluate construction safety risk.
tality and specific construction planning data which are accurate and Hallowell et al. [36] established a safety risk scale that considers all
objective. A plug-in that links BIM with safety risk data is developed in possible likelihoods and consequences. Azadeh-Fard et al. [37] intro
Autodesk Revit. The plug-in can automatically calculate construction duced a new consequence scoring system that considers multiple injury
safety risk to help architects and structural designers quickly select consequence factors, and employed it as part of a novel risk assessment
design alternatives that optimize construction safety. In addition, a case matrix. Mohandes et al. [38] developed a novel risk quantification
study is presented to demonstrate the feasibility and effectiveness of the model with a probability parameter and severity parameter where the
proposed method for construction safety risk assessment at the design Fuzzy Best Worst Method is exploited to obtain the weights of each
stage. parameter to calculate the safety risks of construction workers.
Exposure is also an important index of safety risk evaluation. The US
2. Literature review Occupational Information Network (O*NET) program [39,40] empha
sized that construction work conditions involve many hazardous expo
2.1. Safety risk assessment in construction sures such as contaminants, noise, fall risks and other hazards. The
program quantified the hazard exposure by giving an exposure score to
Scholars have proposed several methods from different perspectives each occupation according to the number of workers and exposure
to quantify construction safety risk, which can be summarized into three duration. Some scholars proposed that exposure can be described as the
aspects. The first aspect is a project-based approach that concentrates on time of contact with the dangerous situation [41]. Risk can be identified
analyzing the causes of construction hazards and comprehensively as the product of likelihood, consequence and exposure [42–45]. Frijters
estimating the construction safety risk level. Some scholars express that and Swuste [46] utilized a risk quantification table to estimate the
organizing for safety supervision and equipment management, compli likelihood and consequence, and determined the exposure by calcu
ance with applicable safety regulations, attitudes about risk taking and lating the total duration of the construction processes. Hallowell and
management behavior, and high quality and frequent safety communi Gambatese [28] proposed a risk scale that includes likelihood and
cation are critical factors that affect the safety behavior of construction consequence, and calculated exposure in terms of worker-hours. Gang
workers [20,21]. Gunduz and Laitinen [22] proposed a dynamic risk olells et al. [47] rated the likelihood and consequence of a construction
assessment model that consists of a 14-item checklist and a two-scale activity according to the opinions of experts, and assessed the exposure
risk matrix. Risk factors include order and tidiness, chemical hazards, according to construction project documents. Dharmapalan et al. [48]
psychology factors, etc. It is necessary to establish various types of safety distributed questionnaires to general contracting superintendents to
communications between all parts [23]. The safety risk assessment determine the risk scale of likelihood and consequence, and quantified
methods include fuzzy sets, the analytic hierarchy process and the the exposure based on worker-hours per unit.
weighted aggregated sum product assessment method [24–26]. However, the values of these three indexes are typically conducted
The second approach is activity-oriented, which highlights creating a through expert interviews and questionnaires, which are easily influ
work breakdown structure, analyzing related hazards, and hierar enced by subjective judgment [49]. Therefore, solving superior data
chically evaluating construction safety risk. Gurcanli et al. [27] estab source issues, which should be accurate and objective, is urgent.
lished the work breakdown structures of 25 building projects and listed
all construction activities of each construction project to determine 2.3. Application of BIM in safety risk
different safety risks. Hallowell and Gambatese [28] identified 13 con
struction activities of building concrete formwork, and quantified each As an object-oriented parametric digital representation, BIM is
2
Y. Lu et al. Automation in Construction 124 (2021) 103553
Data Collection
Fatality Days-Away Injury Unit Labor
Hours
CFOI Number Number SOII
Required Labor
Resource RSMeans
Construction
Occupation Material
Data Mapping
BIM Model
Identity
IDs of Building
Elements
Establish
Construction
Process
Acquire Acquire
Construction
Quantity
Occupation
Select Select
Injury, Fatality Unit Labor
Data Hours Data
Data Analysis
API
Automatic
Calculation of Safety
Risk
becoming a systematic tool that is changing the project planning, design, identity and prevent the hazards associated with falls. Guo et al. [53]
construction, operation and maintenance phases. In the design stage, combined BIM with design safety rules to automatically identify possible
BIM is an effective way to link risk data with design elements in BIM safety problems in the design phase. Hussein et al. [54] developed a risk
[50]. Many scholars combine BIM with relevant databases to realize the review system that was integrated with BIM to help architects and
concept of PtD, which aims to help architects and structure designers structural designers check design elements. Cortés-Pérez et al. [55] in
identify hazards and provide solutions [51]. Two main trends exist in tegrated an occupational health and safety risk assessment into BIM
achieving PtD: using BIM software for clash detection and integrating projects in Spain and parameterized BIM objects according to Spanish
BIM and knowledge-based systems to detect hazards [9]. For instance, regulations and hazard preventive measures in the design phase. The
Hu et al. [52] created a model of a continuous and dynamic simulation assessment results were shown with color codes based on the risk rating
of the entire construction process and then applied BIM technology for level of each object and then the corresponding measures were
collision inspection. Zhang et al. [12] integrated safety rules into BIM to implemented.
3
Y. Lu et al. Automation in Construction 124 (2021) 103553
The methodology in this paper includes a theoretical framework and Based on the information stored in these databases, the indexes of
a developed plug-in. The proposed theoretical framework includes data construction safety risk assessment can be calculated automatically ac
collection, data mapping and data analysis, which are employed to cording to the corresponding formulas. Detailed calculation methods
quantitatively calculate construction safety risks by occupation, as will be introduced in Section 3.1.3. All the indexes and their data sources
shown in Fig. 1. The developed plug-in implements the proposed theo are preliminarily listed in Table 1 to illustrate the relationship between
retical framework and serves as a platform to automatically calculate the risk evaluation process and the databases.
construction safety risk.
3.1.2. Data mapping
3.1. Proposed theoretical framework BIM is a modeling technique that is oriented to family objects. By
establishing a custom family library of basic building components, more
3.1.1. Data collection dimensional information can be integrated beyond dimensional infor
As a visual modeling tool, BIM can provide the information needed in mation. The families were divided according to the construction pro
the planning, design, construction and operation stages. However, BIM cesses. The two previously mentioned databases were linked to BIM, and
lacks data related to safety risk assessment. Safety risk data should be DB Link was applied to extract and update data. As shown in Fig. 2, the
systematically captured in the data collection phase. Therefore, two process of data mapping includes two steps. First, the required labor
databases that integrate all the data for risk assessment were newly built resources, required occupation ID, and unit labor hour of each con
in the data collection phase: The Injury and Fatality Database and the struction process stored in the Construction Planning Database were
Construction Planning Database. The former database recorded the ca mapped into the extension parameters of the Revit family objects. Sec
sualty information and the employment size of each construction ond, the required occupation IDs obtained from the first mapping were
occupation. The latter collected data on the specific construction pro mapped to the Injury and Fatality Database. Then, the likelihood and
cess, unit labor hours, and required labor resources. All the information consequence indexes of the corresponding occupations can be obtained.
aggregated in these databases were based on the US construction The quantities of the construction materials are automatically calculated
industry. in Revit. The exposure index can be calculated according to the quan
In the Injury and Fatality Database, the casualty types were divided tities and unit labor hour. Overall, data mapping is responsible for
into days-away injury and fatality, according to the BLS Injuries, Ill linking the collected data with the BIM platform to quantify the safety
nesses, and Fatalities (IIF) program. The BLS IIF program collects annual risk.
data from diverse state, federal, and independent data sources in the US
to identify and verify the number of occupational days-away injuries and 3.1.3. Data analysis
fatalities of all trades [2]. The data of the construction industry extracted The logic diagram of data analysis is shown in Fig. 3. The likelihood
from the 2018 BLS IIF Program were applied in this study. Specifically, index is divided into the fatality frequency index and the days-away
the data of days-away injury were collected from the Survey of Occu injury frequency index. The fatality frequency index for occupation i
pational Injury and Illness (SOII) of the IIF program. The fatality data (FFIi) is calculated by the number of fatalities (NFi) and employment size
were collected from the Census of Fatal Occupational Injuries (CFOI) of (ESi). The days-away injury frequency index for occupation i (DAIFIi) is
the IIF program. The BLS SOII only collects data from the private sectors, calculated by the number of injuries (NIi) and employment size (ESi), as
while the BLS CFOI collects data from the private sectors, government shown in Eqs. (1) and (2) [31].
workers, and self-employed workers by industry. Therefore, data from
FFI i = NF i /ESi (1)
the private sectors were used in this study for consistency. Nineteen
construction occupations were selected for analysis in this paper based DAIFI i = NI i /ESi (2)
on the 2010 Standard Occupational Classification (SOC) system. The
total number of workers employed in the private sector were collected To compare with the average occupational injury rate, the relative
from the BLS Occupational Employment Statistics (OES). All of these fatality frequency index (RFFIi) and relative days-away injury frequency
data are integrated in the Injury and Fatality Database. index (RDAIFIi) for occupation i are created, as shown in Eq. (3) and Eq.
In the Construction Planning Database, the unit labor hours, required (4), respectively.
labor resources and construction materials were acquired from /
RFFI i = FFI i FFI average (3)
RSMeans. RSMeans is an existing mature construction cost database
owned by Gordian Company; the database contains the most compre /
RDAIFI i = DAIFI i DAIFI average (4)
hensive data of unit material costs, unit labor costs, unit equipment
costs, labor sources, city cost indexes, equipment rentals, location fac Considering the same consequence of fatality, the relative fatality
tors and other information of construction projects in North America severity for occupation i (RFSi) is set to 1. For the relative days-away
[57]. RSMeans can reflect the average level of the US construction in injury severity for occupation i (RDAISi), the median days-away for
dustry. In this study, only three types of data were applied to calculate occupation i (MDAi) is combined to consider the different levels of
the exposure index: the unit labor hours, required labor resources, and severity of construction occupational injuries, as shown in Eqs. (5) and
construction materials. Therefore, only these three types of data were (6).
separately acquired from RSMeans to integrate into the Construction
RFSi = 1 (5)
Planning Database.
4
Y. Lu et al. Automation in Construction 124 (2021) 103553
Revit data table Construction Planning Database Injury and Fatality Database
RFR1 RDAIR1
TFR TDAIR
/ /
RDAISi = MDAi MDAaverage (6) REI i = EI i EI average (9)
The exposure index for occupation i (EIi) is determined by the Safety risk is identified as the product of likelihood, consequence and
quantity (Qj) related to the material of construction process j, and the exposure. The relative fatality risk (RFRi) and relative days-away injury
unit labor hours involved in construction process j. Moreover, the risk (RDAIRi) for occupation i are proposed, as shown in Eqs. (10) and
average exposure index (EIaverage) is determined by the total exposure (11). A construction process usually requires that multiple occupations
index (EItotal) and the total number of occupations (n) in the construction to collaborate, so the total risk of the construction process is the sum of
process. To compare with the average exposure index of the occupa the occupational risks. The total fatality risk (TFR) and total days-away
tions, the relative exposure index for occupation i (REIi) is proposed, as injury risk (TDAIR) are proposed, as shown in Eq. (12) and Eq. (13),
shown in Eqs. (7), (8) and (9). respectively.
∑
m
RFRi = RFFI i × REI i × RFSi (10)
EI i = Qj × Unit labor hoursj (7)
(11)
j=1
RDAIRi = RDAIFI i × REI i × RDAISi
/
EI average = EI total n (8)
RFFI
Automatic calculation of
RFS
Building component BIM for occupational construction safety risk:
RDAIFI
family library safety risk assessment TFR
RDAIS
TDAIR
REI
5
Y. Lu et al. Automation in Construction 124 (2021) 103553
∑
n calculation process of construction safety risk is shown in Fig. 4.
TFR = RFRi (12)
i=1
3.2.2. Main development
∑
n To realize the proposed theoretical framework and embed the risk
TDAIR = RDAIRi (13) data analysis into the BIM model, a functional plug-in is designed and
developed. As previously mentioned, relevant risk data can be collected
i=1
6
Y. Lu et al. Automation in Construction 124 (2021) 103553
Table 2
Available data.
BIM model Construction process Quantity Required labor Unit labor
resource hours
Plan A
Shop-Fabricated Wood Trusses 5 2 Carpenters 0.800 h/
Roof Trusses 2 Construction Ea.
Common wood, 2′′ × 4′′ metal plate laborers
connected, 24” O.C., 4/12 slope, 1 Construction
1’overhang, 32′ span equipment operator
Prefinished Plywood Paneling 3202 2 Carpenters 0.032 h/
Paneling, Plywood S.F.
Plywood, prefinished, 1/4”thick, 4′ × 8′
sheets with vertical grooves. Birch- faced,
minimum
Damp proofing 3202 1 Roofer 0.012 h/
Bituminous Asphalt Coating S.F.
Brushed on, below grade, 1 coat
Metal Shingles 3202 1 Carpenter 0.016 h/
Aluminum shingles S.F.
Mill finish, 0.019 thick
Plan B
Shop-Fabricated Wood Trusses 5 2 Carpenters 0.889 h/
Roof Trusses 2 Construction Ea.
Common wood, 2′′ × 4′′ metal plate laborers
connected, 24” O.C., 8/12 slope, 1 Construction
1’overhang, 32′ span equipment operator
Prefinished Plywood Paneling 3651 2 Carpenters 0.038 h/
Paneling, Plywood S.F.
Plywood, prefinished, 3/4′′ thick, stock
grades, minimum
Damp proofing 3651 1 Roofer 0.016 h/
Bituminous Asphalt Coating S.F.
Brushed on, below grade, 2 coats
Metal Shingles 3651 1 Carpenter 0.016 h/
Aluminum shingles S.F.
Mill finish, 0.020 thick
extended to the attributes of the corresponding coded family. The 3.2.3. User interface
element filter (FilteredElementCollector) in Revit API is used to traverse Once a building design model is loaded into Revit, the user interface
all the objects in the BIM model and return the objects that satisfy the of the plug-in can be launched through the add-in menu. As shown in
conditions to extract the number of standardized components and Fig. 5, the interface consists of four modules. In the basic information
required completion hours. module, the interface automatically obtains the ID and the name of the
Step2: Realizing information interaction between databases and BIM building element selected by the user. In the second module, users can
model based on Revit DB Link. enter and select the intended construction process. The number of
To calculate the safety risks of all occupations at a construction site, workers, occupation types and unit labor hours in the Construction
the Construction Planning Database and Injury and Fatality Database Planning Database can be accessed according to the chosen construction
need to be connected by the occupations ID, which extend the RFFI, materials. The quantities of the construction materials are automatically
RDAIFI, RFS, and RDAIS data of each occupation in the Injury and Fa calculated. Different quantity calculation methods exist for different
tality Database to the occupations’ information. Considering that all construction components. Three types of construction materials exist;
casualty data are updated annually, the Revit DB Link, which is a plug-in the quantities of these materials are presented by the number, length
that is often used to maintain the connection between the BIM model and surface area. 1) The number of certain materials can be obtained
and the SQL Server database, can be used to import the updated data to directly in the self-contained list of building categories. 2) The lengths of
the BIM model to render the calculation of occupational safety risks certain materials can be identified by the automatic dimensioning
more accurate. function in Revit. 3) The surface areas of certain materials can be
Step3: Inputting the calculation results on the plug-in interface based on calculated by programming in Microsoft Visual Studio and using the
Visual Studio. external command mode to load, so as to achieve Revit secondary
Windows Presentation Foundation is employed to realize this func development. In the third module, the interface can automatically
tion. Windows Presentation Foundation provides a unified program retrieve and calculate the corresponding relative fatality frequency
ming model, language and framework, which can be used to design a index (RFFI), relative fatality severity (RFS), relative days-away injury
new multimedia interactive user graphical interface. When using frequency index (RDAIFI), and relative days-away injury severity
Microsoft Visual Studio and Revit API to develop Windows Presentation (RDAIS) within the Injury and Fatality Database. The exposure index (EI)
Foundation, users can create interfaces that meet their needs and access and relative exposure index (REI) can also be calculated automatically
all information obtained from the BIM model, the Injury and Fatality with the data in the second module. The relative fatality risk (RFR) and
Database and the Construction Planning Database. In this study, all the relative days-away injury risk (RDAIR) can be calculated with the rela
formulas in Section 3.1.3 are compiled into callable functions in Visual tive fatality frequency index (RFFI), relative fatality severity (RFS),
Studio. After reading the RFFI, RFS, RDAIFI, RDAIS and EI data that are relative days-away injury frequency index (RDAIFI), relative days-away
synchronously associated with the occupations ID in the plug-in inter injury severity (RDAIS), and relative exposure index (REI). In the last
face, the calculation results of REI, RFR, RDAIR, TFR and TDAIR are module, the interface can automatically calculate the total fatality risk
automatically fed to the plug-in interface through the Revit API. (TFR) and total days-away injury risk (TDAIR) of these occupations.
7
Y. Lu et al. Automation in Construction 124 (2021) 103553
Table 3
Relative frequency and severity for construction occupations.
Construction occupations (SOC code) Employment size Fatality Days-away injury
Roof
(plan A) Shop-Fabricated Wood Trusses Prefinished Plywood Paneling Dampproofing Metal Shingles
1
2
2 Construction 2 1 1
Construction
Carpenters equipment Carpenters Roofer Carpenter
laborers
operator
Likelihood Consequence Exposure Likelihood Consequence Exposure Likelihood Consequence Exposure Likelihood Consequence Exposure
0.7 1.3 1 0.9 3.2 1.6 1.5 1 0.9 0.1 0.7 0.4 1 2.3 0.1 4.6 1.2 1 0.7 0.8
6.2 4.6
Roof
(plan B) Shop-Fabricated Wood Trusses Prefinished Plywood Paneling Dampproofing Metal Shingles
1
2
2 Construction 2 1 1
Construction
Carpenters equipment Carpenters Roofer Carpenter
laborers
operator
Likelihood Consequence Exposure Likelihood Consequence Exposure Likelihood Consequence Exposure Likelihood Consequence Exposure
0.7 1.3 1 0.9 3.1 1.6 1.5 1 0.9 0.1 0.7 0.4 1 2.3 0.1 4.6 1.2 1 0.7 0.9
6.6 4.6
4. Case study Construction Planning Database provides required labor resource and
unit labor hours for construction processes. According to quantity
To demonstrate the theoretical framework and operability of the calculation rules of the US Construction Specification Institute, the
developed plug-in, a case study is presented. This case is a three-story quantity of construction material is calculated based on their actual
private building that is mainly employed to analyze the construction data. In the BIM environment, the quantities of the construction mate
safety risk involved in roofing work. Roofing work includes tasks such as rials can be automatically calculated. In this case study, the quantity of
cutting, aligning and attaching plywood sheets, asphalt felt, shingles, wood trusses is presented by the number of this kind of material. The
flashing to the roof structure. In this case, two construction plans have quantities of plywood, damp proofing and metal shingles of the roof
been established, namely, plan A and plan B, as shown in Table 2. The element are calculated by the surface area of the single-story roof. The
8
Y. Lu et al. Automation in Construction 124 (2021) 103553
corresponding secondary development code mainly realizes the The manual calculation result matches the automatic calculation
following functions: (1) identify all these layers for calculation, (2) result obtained from the developed plug-in. This result validates that
determine the angles between the normal vector of each surface and the using this plug-in in Revit can quickly calculate the construction safety
positive Z-axis, (3) identify all faces with angles greater than 0, and (4) risk of different design plans in a very short time and shows that the
sum the areas of all these selected surfaces. plug-in developed in this paper has excellent engineering application
According to Eqs. (1), (2), (3), (4), (5) and (6), Table 3 shows the value.
likelihood and consequence index of 19 construction occupations using The final results obtained from the developed plug-in of this case
the injury and fatality data of 2018. Only 4 kinds of occupations are show that plans A and B have the same total days-away injury risk
shown in this case. The findings conclude that the relative fatality fre (TDAIR), but plan A has a lower total fatality risk (TFR). Architects and
quency index (RFFI) of roofers (47–2180) and the relative days-away structural designers should quickly select design plan A. Fig. 6 shows
injury frequency index (RDAIFI) of construction laborers (47–2060) that carpenters (47–2030) and roofers (47–2180) have a higher relative
are relatively high. Additionally, the relative days-away injury severity fatality risk (RFR) and relative days-away injury risk (RDAIR) in plan A.
(RDAIS) of construction equipment operators (47–2070) is 3.3 times Thus, the architects and structural designers should focus on the con
that of roofers (47–2180). struction work of carpenters and roofers. Effective protective design
The relative exposure index (REI), relative fatality risk (RFR) and should be considered at the design stage.
relative days-away injury risk (RDAIR) for construction occupations in
plan A and plan B are calculated by Eqs. (7), (8), (9), (10) and (11). The 5. Discussion
total fatality risk (TFR) and total days-away injury risk (TDAIR) for
roofing work in plan A and plan B can be calculated by Eqs. (12) and 5.1. Generality of the methodology
(13), respectively, as shown in Fig. 6.
In this process, Revit API is used to convert the data into computer All the data of this research were collected from the US construction
language. The specific operations are listed as follows: (1) selecting roof industry. The numbers of occupational days-away injuries and fatalities
element; (2) loading the developed plug-in; (3) selecting the corre collected from the 2018 BLS IIF Program and the unit labor hours,
sponding construction process in the user interface; (4) clicking the required labor resources and construction materials extracted from the
“Calculation” button in sequence to realize automatic calculation of the RSMeans Database were based on the US construction industry.
construction safety risk. The automatic calculation results with plan A Different countries have different casualty statistics, which reflects
and plan B are shown in Fig. 7 and Fig. 8, respectively. various safety management situations. Abueisheh et al. [4] investigated
9
Y. Lu et al. Automation in Construction 124 (2021) 103553
the implementation situation of design for safety in the Palestine con and economic situations may also be considered. From the perspective of
struction industry which differed from the developed countries and construction companies, policy requirements, the extent to which death
generated effective strategies that have been implemented in Palestine. and injury indexes affect the reputation of the company, and the legal
Therefore, the calculation results of the case study cannot reflect the penalties for death and injury need to be considered while determining
actual safety risk situation in other countries. However, the proposed the weights.
quantitative risk assessment method in Section 3 can be directly applied
to other countries if they have detailed statistics on injuries and fatalities 6. Conclusion
and the quota system data of different construction occupations.
As a proactive safety intervention for reducing the construction
5.2. Integration of TFR and TDAIR safety risk, the use of CHPtD is recommended in practice. The technical
issue that underlies CHPtD is how to use risk-based methods to select
When selecting design alternatives in practice, in some situations, design alternatives that optimize construction safety. This paper pro
the total fatality risk (TFR) index of one design plan is higher than posed a quantitative construction safety risk assessment method that
another design plan, while the total days-away injury risk (TDAIR) index consists of three indexes: likelihood, consequence and exposure. These
is lower. More than two design alternatives may also exist. In these indexes can be calculated using occupational injury, fatality and specific
possible scenarios, to the separate calculation of the values of the two construction planning data which are accurate and objective. These data
indexes is not enough. The designers should comprehensively consider are closely related to the design features of a building project. The
the plan according to the policy, expert opinions and other circum technical issue that underlies CHPtD is how to use novel tools to improve
stances. By the certain weight determination method, the comprehen the selection of an optimal design solution. A plug-in that links BIM with
sive objective function of risk can be determined by combining both the safety risk data has been developed in Autodesk Revit. This plug-in can
TFR index and the TDAIR index. Similarly, decisions can be made based automatically calculate construction safety risk to quickly help archi
on the values of this objective function when more than two design al tects and structural designers select design alternatives that optimize
ternatives are available. construction safety. This approach resolves the problem that the archi
To determine the specific weights of these two indexes in the tects and structural designers cannot execute CHPtD in practice due to a
objective function, many factors have to be considered. Differences may lack of safety knowledge. In addition, a case study is presented to
exist in different countries, different regions, and different companies. demonstrate the feasibility and effectiveness of the proposed method for
From the perspective of countries and regions, different social, political construction safety risk assessment at the design stage. The final results
10
Y. Lu et al. Automation in Construction 124 (2021) 103553
obtained from the developed plug-in of the case show that plan A has a [11] M. Behm, Linking construction fatalities to the design for construction safety
concept, Saf. Sci. 43 (8) (2005) 589–611, https://doi.org/10.1016/j.
lower total fatality risk (TFR). In plan A, carpenters (47–2030) and
ssci.2005.04.002.
roofers (47–2180) have a higher relative fatality risk (RFR) and relative [12] S. Zhang, K. Sulankivi, M. Kiviniemi, I. Romo, C.M. Eastman, J. Teizer, BIM-based
days-away injury risk (RDAIR). The architects and structural designers fall hazard identification and prevention in construction safety planning, Saf. Sci.
can consider an effective protective design for the construction work of 72 (2015) 31–45, https://doi.org/10.1016/j.ssci.2014.08.001.
[13] A. Golabchi, S.U. Han, S. Abourizk, A simulation and visualization-based
carpenters and roofers at the design stage. framework of labor efficiency and safety analysis for prevention through design
The data of fatality and days-away injury were collected by 19 and planning, Autom. Constr. 96 (DEC) (2018) 310–323, https://doi.org/10.1016/
construction occupations in this paper. For the same occupation, no j.autcon.2018.10.001.
[14] N. Tymvios, J. Gambatese, D. Sillars, Designer, Contractor, and Owner Views on
further analysis of safety risks arose from different construction activ the Topic of Design for Construction Worker Safety, Construction Research
ities. Neither BLS SOII nor BLS CFOI provides risk data according to the Congress, American Society of Civil Engineers, 2012, pp. 341–355, https://doi.
causes of death or injury of construction workers with the same occu org/10.1061/9780784412329.035.
[15] J.A. Gambatese, M. Behm, S. Rajendran, Design’s role in construction accident
pation. Using the occupational casualty data of different construction causality and prevention: perspectives from an expert panel, Saf. Sci. 6 (4) (2008)
activities for further analysis after refining the statistical caliber is more 675–691, https://doi.org/10.1016/j.ssci.2007.06.010.
reasonable. Furthermore, in the traditional method of procuring- [16] D. Hardison, M. Hallowell, Construction hazard prevention through design: review
of perspectives, evidence, and future objective research agenda, Saf. Sci. 120
construction, design-bid-build, the separation of design and construc (2019) 517–526, https://doi.org/10.1016/j.ssci.2019.08.001.
tion precludes a designer from considering construction site safety [17] R.S. Nizam, C. Zhang, L. Tian, A BIM based tool for assessing embodied energy for
during the design process. In practice, the effect of the CHPtD concept buildings, Build. Energy Efficiency 170 (2018) 1–14, https://doi.org/10.1016/J.
ENBUILD.2018.03.067.
may gain momentum especially in Design-Build (DB), Engineering
[18] K.F. Chien, Z.H. Wu, S.C. Huang, Identifying and assessing critical risk factors for
Procurement Construction (EPC) or Integrated Project Delivery (IPD) BIM projects: empirical study, Autom. Constr. 45 (2014) 1–15, https://doi.org/
project delivery methods where the integration of design and con 10.1016/J.AUTCON.2014.04.012.
struction can be realized [11]. In the design phase, the team consists of [19] W. Solihin, J. Dimyadi, Y.C. Lee, C. Eastman, R. Amor, Simplified schema queries
for supporting BIM-based rule-checking applications, Autom. Constr. 117 (2020),
not only designers, but also architects and engineers [58], which enable https://doi.org/10.1016/J.AUTCON.2020.103248.
a collaborative team relationship. The engineers can provide detailed [20] D. Langford, S. Rowlinson, E. Sawacha, Safety behaviour and safety management:
data of the construction process in advance for better implementation of its influence on the attitudes of workers in the UK construction industry, Eng.
Constr. Archit. Manag. 7 (2) (2000) 133–140, https://doi.org/10.1108/eb021138.
CHPtD. [21] B. Zhang, Z. Chu, L. Cheng, N. Zou, A quantitative safety regulation compliance
level evaluation method, Saf. Sci. 112 (2019) 81–89, https://doi.org/10.1016/j.
Declaration of Competing Interest ssci.2018.10.016.
[22] M. Gunduz, H. Laitinen, Construction safety risk assessment with introduced
control levels, J. Civ. Eng. Manag. 24 (1) (2018) 11–18, https://doi.org/10.3846/
The authors declare that they have no known competing financial JCEM.2018.284.
interests or personal relationships that could have appeared to influence [23] V. Zamani, S.Y. Banihashemi, A. Abbasi, How can communication networks among
excavator crew members in construction projects affect the relationship between
the work reported in this paper. safety climate and safety outcomes? Saf. Sci. 128 (2020) https://doi.org/10.1016/
j.ssci.2020.104737.
Acknowledgements [24] A. Pinto, QRAM, a qualitative occupational safety risk assessment model for the
construction industry that incorporate uncertainties by the use of fuzzy sets, Saf.
Sci. 63 (2014) 57–76, https://doi.org/10.1016/j.ssci.2013.10.019.
The authors wish to express their sincere gratitude to the National [25] T. Dejus, J. Antucheviciene, Assessment of health and safety solutions at a
Social Science Foundation of China (Grant No. 19BGL238), the Natural construction site, J. Civ. Eng. Manag. 19 (5) (2013) 728–737, https://doi.org/
Science Foundation of Jiangsu Province, China (Grant No. BK20181276) 10.3846/13923730.2013.812578.
[26] S. Aminbakhsh, M. Gunduz, R. Sonmez, Safety risk assessment using analytic
and Social Science Foundation of Jiangsu Province, China (Grant No. hierarchy process (AHP) during planning and budgeting of construction projects,
17GLB002). J. Saf. Res. 46 (2013) 99–105, https://doi.org/10.1016/j.jsr.2013.05.003.
[27] G.E. Gurcanli, S. Bilir, M. Sevim, Activity based risk assessment and safety cost
estimation for residential building construction projects, Saf. Sci. 80 (2015) 1–12,
References https://doi.org/10.1016/j.ssci.2015.07.002.
[28] M.R. Hallowell, J.A. Gambatese, Activity-based safety risk quantification for
[1] Ministry of Housing and Urban-Rural Development (MOHURD), Available online: concrete formwork construction, J. Constr. Eng. Manag. 10 (2009) 990, https://
http://www.mohurd.gov.cn/wjfb/201903/t20190326_239913.html, 2019 doi.org/10.1061/(ASCE)CO.1943-7862.0000071.
(accessed on 2 July 2020). [29] B. Esmaeili, M.R. Hallowell, B. Rajagopalan, Attribute-based safety risk assessment.
[2] BLS, Injuries, Illnesses, and Fatalities, Available online: https://www.bls.gov/iif/, I: analysis at the fundamental level, J. Constr. Eng. Manag. 141 (8) (2015), https://
2017 (accessed on 2 July 2020). doi.org/10.1061/(ASCE)CO.1943-7862.0000980.
[3] Health and Safety Executive, Workplace Fatal Injuries in Great Britain, Available [30] S. Baradan, M.A. Usmen, Comparative injury and fatality risk analysis of building
online: http://www.hse.gov.uk/statistics/pdf/fatalinjuries.pdf, 2019 (accessed on trades, J. Constr. Eng. Manag. 132 (5) (2006) 533–539, https://doi.org/10.1061/
2 July 2020). (ASCE)0733-9364(2006)132:5(533).
[4] Q. Abueisheh, P. Manu, A.M. Mahamadu, C. Cheung, Design for safety [31] S. Choe, F. Leite, Assessing safety risk among different construction trades:
implementation among design professionals in construction: The context of quantitative approach, J. Constr. Eng. Manag. 143 (2017), https://doi.org/
Palestine, Safety, Science 128 (2020), https://doi.org/10.1016/J. 10.1061/(ASCE)CO.1943-7862.0001237.
SSCI.2020.104742. [32] Y. Sun, D. Fang, S. Wang, M. Dai, X. Lv, Safety risk identification and assessment
[5] J.A. Gambatese, Research issues in prevention through design, J. Saf. Res. 39 (2) for Beijing Olympic venues construction, J. Manag. Eng. 24 (1) (2008) 40–47,
(2008) 153–156, https://doi.org/10.1016/j.jsr.2008.02.012. https://doi.org/10.1061/(ASCE)0742-597X(2008)24:1(40).
[6] W. Creaser, Prevention through design (PtD) safe design from an Australian [33] B. Esmaeili, M.R. Hallowell, B. Rajagopalan, Attribute-based safety risk assessment.
perspective, J. Saf. Res. 39 (2) (2008) 131–134, https://doi.org/10.1016/j. II: predicting safety outcomes using generalized linear models, J. Construct. Eng.
jsr.2008.02.018. Manag. 141 (8) (2015), https://doi.org/10.1061/(ASCE)CO.1943-7862.0000981.
[7] F.A. Manuele, Prevention through design (PtD): history and future, J. Saf. Res. 39 [34] V. Sousa, N.M. Almeida, L.A. Dias, Risk-based management of occupational safety
(2) (2008) 127–130, https://doi.org/10.1016/j.jsr.2008.02.019. and health in the construction industry - part 1: background knowledge, Saf. Sci.
[8] M.D. Martinez Aires, M.C. Rubio Gamez, A. Gibb, Prevention through design: the 66 (2014) 75–86, https://doi.org/10.1016/J.SSCI.2014.02.008.
effect of European directives on construction workplace accidents, Saf. Sci. 48 (2) [35] B. Esmaeili, M. Hallowell, Integration of safety risk data with highway construction
(2010) 248–258, https://doi.org/10.1016/j.ssci.2009.09.004. schedules, Constr. Manag. Econ. 31 (6) (2013) 528, https://doi.org/10.1080/
[9] J.F. Yuan, X.W. Li, X. Xiahou, N. Tymvios, Z. Zhou, Q.M. Li, Accident prevention 01446193.2012.739288.
through design (PtD): integration of building information modeling and PtD [36] M.R. Hallowell, W. Blaney, J. Teizer, Application of sensing technology to safety
knowledge base, Autom. Constr. 102 (2019) 86–104, https://doi.org/10.1016/j. management, in: Construction Research Congress 2010: Innovation for Reshaping
autcon.2019.02.015. Construction Practice - Proceedings of the 2010 Construction Research Congress,
[10] T.M. Toole, J. Gambatese, The trajectories of prevention through design in 2010, pp. 31–40, https://doi.org/10.1061/41109(373)4.
construction, J. Saf. Res. 39 (2) (2008) 225–230, https://doi.org/10.1016/j. [37] N. Azadeh-Fard, A. Schuh, E. Rashedi, J.A. Camelio, Risk assessment of
jsr.2008.02.026. occupational injuries using accident severity grade, Saf. Sci. 76 (2015) 160–167,
https://doi.org/10.1016/j.ssci.2015.03.002.
11
Y. Lu et al. Automation in Construction 124 (2021) 103553
[38] S.R. Mohandes, H. Sadeghi, A. Mahdiyar, S. Durdyev, Assessing construction [49] P.K. Marhavilas, D. Koulouriotis, V. Gemeni, Risk analysis and assessment
labors’ safety level: a fuzzy MCDM approach, J. Civ. Eng. Manag. 26 (2) (2020) methodologies in the work sites: on a review, classification and comparative study
175–188, https://doi.org/10.3846/JCEM.2020.11926. of the scientific literature of the period 2000-2009, J. Loss Prev. Process Ind. 24 (5)
[39] The Center for Construction Research and Training, Construction Chart Book, (2011) 477–523, https://doi.org/10.1016/j.jlp.2011.03.004.
Available online: https://www.cpwr.com/sites/default/files/publications/The_6th [50] S. Azhar, M. Khalfan, T. Maqsood, Building information modelling (BIM): now and
_Edition_Construction_eChart_Book.pdf, 2018 (accessed on February 2018). beyond, Construct. Econom. Build. 12 (4) (2015), https://doi.org/10.5130/AJCEB.
[40] U.S. Department of Labor, Employment and Training Administration, O*NET, v12i4.3032.
2017. Available online: http://www.onetonline.org/ (Accessed on April 2017). [51] J.A. Gambatese, J.W. Hinze, C.T. Haas, Tool to design for construction worker
[41] M.R. Hallowell, A Formal Model of Construction Safety and Health Risk safety, J. Archit. Eng. 3 (1) (1997) 32–41, https://doi.org/10.1061/(ASCE)1076-
Management, Ph.D, Thesis, Oregon State University, Oregon State, USA, 2008, htt 0431(1997)3:1(32).
p://hdl.handle.net/1957/8541. [52] Z. Hu, J. Zhang, Z. Deng, Construction process simulation and safety analysis based
[42] O.A. Jannadi, S. Almishari, Risk assessment in construction, J. Constr. Eng. Manag. on building information model and 4D technology, Tsinghua Sci. Technol. 13 (S1)
129 (5) (2003) 492–500, https://doi.org/10.1061/(ASCE)0733-9364(2003)129:5 (2008) 266–272, https://doi.org/10.1016/S1007-0214(08)70160-3.
(492). http://hdl.handle.net/1957/8541. [53] H. Guo, Y. Yu, W. Zhang, Y. Li, BIM and safety rules based automated identification
[43] G. Reniers, W. Dullaert, B. Ale, K. Soudan, Developing an external domino accident of unsafe design factors in construction, Procedia Eng. 164 (2016) 467–472,
prevention framework: Hazwim, J. Loss Prev. Process Ind. 18 (3) (2005) 127–138, https://doi.org/10.1016/j.proeng.2016.11.646.
https://doi.org/10.1016/j.jlp.2005.03.002. [54] M.A. Hossain, E.L.S. Abbott, D.K.H. Chua, T.Q. Nguyen, Y.M. Goh, Design-for-
[44] F. Henselwood, G. Phillips, A matrix-based risk assessment approach for addressing safety knowledge library for BIM-integrated safety risk reviews, Autom. Constr. 94
linear hazards such as pipelines, J. Loss Prev. Process Ind. 19 (5) (2006) 433–441, (2018) 290–302, https://doi.org/10.1016/j.autcon.2018.07.010.
https://doi.org/10.1016/j.jlp.2005.10.005. [55] J.P. Cortés-Pérez, A. Cortés-Pérez, P. Prieto-Muriel, BIM-integrated management of
[45] P.K. Marhavilas, D.E. Koulouriotis, A risk-estimation methodological framework occupational hazards in building construction and maintenance, Autom. Constr.
using quantitative assessment techniques and real accidents’ data: application in 113 (2020), https://doi.org/10.1016/J.AUTCON.2020.103115.
an aluminum extrusion industry, J. Loss Prev. Process Ind. 21 (6) (2008) 596–603, [56] Z. Jin, J. Gambatese, D. Liu, V. Dharmapalan, Using 4D BIM to assess construction
https://doi.org/10.1016/j.jlp.2008.04.009. risks during the design phase, Eng. Constr. Archit. Manag. 26 (11SI) (2019)
[46] A.C.P. Frijters, P.H.J.J. Swuste, Safety assessment in design and preparation phase, 2637–2654, https://doi.org/10.1108/ECAM-09-2018-0379.
Saf. Sci. 46 (2) (2008) 272–281, https://doi.org/10.1016/j.ssci.2007.06.032. [57] RSMeans From Gordian, The 2020 RSMeans Building Construction Costs Book,
[47] M. Gangolells, M. Casals, N. Forcada, X. Roca, A. Fuertes, Mitigating construction Available Online, https://www.rsmeans.com/products/books/2020-building-cons
safety risks using prevention through design, J. Saf. Res. 41 (2) (2010) 107–122, truction-costs-book, 2020.
https://doi.org/10.1016/j.jsr.2009.10.007. [58] T.M. Toole, J.A. Gambatese, D.A. Abowitz, Owners’ role in facilitating prevention
[48] V. Dharmapalan, J.A. Gambatese, J. Fradella, A.M. Vahed, Quantification and through design, J. Prof. Issues Eng. Educ. Pract. 143 (2017), https://doi.org/
assessment of safety risk in the design of multistory buildings, J. Constr. Eng. 10.1061/(ASCE)EI.1943-5541.0000295.
Manag. 141 (2015), https://doi.org/10.1061/(ASCE)CO.1943-7862.0000952.
12