New Method for Measuring Safety Risk in Construction Activities: TDA

This document introduces the Task Demand Assessment (TDA) methodology, a new technique for measuring the safety risk of construction activities. The TDA quantifies the "task demand" of operations based on activity characteristics and observable risk factors, independent of worker capabilities. The paper presents findings from its initial implementation, demonstrating its feasibility on roofing and concrete paving operations. It illustrates how TDA can compare different production scenarios and measure the effect of production variables on accident potential. The methodology integrates productivity and safety risk analysis, enabling simulation and evaluation of construction operations. The paper also discusses the selection of task demand factors, limitations, and the need for further research, positioning TDA as a valuable tool for researchers and practitioners in construction operation analysis and design. The method is based on ergonomic assessment and cognitive perspectives.

New Method for Measuring the Safety Risk of Construction
Activities: Task Demand Assessment
Panagiotis Mitropoulos, Ph.D., A.M.ASCE1; and Manoj Namboodiri, M.S.2
Abstract: The task demand assessment 共TDA共is a new technique for measuring the safety risk of construction activities and analyzing
how changes in operation parameters can affect the potential for accidents. TDA is similar to observational ergonomic methods—it does
not produce estimates of probabilities of incidents, but it quantifies the “task demand” of actual operations based on characteristics of the
activity and independent of the workers’ capabilities. The task demand reflects the difficulty to perform the activity safely. It is based on
共1共the exposure to a hazard and 共2共the presence and level of observable task demand factors—that is, risk factors that can increase the
potential for an accident. The paper presents the findings from the initial implementation of TDA and demonstrates its feasibility and
applicability on two different operations: a roofing activity and a concrete paving operation. Furthermore, the paving case illustrates how
the TDA method can compare different production scenarios and measure the effect of production variables on the accident potential. The
findings indicate that the method can be applied on activities of varying complexity and can account for several risks and task demand
factors as required by the user. The selection of task demand factors is a key issue for the validity of the method and requires input from
the crew and safety management. The limitations of the methodology and the need for further research are discussed. Overall, TDA
provides a tool that can assist researchers and practitioners in the analysis and design of construction operations.
DOI: 10.1061/共ASCE共CO.1943-7862.0000246
CE Database subject headings: Safety; Risk management; Construction; Methodology .
Author keywords: Safety risk assessment; Construction safety; Task demands; Risk assessment methodology .
Introduction
Designing operations that are at the same time highly productive
and highly safe remains a significant challenge, as indicated by
the accident statistics. Productivity analysis methods such as time
studies 共Oglesby et al. 1989; Howell et al. 1993共and simulation
techniques such as CYCLONE 共Halpin and Woodhead 1976; Io-
annou 1989共and STROBOSCOPE 共Martinez and Ioannou 1999 共
provide the ability to analyze the productivity of an operation,
understand the impact of different operational variables and iden-
tify changes that improve productivity. However, our ability to
analyze or simulate construction activities with regards to safety
risks is limited. Currently, there is no methodology that allows
researchers and practitioners to measure how the safety risk varies
over the duration of the activity and evaluate how the production
variables affect the safety risk of the operation.
The long-term goal of this research is to provide methods and
techniques to assist in designing productive and safe operations.
This requires the ability 共and the methodologies 共to 共1共quantify
the safety risk of actual operations and 共2共analyze the effect of
production variables on the safety risk. Toward this goal, this
research developed the task demand assessment 共TDA共method-
ology. This is a new technique developed by the writers for mea-
suring the accident potential of an activity, based on the
production variables. The TDA method is based on a cognitive
perspective—the premise that the attributes of the task and the
environment affect the likelihood of accidents during a construc-
tion operation 共. Methodologically, the method is based on ergo-
nomic assessment methods; it quantifies the task demands due to
the risk factors involved in the operation and independent of the
workers’ abilities and human factors. In addition to quantifying
the task demands, this methodology integrates the analysis of pro-
ductivity with the analysis of safety risk, based on operational
variables. The method enables simulation of construction opera-
tions and analysis of different operations for both productivity
and safety risk.
The paper introduces the method and demonstrates its feasibil-
ity and applicability for different activities and risks. For this, the
TDA method is used to analyze the safety risk of two cases: a
roofing operation and a concrete paving operation. The discussion
summarizes the contribution of the method to research and prac-
tice and identifies its limitations.
Background
This section reviews current methods to measure the safety risk of
an operation and identifies the strengths and limitations of the
different methods. Such methods can be grouped in three main
categories: 共1共 a priori risk estimates using expert opinions or
statistical data; 共2共 compliance measures of safe conditions and
behaviors; and 共3共methods to calculate ergonomic and cognitive
1Assistant Professor, Del E. Webb School of Construction, Arizona
State Univ., P.O. Box 870204, Tempe, AZ 85287. E-mail: takism@
asu.edu
2Senior Consultant, Faithful and Gould; formerly, Graduate Research
Assistant, Del E. Webb School of Construction, Arizona State Univ.
Note. This manuscript was submitted on June 5, 2009; approved on
June 15, 2010; published online on June 16, 2010. Discussion period
open until June 1, 2011; separate discussions must be submitted for indi-
vidual papers. This paper is part of the Journal of Construction Engi-
neering and Management, Vol. 137, No. 1, January 1, 2011. ©ASCE,
ISSN 0733-9364/2011/1-30–38/$25.00.
30 / JOURNAL OF CONSTRUCTION ENGINEERING AND MANAGEMENT © ASCE / JANUARY 2011
J. Constr. Eng. Manage. 2011.137:30-38.
Downloaded from ascelibrary.org by UNIVERSITE LAVAL on 07/02/14. Copyright ASCE. For personal use only; all rights reserved.

demands. Observational methods are reviewed in more detail as
they provide the methodological basis for the TDA method.
Risk Estimates
According to the National Safety Council 共2009共“risk is a mea-
sure of the probability and severity of adverse effects.” Methods
to quantify the safety risk of construction operations have focused
on assessment of the probability 共frequency共of incidents and con-
sequences. For example, Brauer 共1994共quantified the frequency
of event occurrence in subjective levels such as frequent, prob-
able, occasional, remote, and improbable. These methods differ
primarily in the approach they use to assess the probability and
severity. The two main approaches are based on expert opinions
and statistical data.
Several researchers proposed subjective expert assessments to
quantify the safety risk. Methods using expert opinions differ in
the factors that they take into consideration. For example, Sun et
al. 共2008共 used the analytic hierarchy process and quantified 25
risk factors 共e.g., schedule pressure from client 共by asking expert
participants to rate each component on a subjective Likert scale.
Lee and Halpin 共2003共developed a fuzzy logic system that uses
expert input with regards to three factors: preplanning, training,
and supervision. Jannadi and Almishari 共2003共 used risk scores
provided by the user to populate a risk assessor model with se-
verity, probability, and exposure inputs. Yi and Langford 共2006共
proposed that the risk of an activity depends on the process risk,
human resources risk, technology risk, and environment risk. Hal-
lowell and Gambatese 共2009共used expert opinion to quantify the
probability and consequences of incident events for formwork
activities. Everett 共1999共 analyzed ergonomic risks associated
with different construction processes 共e.g., install drywall and
light gauge steel partitioning共by identifying seven ergonomic risk
factors and having experts assess the presence of such factors on
a scale of 1 共low共to 3 共high共.
Other researchers have used statistical injury data to develop a
priori estimates of safety risks. Baradan and Usmen 共2006共quan-
tified the risk for various construction trades using data published
by the Bureau of Labor Statistics. Wang et al. 共2006共 estimated
the safety risk on projects using the amount of labor hours with
the probability and costs 共consequences共 for different incident
types 共such as falls, struck by, etc.共which were calculated based
on statistical injury data from BLS.
The aforementioned methods provide an a priori estimate of
the activity risk. These assessments, however, are independent of
how the operation is actually performed. To use a cost control
analogy, these systems provide estimates but not actual costs and
do not capture or describe the actual safety risk over the duration
of the operation. Ways to assess the safety risk of actual/observed
operations include compliance metrics and ergonomic methods
共which, however, are limited to ergonomic hazards 共.
Compliance Measurements
Risk assessment is an important part of safety management. In
construction a typical approach to risk assessment consists of haz-
ard identification before the activity, such as job hazard analysis
and identification of the safety measures needed. Evaluations of
workplace safety have focused primarily on observations of un-
safe conditions and behaviors. Such methods typically measure
the frequency 共percent共 of safe-unsafe behaviors or conditions.
Such metrics are also used in behavior based safety 共Krause 1997;
Geller 2005共. Such assessments are independent of operational
variables as they do not account for the production factors that
increase the unsafe behaviors.
Ergonomic Assessment Methods
Several ergonomic methods have been developed to evaluate the
potential for musculoskeletal disorders 共MSDs共based on charac-
teristics of the operation. Such methods evaluate the physical de-
mands on workers based on specific characteristics of the activity
共physical loads, posture, etc.共. These methods do not calculate the
likelihood of injuries 共as this depends on many other factors 共, but
assess the ergonomic task demands, which create the potential for
injury. Depending on the data collected, ergonomic assessment
methods are of three types: 共1共 self-reports; 共2共 physiological
measurements; and 共3共 observational methods 共Li and Buckle
1999; David 2005 共.
Self-reports from workers can be used to collect data on work-
place exposure to both physical and psychosocial factors by using
methods that include worker diaries, interviews, and question-
naires. Physiological measurements using monitoring instruments
that rely on sensors attached directly to the subject for the mea-
surement of exposure variables at work 共David 2005 共. Such as
heart rate elevations, percentage of available heart range rate,
oxygen consumption, electromyography, etc. 共Abdelhamid and
Everett 2002; Bernold et al. 2001; Saurin and Guimarães 2006;
Chang et al. 2009共.
Observational methods calculate the ergonomic demand based
on observation of key ergonomic risk factors for different body
parts. Observational methods vary in terms of complexity 共David
2005共and the risk factors they account for. The main risk factors
related to MSDs are the load, the posture, and the frequency, but
others are often accounted for such as task duration, vibration,
etc.
Observational methods include the Ovako working-posture
analysis system 共OWAS共 共Karhu et al. 1977共, the cube model
共Kadefors 1994, 1997; Sperling et al. 1993; Rwamamara 2007 共,
Ergo-SAM 共Laring et al. 2002; Christmansson et al. 2000共, and
the quick exposure check 共QEC共 共David et al. 2008 共. OWAS is a
work-sampling technique that enables an observer to assess and
record data on a number of factors using specifically designed
proforma sheets. OWAS evaluates the ergonomic workload of
different body parts, based on the posture and load. Buchholz et
al. 共1996共 developed a modified OWAS for construction called
PATH 共posture, activity, tools, and handling共as a work sampling
method to evaluate ergonomic risk factors.
The cube model calculates ergonomic demand based on the
assessment of three risk factors: 共1共the force that a worker exerts;
共2共the work posture; and 共3共the frequency of the force exertion.
Criteria are formulated to identify in operative terms what consti-
tutes low, moderate, and high demands for each factor. The model
assigns a score of 1 for low demand, 2 for moderate, and 3 for
high demand. The overall ergonomic demand is calculated as the
product of the three scores: Total demand= force⫻ posture
⫻ frequency—the score can range from 1 to 27. The demand is
considered acceptable if less than 5, conditionally acceptable if
between 5 and 10 and unacceptable if greater than 10. Ergo-SAM
is a more recent methodology based on the cube model developed
to assist industrial production engineers predict the ergonomic
demands of production activities at the design stage of the pro-
duction 共Laring et al. 2002 共. QEC was developed in the U.K. and
JOURNAL OF CONSTRUCTION ENGINEERING AND MANAGEMENT © ASCE / JANUARY 2011 / 31
J. Constr. Eng. Manage. 2011.137:30-38.
Downloaded from ascelibrary.org by UNIVERSITE LAVAL on 07/02/14. Copyright ASCE. For personal use only; all rights reserved.

calculates the ergonomic demands as a combination of the above
three factors, as well as additional factors such as vibration and
duration 共David et al. 2008共.
Cognitive Assessment Methods
In cognitive systems engineering, techniques have been devel-
oped to assess the mental load on operators working with human-
machine systems. Four techniques are commonly used to assess
mental workload: physiological measures, subjective rating mea-
sures, primary task performance measures, and secondary task
measures 共Meshkati and Loewenthal 1988共. The most widely
used techniques use subjective workload assessments. Subjective
mental workload is defined as the subject’s direct estimate or
comparative judgment of the mental or cognitive workload expe-
rienced at a given moment 共Reid and Nygren 1988 共. Such tech-
niques are the National Aeronautics and Space Administration
task load index 共Hart and Staveland 1988 共 and the subjective
workload assessment technique 共Reid and Nygren 1988 共. Even
though much effort has been made to develop objective measures
of workload, subjective workload assessment techniques continue
to be popular due to their ease of use, general nonintrusiveness,
low cost, high face validity, and known sensitivity to workload
variations 共Reid and Nygren 1988 共.
Limitations of Risk Assessment Methods
In summary, the existing methods for assessing safety risk can
provide a priori risk estimates and can measure the frequency of
unsafe behaviors or conditions. These techniques do not provide a
way to assess the potential for accidents based on the actual ex-
ecution of the operation. Ergonomic methods provide this capa-
bility but only for ergonomic hazards, and not for traumatic injury
incidents, such as falls, struck by, etc. To address this limitation,
this research develops the TDA methodology.
Task Demands Assessment Methodology
Background
The TDA methodology is a new technique developed by the writ-
ers in order to analyze the effect of production variables on the
accident potential of an activity 共Namboodiri 2007共. This was
needed in order to analyze construction operations for both pro-
ductivity and safety. With regards to productivity, time studies
共Oglesby et al. 1989共were used to analyze an operation and iden-
tify changes to improve productivity. However, there was no spe-
cific methodology to indicate if the accident potential of the
operation changed. Thus, the goal of the initial study was to de-
velop an analytical method to quantify the activity risk and com-
pare similar operations with regards to their accident potential.
The methodology was developed based on observation and
analysis of three concrete operations: 共1共 a concrete paving op-
eration; 共2共 a formwork operation for concrete diaphragm walls
on a bridge; and 共3共 formwork operation 共Namboodiri 2007共.
Later, the method was used to analyze the safety risk of other
activities—roofing, framing, etc. The paper reports the results
from the early use of the TDA method—the results identify the
strengths and limitations of the methodology and indicate direc-
tions for further research.
Theoretical Foundation
The TDA methodology has two key points of departure: 共1共with
regards to methodology, TDA is based on observational methods
for ergonomic assessment reviewed in the previous section and
共2共 theoretically, TDA is based on the task demand-capability
model for construction safety 共Mitropoulos et al. 2009共. Accord-
ing to that model, when a worker is exposed to a hazard, the
likelihood of incidents depends on the task demands and applied
capabilities—it increases when task demands increase and re-
duces when applied capability increases. As shown in Fig. 1, task
demands depend on factors related to the task 共support conditions,
tools, loads, etc.共the environment 共other activities, weather con-
ditions共, and worker’s behaviors 共postures, etc.共. The applied ca-
pability depends on the worker’s skill and capability, human
factors 共such as fatigue, etc.共 and the level of activation. The
purpose of the TDA method is to develop an objective assessment
of the task demands with regards to specific hazards, independent
of the workers’ applied capability.
Key Characteristics
With regards to scope, this method focuses on traumatic injuries
and does not capture the risks arising from overexertion injuries,
physical fatigue, and/or occupational illnesses. Its main purpose is
to be used for detailed analysis of activities in the same way that
time studies are used to analyze/improve productivity, and ergo-
nomic studies are used to analyze/reducing ergonomic loads. The
TDA method has the following key characteristics: it is an obser-
vational method that provides an objective assessment of an ac-
tivity’s task demand based on observable risk factors. First, TDA
provides an assessment of an activity’s task demand based on
observable risk factors. The method does not provide estimates of
safety risk in terms of probabilities and consequences. It quanti-
fies the task demand—that is, the difficulty to do the work safely
based on observable risk factors of the specific operation. Second,
it is an observational method that enables detailed analysis of an
activity. Based on observation/videotaping of the operation, it cal-
culates the accident potential over the duration of the activity for
the risks examined, thus producing a “risk profile” of the actual
activity. Third, TDA is an objective assessment 共as opposed to a
subjective assessment 共because it quantifies the potential for ac-
cident independent of the capabilities of the worker performing
the task.
Fig. 1. Task demand-capability model of construction safety
共adapted from Mitropoulos et al. 2009共
32 / JOURNAL OF CONSTRUCTION ENGINEERING AND MANAGEMENT © ASCE / JANUARY 2011
J. Constr. Eng. Manage. 2011.137:30-38.
Downloaded from ascelibrary.org by UNIVERSITE LAVAL on 07/02/14. Copyright ASCE. For personal use only; all rights reserved.

Analysis Steps
To calculate the task demands the TDA method involves the fol-
lowing steps.
Understand the Operation
Interviews with supervisors and crews develop an understanding
of the operation—the layout, estimated productivity, crew com-
position, preparation, work distribution and tasks, coordination
needs, risks and hazards, etc.
Identify the High Consequence Incidents and Hazards
Ergonomic methods analyze the ergonomic demands separately
for different ergonomic incidents—back, shoulders, etc. In a simi-
lar way, the TDA analyzes the task demand separately for differ-
ent potential incidents. For example “fall from elevation” is the
potential incident; the unprotected edge is the physical hazard. In
the cases used to develop the method, the researchers focused on
the incident events with the highest severity potential. These are
the potential injuries with the highest cost 共individual suffering
and claims cost for the organization 共. These potential incidents
are identified through review of the task hazard analysis per-
formed by crew, and discussions with the supervisor and crew
about what they consider the most significant risks.
Identify the Conditions of Exposure
The exposure to a hazard is calculated as the percentage of the
time that a worker is exposed to the relevant hazard. The percent-
age of time is used to normalize and compare different operations,
and is consistent with measurements in ergonomics methods. For
each hazard, the analyst and practitioners need to define what
conditions constitute “exposure to the hazard.” This can be based
on relevant regulations—for example, workers were considered
exposed to a fall hazard if they were less than 6 ft away from an
elevated edge or opening. The exposure conditions must be ob-
servable and unambiguous 共for example, 6 ft from the edge 共, so
that it does not depend on the observer’s judgment. Safety mea-
sures often reduce exposure to hazards; for example, fall arrest
equipment protects workers from falls. The TDA method can
measure the accident potential with or without the safety mea-
sures used—for example, with regards to the fall from elevation
hazard, the method can calculate the exposure of a worker to
heights above 6 ft, with or without the use of fall arrest equip-
ment.
Identify Task Demand Factors and Values
The potential for accident depends on the presence and level of
risk factors 共task demand factors 共 that relate to each risk exam-
ined. A central issue in the TDA method is the identification of the
risk factors that determine the task demand. In ergonomics the
three primary risk factors for MSDs are force, work posture, and
frequency. However, with regards to traumatic injuries 共falls,
struck by, etc.共there is no established framework of risk factors.
The risk factors depend on the hazard examined and the opera-
tion; in the TDA method, these factors are determined by observ-
ing and analyzing the operation and interviewing the activity
participants. For example, for a roofing activity the primary risk is
fall from elevation and risk factors related to falls include the roof
slope, worker’s movement, etc.
In the TDA method each task demand factor is assigned a
value of low, moderate or high. The numerical values used are 1
for low, 3 for moderate and 9 for high. This approach is similar to
the cube model and other ergonomic methods. In order to assess
the task demand, the analyst needs to specify the operative criteria
that determine if a task demand is low, moderate, or high. For
example, if a task demand factor is “distance from the unpro-
tected edge,” the task demand values may be “high” if the dis-
tance is less than 6 ft, moderate if the distance is between 6 and
10 ft, and low if the distance is greater than 10 ft. Such conditions
need to be specified in an unambiguous way, to minimize ob-
server bias. The criteria for low, moderate and high task demand
should be developed with input from the work crew.
Calculate the Task Demand
Once the exposure and task demand factors have been defined,
the analyst reviews the video of the operation and records how the
task demand values for each factor change over time. This pro-
duces a graph that indicates the level of task demand over time
for each task demand factor. When there are more than one task
demand factors, the overall task demand is calculated as the sum
of the individual task demands. The accidents potential for a par-
ticular hazard is calculated as the task demand level multiplied by
the percent of exposure time. For example, if 50% of the time the
task demand is low 共1共, 30% of the time it is moderate 共3共 and
20% of the time it is high 共9共, the overall task demand would be
0.50⫻ 1 + 0.30⫻ 3 + 0.20⫻ 9 = 3.2. This reflects the average task
demand during the activity.
The two case studies that follow illustrate how TDA quantifies
the safety risk for different activities and risks. The paving case
also illustrates how TDA can analyze different production sce-
narios and compare them with regards to productivity and safety
risk.
Roofing Operation
This case study analyzes the accident potential of a residential
roofing activity. The activity is the installation of three 10-ft units
of flashing, by two workers on a two story house with roof slope
greater than 5:12. The activity involved three tasks: transporting
the material, preparing the material for installation by shaping
them for easier positioning, and installing them by nailing them at
the eave. The duration of the three cycles was 4 min and 10 s.
Identify Key Hazards
The primary risk in the roofing operation is fall from elevation.
Determine Exposure Conditions
The roofers were considered exposed to the fall hazard the entire
time they were on the roof. The crew did not use any perimeter
protection or fall arrest equipment 共these measures are not re-
quired in residential construction operations 共.
Identify Task Demand Factors
Discussions with the crew identified three main task demand fac-
tors that affect the difficulty of performing the work safely: 共1共
roof slope; 共2共 distance from the edge; and 共3共 workers’ move-
ment. Table 1 summarizes the conditions that affect the level of
task demand for each factor.
JOURNAL OF CONSTRUCTION ENGINEERING AND MANAGEMENT © ASCE / JANUARY 2011 / 33
J. Constr. Eng. Manage. 2011.137:30-38.
Downloaded from ascelibrary.org by UNIVERSITE LAVAL on 07/02/14. Copyright ASCE. For personal use only; all rights reserved.

Calculate Accident Potential
Using the video of the activity, the task demand level for each
factor was identified for the duration of the work. Fig. 2 illustrates
the task demand factors and the cumulative task demand for one
worker. The combined task demand is a sum of the individual task
demand factors. The method provides a way to capture and quan-
tify the presence and level of task demands factors, and the task
demand score reflects the overall level of difficulty.
Concrete Paving Operation
This case study analyzes a concrete paving operation. The analy-
sis first examines the productivity and safety risk of the observed
operation. Then it analyzes the productivity and accident potential
for different production variables in order to identify the effect of
production changes on accident potential and productivity.
The project involved upgrading of airport taxiways from as-
phalt concrete to Portland cement concrete pavement. The opera-
tion studied was the paving of one section performed in one night
shift 8 h long. The section was 300–350 ft long, 37.5 ft wide by
19.5 in. deep and lied between two paved sections. The concrete
was supplied by a batching plant at the airport and was delivered
to the paving area by eight trucks. The paving operation involved
three distinct activities: 共1共 placing the concrete in front of the
paver; 共2共 paving; and 共3共 concrete finishing behind the paver.
The three activities needed to stay relatively close to each other,
so that the speed of the paver and the workability of the concrete
were not affected.
Fig. 3 illustrates the layout of the operation and the crew in-
volved. The section was divided in two lanes where the trucks
backed up and dumped the concrete. The concrete delivery rate
was one truck every 2.5 min at each lane. Each truck load pro-
vided concrete for approximately 4 ft in front of the paver in each
lane. During the operation, the paver reached and maintained a
speed of 1 ft per minute. The paving crew consisted of 10 workers
共not counting the superintendent, the paver operator, and the me-
chanical foreman共. The general foreman supervised the entire
crew and was in constant communication with the batching plant.
Placing the Concrete in Front of the Paver
The two laborers and the superintendent performed the following
tasks: 共1共 the laborers guided the trucks as they back up to the
right location, opened the truck gate, and guided the truck to
dump the concrete. The cycle time of dumping the concrete var-
ied from 52 to 72 s 共62 s on average 共. Each laborer was assigned
one lane in which to guide the trucks; 共2共set the dowel baskets
every 8 ft 共which is every two trucks for each side 共. It took two
people to move the dowel baskets in place; if the other laborer
was busy guiding a truck, the first laborer had to wait or the
superintendent was helping. After setting the dowel baskets, one
laborer fixed the dowel basket onto the ground with a nail gun.
Setting the baskets took 14 s on average and fixing the baskets
took 60 s.
The superintendent supervised this activity. His primary con-
cern was to maintain the smooth and continuous delivery of con-
crete so that the paver, once started, did not stop. Another
Table 1. Task Demand Factors and Conditions Determining Level of Task Demands
Task demand factor
Task demand level
Low 共value= 1共 Moderate 共value= 3共 High 共value= 9共
Roof slope No slope Slopeⱕ 5 : 12 Slopeⱖ 5 : 12
Distance from edge At ridge More than 6 ft from edge Less than 6 ft from edge
Body movement Stationary Moving forward Moving backward
Fig. 2. Task demands for roofing activity
Fig. 3. Layout of paving operation
34 / JOURNAL OF CONSTRUCTION ENGINEERING AND MANAGEMENT © ASCE / JANUARY 2011
J. Constr. Eng. Manage. 2011.137:30-38.
Downloaded from ascelibrary.org by UNIVERSITE LAVAL on 07/02/14. Copyright ASCE. For personal use only; all rights reserved.

important concern was to support the laborers in front of the
paver—the superintendent sometimes helped guide the trucks and
place dowel baskets.
Paving
The speed of the paver was 1 ft per minute. The paver operator
took instructions from the superintendent regarding the speed of
the paver.
Finishing behind the Paver
The finishing crew vibrated and finished the concrete. The crew
included four finishers, two laborers, and a finishing foreman who
performed the following tasks: vibrated and troweled the con-
crete, checked paver alignment and concrete edge, checked the
concrete surface, and added or removed excess concrete.
Productivity and Safety Analysis
Because of the linear nature of the project the productivity in this
case was expressed in linear feet of section paved per minute.
Thus, the productivity of the operation was indicated by the speed
of the paver, which was 1 ft per minute. The TDA included the
following steps.
Identify Key Risks
Before the operation the superintendent, foreman, and crew per-
formed a task hazard analysis and discussed safety issues. The
most significant injury risks were identified in the activity “plac-
ing concrete”: 共1共getting hit by the moving truck and 共2共getting
hit by the paver. Thus, the analysis focused these two potential
incidents and the two laborers in front of the paver.
Determine Exposure Conditions
For the hazard “hit by paver,” it was considered that exposure
exists when the worker is less than 6 ft from the paver. Because
only a portion of the operation was videotaped 共about 30 min
which also included the finishing 共, the workers’ actual positions
were not recorded throughout the operation. The observations in-
dicated that in general, the workers were standing near the end of
the concrete buffer. This was always the case when the workers
were guiding the trucks; for the rest of the time it is a conserva-
tive estimate. Hence, this assumption was used for the analysis.
The length of concrete in front of the paver was calculated based
on the truck arrival rate and the paver’s speed; each truck added 4
ft of concrete, and every minute the concrete was reduced by 1 ft.
For the hazard “hit by truck,” the exposure to the hazard starts
from the moment when the truck starts backing up, to the point
when the truck leaves. The rest of the time, the workers are not
exposed to this hazard.
Determine the Task Demand Factors
For the hazard hit by paver, it was considered that the task de-
mand depends on the distance between the worker and the
paver—the smaller the distance, the greater the task demand. If
the distance was between 4–6 ft, the task demand was considered
low, between 2–4 ft moderate, and less than 2 ft high. For the
hazard hit by truck, the task demand was determined as follows,
based on observations of the operation and discussion with the
superintendent. When a laborer guided the truck, the worker’s full
attention was at the truck and the worker and the driver were in
direct communication. In this case, the task demand was consid-
ered low 共1共. If the truck was backing up while the laborer was
performing another task behind the truck 共setting the dowel bas-
kets共 and the superintendent was monitoring the truck, then the
task demand was considered moderate 共3共. In two instances the
superintendent stopped the truck until the laborer finished. Fi-
nally, if the truck was backing up while the laborer was behind
the truck performing another task and no one else was monitoring
the truck, then the task demand was considered high 共9共.
Calculate the Task Demand
Fig. 4 shows the task demand charts for Laborer B. The analysis
uses the average task durations. Truck arrives in each lane every
2.5 min. In the analysis, the truck waits for the laborer to guide it;
hence, the task demand for hit by truck is always low. The task
demand for hit by paver depends on the workers distance from the
paver.
As shown in Fig. 4, the demands for both hazards may in-
crease simultaneously; for example, in minute 4:00–5:00, the
worker is very close to the paver while at the same time the truck
is backing up. The combined presence of task demand factors
may further increase the task demand; while the attention of the
Fig. 4. Task demands for Laborer B
JOURNAL OF CONSTRUCTION ENGINEERING AND MANAGEMENT © ASCE / JANUARY 2011 / 35
J. Constr. Eng. Manage. 2011.137:30-38.
Downloaded from ascelibrary.org by UNIVERSITE LAVAL on 07/02/14. Copyright ASCE. For personal use only; all rights reserved.

worker is on the truck, this may increase the risk of hit by paver.
Table 2 summarizes the overall task demand for each hazard and
laborer.
Analysis of Alternative Scenarios
The TDA method can be used to analyze how changes in the
production variables 共paver speed, truck arrival, and work rules 共
would affect productivity and safety. To illustrate this, the task
demand was calculated for three different production scenarios as
shown in Table 3. In each scenario, the differences in paver speed
and truck arrival resulted in different task demands.
Scenario 2 indicates increased risk of hit by paver because the
greater speed of the paver results in a smaller concrete buffer. In
Scenario 3, trucks arrive more frequently, hence the safety risk
increased due to increased exposure to trucks. Often the trucks
arrive while the laborers are setting or fixing the dowel baskets.
However, the trucks wait for the worker to guide them to position.
Thus, the task demands due to trucks did not increase 共the task
demand value remained 1 共. In Scenario 4, the trucks do not wait
for the workers to guide them. As a result, there is increased risk
as the trucks back up while the laborers are fixing the dowel
baskets. In the actual operation, this problem is addressed orga-
nizationally, having the superintendent watch out for the laborers
as an important part of his functions. This element was not in-
cluded in Scenario 3b, but it could be easily addressed in another
scenario. Fig. 5 summarizes the productivity 共paver speed 共 and
accident potential of the scenarios analyzed.
Discussion
The TDA methodology provides a new way to measure the safety
risk of construction activities. However, the method does not es-
timate the safety risk in terms of probability and consequences,
but it quantifies an activity’s safety difficulty 共task demand共based
on characteristics of the task. In a way parallel to ergonomic
methods, it measures the task demand based on the exposure and
level of risk factors, and independent of the workers capability.
Flexibility
The TDA method can be used for different activities and risks. A
key element of the methodology is the identification of the risk
factors affecting the task demand for each hazard. Because each
activity analyzed may involve different hazards and risk factors,
these factors are not predetermined 共as in the ergonomic meth-
ods共, but need to be identified in every case. The number of haz-
ards and task demand factors that need to be considered depends
on the complexity of the activity analyzed. However, the basic
method remains the same independent of the number of factors
considered.
Comprehensiveness
In terms of comprehensiveness, that is the task demand factors
taken into consideration, the method allows a great degree of
flexibility to the user to analyze as many hazards and factors as
needed.
Validity
The validity of the method depends largely on the selection of the
task demand factors. This should be done with input from the
production and safety personnel as this will assure that the impor-
tant factors are accounted for, and will ensure the validity of the
analysis. Furthermore, establishing objective criteria for low,
moderate, and high demand values reduces observer’s bias.
Similar to ergonomics methods, validation of the TDA results
is a challenge. Ideally, the validation of the method would require
comparison of the results with an accepted way to measure the
task difficulty. Currently, however, there is no other objective
metric of task demand for traumatic injury risks. Thus, the main
alternative is to use the subjective assessment of the workers who
experience the task demands. For this, it is important to use the
crew’s input with regards to the task demands factors.
Contribution to Operations Improvement
The TDA method provides a systematic approach for identifying
interventions and safety improvements. First, it clearly identifies
Table 2. Task Demand for Each Hazard and Worker
Worker
Risk
Hit by truck Hit by paver Combined
Laborer A 40%⫻ 1 = 0.40 25%⫻ 1+25% ⫻ 3 = 1.000 1.40
Laborer B 40%⫻ 1 = 0.40 30%⫻ 1+30% ⫻ 3 + 7.5%⫻ 9 = 1.875 2.275
Total 0.80 2.875 3.675
Table 3. Scenarios Analyzed and Summary of Results
Scenario
Paver speed
共ft/min共
Truck arrival per side
共min共 Who guides truck Safety risk “truck” Safety risk “paver” Combined task demand
Actual case 1.0 2.5 Worker 0.80 2.875 3.675
Scenario 2 1.5 2.5 Worker 0.80 6.100 6.900
Scenario 3 1.5 2.0 Worker 1.00 3.850 4.850
Scenario 4 1.5 2.0 No one 2.20 2.875 5.075
36 / JOURNAL OF CONSTRUCTION ENGINEERING AND MANAGEMENT © ASCE / JANUARY 2011
J. Constr. Eng. Manage. 2011.137:30-38.
Downloaded from ascelibrary.org by UNIVERSITE LAVAL on 07/02/14. Copyright ASCE. For personal use only; all rights reserved.

conditions of exposure and the task demand factors. This creates
a deeper understanding of the operation, the key safety variables
and potential trade-offs between productivity and safety. Second,
it increases understanding of the aspects of the operation that can
reduce the exposure to or the level of the task demand factors.
This guides safety improvements efforts toward changes in the
work that 共1共reduce exposures and/or 共2共reduce the level of task
demands. Third, it provides a tool to evaluate and compare the
accident potential of alternative production designs as the paving
case illustrated. The method provides a way to capture and quan-
tify the presence and level of task demands factors and the task
demand score reflects the overall level of “safety difficulty” for
the specific operation parameters. The score is not meant to be an
accurate measure of task demand.
Implementation
The TDA method can be used by contractors to analyze selected
operations or to compare different production scenarios, in the
same way they use time studies for productivity analysis or ergo-
nomic studies. The method is not expected to be used for continu-
ous monitoring of operations. Despite the variability and
dynamism of construction activities, the practical value of the
TDA method is the identification of and attention to the underly-
ing task demand factors. However, if the work method or situa-
tion changes in a way that new significant task demand factors are
present, then the activity will have to be analyzed again.
The primary users will by the individuals or groups in the
contractor’s organization who perform the productivity improve-
ment or ergonomic studies 共such as internal analysts or safety
professionals共 or external consultants. Implementation will re-
quire some basic training in the method, and the time to videotape
the operation and perform the analysis.
The methodology has the following limitations and areas of
difficulty:
• Scope. As discussed earlier, this method focuses on traumatic
injuries and does not capture the risks arising from overexer-
tion injuries, physical fatigue, and/or occupational illnesses.
• Task demand values versus probabilities. As discussed, the
TDA method does not estimate the probability that an incident
will occur, but provides a measure of an activity’s accident
potential based on task demand factors. The method does not
correlate the task demand values with probability of incidents.
This is beyond the scope of the methodology. Furthermore, the
extent to which each factor contributes to the probability of
incidents is also beyond the scope of the methodology and
subject of future safety research. Even with this limitation the
methodology can be used to compare the accident potential
between different designs of the same operations and evaluate
how changes in the activity can reduce the accident potential.
• Effect of multiple task demand factors. The effect of multiple
task demand factors is captured by adding the task demand
values. However, the extent to which the difficulty increases
with the combination of risk demand is beyond the scope of
the TDA method and subject of further safety research.
• Effect of multiple hazards. Similarly, the presence of multiple
hazards may also increase the likelihood of incidents, as the
workers may have to divide their attention 共which is a limited
resource共 to the task and the multiple hazards. The effect of
combined hazards on the safety risk needs to be further inves-
tigated, as the safety risk may increase disproportionately
when two or more task demand factors occur simultaneously.
• Task demand metric. The calculation of the overall task de-
mand reflects the average task demand over the duration of the
activity. Another metric for the accident potential is the fre-
quency of unacceptable task demand levels, that is, the percent
of time that the task demand exceeds some limit of acceptabil-
ity. For example, in the cube method, for the three ergonomic
risk factors 共load, posture, and frequency 共, the demand can
range from 1 to 27 and the acceptable level is set at 5. The
level of acceptability would have to be determined by the
management and the crew. The value of the methodology is
that it brings all these safety issues into focus.
Despite the limitations, the TDA methodology provides a use-
ful tool that researchers and practitioners can use to measure and
improve both the productivity and safety of construction opera-
tions. The TDA method provides researchers and practitioners
with a tool for analyzing the accident potential of different pro-
duction designs and identifies how changes in the operation affect
the accident potential. This ability will facilitate the design of
safer and more productive operations.
Acknowledgments
The writers would like to thank the management and field person-
nel of the participating crews for taking the time to explain their
work. The research described in this paper is conducted with the
support of NSF and the CAREER Award, Grant No. 0645139.
References
Abdelhamid, T. S., and Everett, J. G. 共2002共. “Physiological demands
during construction work.” J. Constr. Eng. Manage., 128共5共, 427–
437.
Baradan, S., and Usmen, M. A. 共2006共. “Comparative injury and fatality
risk analysis of building trades.” J. Constr. Eng. Manage., 132共5共,
533–539.
Bernold, L. E., Lorenc, S. J., and Davis, M. L. 共2001共. “Technological
intervention to eliminate back injury risks for nailing.” J. Constr. Eng.
Manage., 127共3共, 245–250.
Brauer, R. L. 共1994共. “Risk management and assessment.” Safety and
health for engineers, Van Nostrand Reinhold, New York, 543–572.
Buchholz, B., Paquet, V., Punnett, D., Lee, D., and Moir, S. 共1996共.
“PATH: A work sampling-based approach to ergonomic job analysis
for construction and other non-repetitive work.” Appl. Ergon, 27共3共,
177–187.
Chang, F., Sun, Y., Chuang, K., and Hsu, D. 共2009共. “Work fatigue and
physiological symptoms in different occupations of high-elevation
construction workers.” Appl. Ergon, 40共4共, 591–596.
Christmansson, M., Falck, A.-C., Ampazis, J., Forsman, M., Rasmusson,
Fig. 5. Productivity and task demands for the scenarios examined
JOURNAL OF CONSTRUCTION ENGINEERING AND MANAGEMENT © ASCE / JANUARY 2011 / 37
J. Constr. Eng. Manage. 2011.137:30-38.
Downloaded from ascelibrary.org by UNIVERSITE LAVAL on 07/02/14. Copyright ASCE. For personal use only; all rights reserved.

L., and Kadefors, R. 共2000共. “Modified method time measurements
for ergonomic planning of production systems in the manufacturing
industry.” Int. J. Prod. Res., 38共17共, 4051–4059.
David, G., Woods, V., Li, G., and Buckle, P.共2008共. “The development of
the quick exposure check 共QEC共for assessing exposure to risk factors
for work-related musculoskeletal disorders.” Appl. Ergon, 39共1共, 57–
69.
David, G. C. 共2005共. “Ergonomic methods for assessing exposure to risk
factors for work-related musculoskeletal disorders.” Occup. Med.
(Chic. Ill), 55共3共, 190–199.
Everett, J. 共1999共. “Overexertion injuries in construction.” J. Constr. Eng.
Manage., 125共2共, 109–114.
Geller, E. S. 共2005共. “Behavior-based safety and occupational risk man-
agement.” Behav. Modif, 29共3共, 539–561.
Hallowell, M. R., and Gambatese, J. A. 共2009共. “Activity-based safety
risk quantification for concrete formwork construction.” J. Constr.
Eng. Manage., 135共10共, 990–998.
Halpin, D. W., and Woodhead, R. W. 共1976共. Design of construction and
process operations, Wiley, New York.
Hart, S. G., and Staveland, L. E. 共1988共. “Development of NASA-TLX
共task load index 共: Results of empirical and theoretical research.”
Human mental workload, P. A. Hancock and N. Meshkati, eds., North
Holland, Amsterdam, The Netherlands, 139–183.
Howell, G. A., Laufer, A., and Ballard, G. 共1993共. “Interaction between
subcycles: One key to improved methods.” J. Constr. Eng. Manage.,
119, 714.
Ioannou, P. G. 共1989共. “UM-CYCLONE discrete event simulation sys-
tem, reference manual.” Rep. No. UMCE-89-11, Dept. of Civil Engi-
neering, Univ. of Michigan, Ann Arbor, Mich.
Jannadi, O., and Almishari, S. 共2003共. “Risk assessment in construction.”
J. Constr. Eng. Manage., 129共5共, 492–500.
Kadefors, R. 共1994共. “An ergonomic model for workplace assessment.”
12th Triennial Congress of the Int. Ergonomics Association, Vol. 5,
Human Factors Association of Canada, Toronoto, 210–212.
Kadefors, R. 共1997共. “Evaluation of working situations using the cube
model approach.” Proc., 13th Triennial Congress of the Int. Ergonom-
ics Association, Vol. 4, Finnish Institute of Occupational Health, Hel-
sinki, 174–176.
Karhu, O., Kansi, P., and Kuorinka, I. 共1977共. “Correcting working pos-
tures in industry: A practical method for analysis.” Appl. Ergon, 8,
199–201.
Krause, T. 共1997共. The behavior-based safety process: Managing involve-
ment for an injury-free culture, Van Nostrand Reinhold, New York.
Laring, J., Forsman, M., Kadefors, R., and Ortengren, R. 共2002共. “MTM-
based ergonomic workload analysis.” Int. J. Ind. Ergonom., 30, 135–
148.
Lee, S., and Halpin, D. 共2003共. “Predictive tool for estimating accident
risk.” J. Constr. Eng. Manage., 129共4共, 431–436.
Li, G., and Buckle, P. 共1999共. “Current techniques for assessing physical
exposure to work-related musculoskeletal risks, with emphasis on
posture-based methods.” Ergonomics, 42, 674–695.
Martinez, J. C., and Ioannou, P. G. 共1999共. “General-purpose systems for
effective construction simulation.” J. Constr. Eng. Manage., 125共4共,
265–276.
Meshkati, N., and Loewenthal, A. 共1988共. “An eclectic and critical re-
view of four primary mental workload assessment methods: A guide
for developing a comprehensive model.” Human mental workload, P.
A. Hancock and N. Meshkati, eds., Elsevier, North-Holland, Amster-
dam, The Netherlands, 251–267.
Mitropoulos, P., Cupido, G., and Namboodiri, M. 共2009共. “Cognitive ap-
proach to construction safety: Task demand-capability model.” J.
Constr. Eng. Manage., 135共9共, 881–889.
Namboodiri, M. 共2007共. “The effect of production practices on the like-
lihood of accidents.” Master’s thesis, Del E. Webb School of Con-
struction, Arizona State Univ., Tucson, Ariz.
National Safety Council. 共2009共. “Occupational safety terms and con-
cepts.” 共http://www.nsc.org/resources/issues/safetyknow/safety_terms.
aspx共 共May 2009共.
Oglesby, C. H., Parker, H., and Howel, G. 共1989共. Productivity improve-
ment in construction, McGraw-Hill College, New York.
Reid, G. B., and Nygren, T. E. 共1988共. “The subjective workload assess-
ment technique: A scaling procedure for measuring mental workload.”
Human mental workload, P. A. Hancock and N. Meshkati, eds.,
Elsevier, North-Holland, Amsterdam, The Netherlands, 185–218.
Rwamamara, R. A. 共2007共. “Risk assessment and analysis of workload in
an industrialized construction process.” Construction Information
Quarterly, 9共2共, 80–85.
Saurin, A. T., and Guimarães, L. B. 共2006共. “Ergonomic assessment of
suspended scaffolds.” Int. J. Ind. Ergonom., 36共3共, 229–237.
Sperling, L., Dalhman, S., Kilbom, A., and Kadefors, R. 共1993共. “A cube
model for the classification of work with hand tools and the formula-
tion of functional requirements.” Appl. Ergon, 24, 212–220.
Sun, Y., Fang, D., Wang, S., Dai, M., and Lv, X. 共2008共. “Safety risk
identification and assessment for Beijing Olympic venues construc-
tion.” J. Manage. Eng., 24共1共, 40–47.
Wang, W. C., Liu, J. J., and Chou, C. C..共2006共. “Simulation based safety
evaluation model integrated with network schedule.” Autom. Constr.,
15共3共, 341–354.
Yi, K., and Langford, D. 共2006共. “Scheduling-based risk estimation and
safety planning for construction projects.” J. Constr. Eng. Manage.,
132共6共, 626–635.
38 / JOURNAL OF CONSTRUCTION ENGINEERING AND MANAGEMENT © ASCE / JANUARY 2011
J. Constr. Eng. Manage. 2011.137:30-38.
Downloaded from ascelibrary.org by UNIVERSITE LAVAL on 07/02/14. Copyright ASCE. For personal use only; all rights reserved.