Documente Academic
Documente Profesional
Documente Cultură
Abstract: Airport expansion projects often require the presence of construction personnel, material, and equipment near airport secure
areas/facilities, leading to an increase in the level of risk to airport security. Construction planners and airport operators need to carefully
Downloaded from ascelibrary.org by Universitas Indonesia on 02/09/18. Copyright ASCE. For personal use only; all rights reserved.
study this challenge and implement active measures in order to minimize construction-related security breaches and comply with all
relevant Federal Aviation Administration guidelines. This paper presents the development of an advanced multiobjective optimization
model for planning airport construction site layouts that is capable of minimizing construction-related security breaches while simulta-
neously minimizing site layout costs. The model incorporates newly developed criteria and performance metrics that enable evaluating
and maximizing the construction-related security level in operating airports. The model is developed using a multiobjective genetic
algorithm, and an application example is analyzed to demonstrate the use of the model and its unique capability of generating a wide
spectrum of optimal trade-offs between construction-related airport security and site layout costs.
DOI: 10.1061/共ASCE兲0733-9364共2008兲134:1共40兲
CE Database subject headings: Optimization; Evolutionary computation; Security; Airport construction; Site evaluation; Site
preparation, construction; Construction management; Risk management.
the FAA security guidelines for airport expansion projects; 共2兲 scan. The FAA-recommended systems include the use of: 共1兲 anti-
formulating planning variables and optimization objectives; and intrusion systems such as closed circuit TV 共CCTV兲 and motion
共3兲 implementing the model as a multiobjective genetic algorithm. detectors; 共2兲 detection technologies such as x-ray scanning and
The following sections provide more detailed descriptions of explosives detection; and 共3兲 security lighting systems 共FAA
these three development phases. 1972; 1988兲, as shown in Table 1. These systems play a major
role in preventing unauthorized access to the construction site and
in detecting potential airport security breaches that may originate
FAA Security Guidelines for Airport Expansion from the construction site.
Projects
The FAA establishes guidelines for airport security in a number of Site Layout Planning Variables and Optimization
advisory circulars and reports, including 共1兲 Recommended Secu- Objectives
rity Guidelines for Airport Planning, Design and Construction
共FAA 2001兲 and 共2兲 Aviation Security 共FAA 1972兲. In order to In order to minimize construction-related security breaches, the
control and minimize construction-related security breaches that impact of the aforementioned site layout measures needs to be
may originate during airport expansion projects, construction evaluated and modeled. In this decision-making problem, the
planners and airport operators need to study and comply with all main goal of the planner is to maximize compliance with relevant
relevant FAA security guidelines in these advisory circulars. In FAA guidelines and minimize construction-related security
these documents, the FAA recommends implementing two site breaches at the least possible cost, as shown in Fig. 1. To accom-
layout planning measures to maintain the security of operating plish this, all relevant site layout planning variables are identified
airports during construction: 共1兲 utilizing security response dis- and grouped in three sets: 共1兲 the planned distance to separate
tances; and 共2兲 installing physical security systems 共FAA 2001, each secure area/facility from the construction site fence; 共2兲 the
1972, 1988兲. types of security control systems that are planned to be installed
First, the implementation of security response distances is an on site; and 共3兲 the locations of temporary facilities on site 共Fig.
important requirement in the security planning of operational air- 1兲. The first and second sets of variables directly affect the control
ports to ensure an adequate and secure distance between potential of construction-related security breaches, while the second and
sources of security breaches and the secure area/facility. For ex- third sets directly influence site layout costs. The following sub-
ample, the distance between a construction site fence and a sections discuss these three sets of decision variables and their
nearby secure facility plays an important role in determining the impact on the optimization objectives.
security level of that facility. Locating the fence very close to a
secure facility would not allow for a suitable response to potential
Variable 1: Security Response Distance
breaches originating from the construction site. Accordingly, in-
creasing the separation distance provides greater flexibility to put The distance separating each secure facility from the construction
in place systems, measures, and procedures that will detect, delay, site plays an important and direct role in determining the effect of
and allow for a suitable security response 共FAA 2001兲. the site layout plan on the level of controlling construction-related
Second, the FAA requires installing physical security systems security breaches and an indirect role on the cost of the site layout
to ensure that authorized construction personnel have limited ac- plan. First, sufficient separation distances between the construc-
cess to only the construction site and that they do not have access tion site fence and secure facilities are needed to establish buffer
to airport secure areas. These physical security systems are placed zones around secure facilities in order to provide airport operators
into two groups: FAA-required systems and FAA-recommended with the opportunity of placing systems, measures, or procedures
systems, as illustrated in Table 1. The FAA-required systems in- that defect, delay, and allow for a response to security breaches
clude the use of: 共1兲 physical barriers such as security fences; and 共FAA 2001兲. Second, the location of the construction site fence
共2兲 access control systems such as keypad entry and fingerprint determines the available space for locating temporary facilities on
site. Accordingly, a change in the fence layout will produce in and around airport construction sites. The two new criteria are
changes in the locations of temporary facilities on site, leading to designed to measure and evaluate the impact of site layout deci-
a possible change in the travel distances and cost of construction sions on satisfying the earlier described FAA guidelines, which
resources among the site facilities 共El-Rayes and Khalafallah require establishing security response distances around secure
2005兲. areas/facilities and the utilization of security systems to prevent
construction-related security breaches. As shown in Fig. 2, the
model uses a weighted average method to combine these two
Variable 2: Security Systems Utilized
newly developed criteria in order to evaluate the overall
The Federal Aviation Administration recommends utilizing effi- construction-related security level on site.
cient security control systems in order to control and minimize The first of these criteria is named the security response dis-
security breaches that may originate from the construction site tance criterion 共SRDC兲. The SRDC can be used by construction
共FAA 1972, 1988兲. These systems include security fences, access planners to specify two types of buffer zones around each secure
control technologies, security lighting, and anti-intrusion systems. facility: the required buffer zone, and the recommended buffer
Each of these systems is associated with a certain cost and effec- zone, as shown in Fig. 3. The purpose of the required buffer zone
tiveness, as shown in Table 1. As such, there is a need to optimize is to define the minimum required separation space between any
the selection and utilization of these systems in order to minimize temporary facility and the secure facility. This zone should be free
construction-related security breaches while keeping the site lay- of any potential sources of security risks to the secure facility. The
out costs at a minimum.
role of the recommended buffer zone, on the other hand, is to rity response distance of secure area/facility e; and E = total num-
establish a security response distance 共df e兲, which is needed to ber of secure areas/facilities.
detect, delay, and allow for a response to security breaches that The second criterion is the security systems criterion 共SSC兲
may originate from the construction site 共FAA 2001兲, as shown in and is developed to evaluate the effect of installing security con-
Fig. 3. trol systems on improving security arrangements in airport con-
The FAA suggests a distance of at least 3 – 9 m 共10– 30 ft兲 struction site layouts. As shown in Table 1, each security system
clearance in order to allow for the detection of a fence breach is associated with a qualitative effectiveness rating that can be
共FAA 2001兲, while specifications in the U.K. recommend a sepa- obtained from available security studies 共FAA 1972; GAO 2002;
ration distance of 30 m between the source of risk and the secure EPA 2006兲. These qualitative security ratings can be used as a
facility 共e.g., an airport terminal兲. These recommended distances guide for planners in estimating quantitative effectiveness scores,
should allow for detecting security breaches and responding by as shown in Table 1. The SSC utilizes the weighted sum of the
closing the secure facility before further penetration 共FAA 2001兲. effectiveness scores of all security systems selected for a site
The newly developed SRDC utilizes two variables to measure layout plan as a measure for evaluating the security level pro-
and evaluate the effectiveness of separating secure areas/facilities vided by the entire set of utilized security control systems on site,
from the construction site: 共1兲 the planned distance 共df e兲, which as shown in Eq. 共3兲.
separates each secure area/facility 共e兲 from the construction site Table 2 provides a simple example to illustrate the required
fence; and 共2兲 the recommended security response distance 共de兲 computations for this newly developed criterion. To calculate the
needed for each and every secure area/facility 共e兲 on site. The SSC in this simple example, the weighted effectiveness of each
SRDC uses the ratio between these variables to identify a security selected security control system needs to be evaluated by multi-
rating score for each secure facility on site, as shown in Fig. 3 and plying the effectiveness of each of these systems 共Eu兲 by its rela-
Eq. 共1兲. For example, Case 1 in Fig. 3 represents a layout where tive significance weight 共wu兲. For example, the weighted
the planned fence separation distance 共df e兲 equals half of the effectiveness of Fence Type III is calculated to be wu ⫻ Eu = 0.3
recommended security response distance 共de兲, and accordingly ⫻ 100% = 30%, as shown in Table 2. Second, the results of these
the response distance rating is calculated to be 50% 关see Eq. 共1兲兴. weighted multiplications are summed up to get the SSC rating
Similarly, Case 2 in Fig. 3 represents the most secure case, where 共i.e., SSC= 30+ 18+ 15+ 15= 78%兲, as shown in Eq. 共3兲 and Table
the fence is located completely outside the recommended buffer 2. It should be noted that the relative significance weights 共wu兲
zone, leading to a perfect score in the response distance rating can be readily estimated from historical data about airport secu-
共i.e., RDRe = 100%兲. The SRDC is then used to calculate the over- rity breaches, such as the data of the FAA enforcement database
all security control rating for all secure areas/facilities in the 共FAA 2006兲.
neighborhood of a construction site by averaging the response From these historical data, wu should represent the ratio be-
distance ratings for all secure areas/facilities on site, as shown in tween the number of security breach incidents involving security
Eq. 共2兲 system u and the total number of security breach incidents due to
deficiencies in all security systems. It is also important to note
冦 冧
df e that the FAA-required systems should be assigned relatively
⫻ 100% 共df e 艋 de兲
Response Distance Rating 共RDRe兲 = de higher significance weights 共wu兲 to distinguish them from the
100% 共df e ⬎ de兲 relatively less important recommended systems, as shown in
Table 2. In this example, the collective weight of the FAA re-
共1兲
quired systems is assumed to represent 60% of the SSC rating
E
兺
U
兺
RDRe
e=1 Security Systems Criterion 共SSC兲 = w uE u 共3兲
Security Response Distance Criterion 共SRDC兲 = u=1
E
where wu = relative significance weight of security system u; Eu
共2兲
= effectiveness of security system u, which ranges from 0 to
where df e = planned separation distance between secure area/ 100%; and U = total number of security system categories
facility e and the construction site fence; de = recommended secu- analyzed.
The overall construction-related security level 共CSL兲 can then ties i and j; I = total number of temporary facilities on site; J
be calculated using the weighted average of the SRDC and SSC, = total number of temporary and fixed-location facilities on site;
as shown in Fig. 3 and Eq. 共4兲. In the present model, the CSL is cu = binary variable to represent the utilization of security system
maximized 关see Eq. 共4兲兴 to ensure the selection of the most effec- u to control security on site; and Ccu = cost of installing, operat-
tive security arrangements in and around airport construction ing, and maintaining security system u on site.
sites. It should be noted that the relative significance weights 共ws1 It should be noted that the travel cost rate of resources 共Cij兲
and ws2兲 of the SRDC and SSC can be estimated using historical between facilities can be estimated based on the planned travel
data such as the data of the FAA enforcement database 共FAA frequency of utilized crews, their hourly cost rates, and their av-
2006兲 in a way similar to the estimation of the relative signifi- erage speeds of traveling 共El-Rayes and Khalafallah 2005兲. For
cance weights of various security systems 共wu兲. example, to transport 600 tons of steel, 350 tons of bricks,
Using this database, ws1 can be estimated as the ratio between 250 tons of tile, 200 tons of granite, 150 tons of lumber, and
the number of security breach incidents involving lack of security 50 tons of stone 共i.e., a total of 1,600 tons兲 between a storage
separation distance and the total number of all security breach facility and a construction site, a crew consisting of a forklift and
accidents. On the other hand, ws2 can be estimated as the ratio an operator is selected 共Freeman 2006兲. The capacity of the crew
between the number of security breach accidents due to deficien- is 2 tons per trip, its hourly cost rate is estimated to be $120/ h
cies in security systems and the total number of all security 共Freeman 2006兲, and its average speed of traveling is estimated to
breach accidents. The developed system also gives the user the be 6 km/ h 共CAT 2005兲. Assuming a job efficiency of 45 min/ h,
flexibility to adjust the default values of these relative significance the travel cost rate 共Cij兲 between the storage facility and the con-
weights if deemed necessary. In such cases, the SRDC should be structed facility in this example is estimated to be 共120兲
given a higher relative weight than the SSC as it represents the ⫻ 共1,600⫻ 2 / 2兲 ⫻ 共60/ 45兲 / 共6 ⫻ 1,000兲 = $42.67/ m.
fundamental concept in security planning 共FAA 2001兲
Maximize Construction-Related Security Level
Multiobjective Optimization Model
= Max:关ws1SRDC + ws2SSC兴 共4兲
where ws1 = relative weight/significance of the security response The present optimization model is implemented as a multiobjec-
distance criterion; and ws2 = relative weight/significance of the se- tive genetic algorithm in order to enable the generation of optimal
curity systems criterion. site layout plans that maximize construction-related airport secu-
rity while keeping the overall site layout cost at a minimum. The
model is implemented using the nondominated sorting genetic
Objective 2: Minimizing Overall Site Layout Cost algorithm-II 共Deb et al. 2000兲. The earlier-described three sets of
The overall site layout cost in this model is grouped into two site layout planning variables are represented in this model by a
main categories: 共1兲 the travel cost of resources among temporary genetic algorithm chromosome 共string of variables兲, as shown in
facilities, which is affected by the travel cost rate of resources Fig. 1. The optimization calculations in the present model 共Fig. 1兲
共Cij兲, and the planned locations of these facilities as well as the are performed in four main steps: 共1兲 generate an initial set of
location of the selected security fence; and 共2兲 the cost of install- random site layout solutions; 共2兲 evaluate and rank the generated
ing security control systems on site. As shown in Eq. 共5兲, this solutions based on nondomination criteria 共Deb et al. 2000兲; 共3兲
overall cost is minimized as the second main objective function in select the best solutions to perform crossover and mutation opera-
the present multiobjective optimization model, which is described tions and generate the next generation of solutions; and 共4兲 repeat
in more detail in the following section the second and third steps over a number of cycles until the gen-
erated solutions converge to an optimal/near optimal set of site
I−1 J
layout solutions.
Minimize Overall Site Layout Cost = Min: 兺 兺 Cijdij
i=1 j=i+1
U
Application Example
+ 兺
u=1
cuCcu 共5兲
An application example is analyzed to illustrate the use of the
where Cij = travel cost rate of resources in $/m of distance traveled present model and demonstrate its capabilities in optimizing con-
between facilities i and j; dij = distance in meters between facili- struction site layouts and generating optimal trade-offs between
ence Foundation under NSF CAREER Awards Nos. CMS nondominated sorting genetic algorithm for multiobjective optimiza-
0238470 and CMS 0626066. Any opinions, findings, conclusions, tion: NSGA-II.” KanGAL Rep. 200001, Indian Institute of Technol-
ogy, Kanpur, India.
or recommendations expressed in this publication are those of the
Elbeltagi, E., Hegazy, T., and Eldosouki, A. 共2004兲. “Dynamic layout of
writers and do not necessarily reflect the views of the National construction temporary facilities considering safety.” J. Constr. Eng.
Science Foundation. Manage., 130共4兲, 534–541.
El-Rayes, K., and Khalafallah, A. 共2005兲. “Trade-off between safety and
cost in planning construction site layouts.” J. Constr. Eng. Manage.,
Notation 131共11兲, 1186–1195.
Environmental Protection Agency 共EPA兲. 共2006兲. “Fences.” 具http://
cfpub.epa.gov/safewater/watersecurity/guide/productguide.cfm?page
The following symbols are used in this paper:
⫽fences典 共April 11, 2006兲.
Ccu ⫽ cost of installing, operating, and maintaining Federal Aviation Administration 共FAA兲. 共1972兲. “Aviation security—
security system u on site; Airports.” Advisory circular no. 107-1, U.S. Dept. of transportation,
Cij ⫽ travel cost rate of resources in $/m of distance Washington, D.C.
traveled between facilities i and j; Federal Aviation Administration 共FAA兲. 共1988兲. “Planning and design
cu ⫽ binary variable to represent the utilization of guidelines for airport terminal facilities.” Advisory circular no. 150/
security system u to control security on site; 5360-13 CHG:1, U.S. Dept. of Transportation, Washington, D.C.
de ⫽ recommended security response distance of secure Federal Aviation Administration 共FAA兲. 共2001兲. “Recommended security
area/facility e; guidelines for airport planning, design and construction.” Rep. No.
DOT/FAA/AR-00/52, R. Lazarick and Cammaroto, U.S. Dept. of
df e ⫽ planned separation distance between security area/
Transportation, Washington, D.C.
facility e and the construction site fence;
Federal Aviation Administration 共FAA兲. 共2003兲. Aviation capacity en-
dij ⫽ distance in meters between facilities i and j;
hancement plan: Building capacity today for the skies of tomorrow,
E ⫽ total number of secure areas/facilities;
FAA Office of System Capacity, Washington, D.C.
Eu ⫽ effectiveness of security system u 共ranges from 0 to Federal Aviation Administration 共FAA兲. 共2006兲. “FAA enforcement data-
100%兲; base records.” 具http://www.faa.gov/foia/faa_enforcement database
I ⫽ total number of temporary facilities on site; records/典 共December 23, 2006兲.
J ⫽ total number of temporary and fixed-location Freeman. 共2006兲. “Exhibitor service manual.” 具http://www.ametsoc.org/
facilities on site; exhibitr/2004exhibitinfo/manuals/exhibitor service manual.pdf典 共De-
U ⫽ total number of security system categories analyzed; cember 23, 2006兲.
ws1 ⫽ relative weight/significance of security response General Accounting Office 共GAO兲. 共2002兲. “National preparedness: Tech-
distance criterion; nologies to secure federal buildings.” Testimony before the Subcom-
ws2 ⫽ relative weight/significance of security systems mittee on Technology and Procurement Policy, Committee on Gov-
criterion; and ernment Reform, U.S. House of Representatives. Washington, D.C.
wu ⫽ relative significance weight of security system u. Hamiani, A. 共1989兲. “Knowledge representation for the site layout prob-
lem.” 6th Conf. on Computing in Civil Engineering, New York, 283–
Subscripts and Superscripts 289.
Hegazy, T., and Elbeltagi, E. 共1999兲. “EvoSite: Evolution-based model
i ⫽ temporary facility counter 共from i = 1 to I兲; and
for site layout planning.” J. Comput. Civ. Eng., 13共3兲, 198–206.
u ⫽ security system category 共from u = 1 to U兲.
Kumara, S. R. T., Kashyap, R. L., and Moodie, C. L. 共1988兲. “Applica-
tion of expert systems and pattern recognition methodologies to fa-
cilities layout planing.” Int. J. Prod. Res., 26共5兲, 905–930.
References Li, H., and Love, P. E. D. 共1998兲. “Site-level facilities layout using ge-
netic algorithms.” J. Comput. Civ. Eng., 12共4兲, 227–231.
123 CCTV Security Camera Surveillance Equipment. 共2006兲. “Complete Mawdesley, M. J., Al-Jibouri, S. H., and Yang, H. 共2002兲. “Genetic al-
4 camera TV system.” 具http://www.123cctv.com/cctv/cctv-systems- gorithms for construction site layout in project planning.” J. Constr.
4.html典 共April 19, 2006兲. Eng. Manage., 128共5兲, 418–426.
AirNav. 共2005兲. “Phoenix Sky Harbor International Airport.” 具http:// Mazzara, A. F., Swanson, D. C., and Nicholas, N. C. 共2003兲. “Sensor
www.airnav.com/airport/KPHX典 共February 13, 2005兲. fence: A new approach to large-perimeter security.” Corrections
Armour, G. C., and Buffa, E. S. 共1963兲. “A heuristic algorithm and com- Today, 65共1兲, 94–95.
puter simulation approach to relative location of facilities.” Manage. Osman, H. M., Georgy, M. E., and Ibrahim, M. 共2003兲. “A hybrid CAD-
Sci., 9共1兲, 294–309. based construction site layout planning system using genetic algo-
Berg, R., and Hinze, J. 共2005兲. “Theft and vandalism on construction rithms.” Autom. Constr., 12共6兲, 749–764.