Complementary Cooperation of CCTV and UAV Systems for Tourism Security and Sustainability

Abstract

The meaning of sustainability is very broad and has many pillars such as the economy, environment and society. In the tourism industry, another important pillar is security. Tourism security affects the satisfaction and pleasure of tourists on a trivial level, and life and injuries on a significant level. Thus, unless security is guaranteed, tourists will not be able to fully enjoy the attractions and will not even consider the travel itself. Such tourist behavior has a significant impact on short-term and long-term tourism sustainability. Therefore, to enhance tourism security, many policies and frameworks have been suggested and announced in recent years. However, without efficient guidelines for the installation and operation of security devices, it may be hard to obtain actual effectiveness. To support real implementation of security systems in the tourism industry, this study quantitatively addresses the installation and operation issue of security devices in an optimal manner. A complementary cooperation of Closed-Circuit Television (CCTV) and Unmanned Aerial Vehicles (UAV) is suggested to efficiently monitor the key locations of tourism destinations and improve the security. Two mathematical models are developed to derive the optimal location of CCTVs, and the optimal operation schedule of UAVs over multiple time periods. Security requirements, service range, and budget are considered as realistic constraints. The validity of the models is demonstrated through a realistic case study of Nice, France.

1. Introduction

The 2005 world submit on social development clarified the three pillars of sustainability as (1) economic development, (2) environmental protection, and (3) social development [1]. The three pillars have served as a common ground for numerous sustainability standards and certification systems in recent years for a variety of industries. However, in the tourism industry, there is another important pillar, which is tourism security. Misdemeanors in a tourism destination affect the satisfaction, pleasure and economic loss of tourists. Moreover, terrible accidents such as terrorism are directly linked to the lives and injuries of tourists. Thus, unless security is guaranteed, tourists will not be able to fully enjoy tourism destination and will not even consider travel itself. Such behavior of tourists can give huge negative effects on the short-term and long-term tourism sustainability such as tourist visits, survival of the local residents, and tourism revenue. The importance of security to tourism sustainability has already been emphasized by researchers. Reference [2] pointed out the importance of tourism security for sustainable tourism. Long-term investment for tourism security was suggested as one of the sustainable tourism strategies. Reference [3] argued that the safety and security issues at tourist sites have a great impact on the tourism industry, not only the decrease of tourists but also the survival and tension of local residents. Reference [4] pointed out the importance of safe and security for sustainability factors and performance of a tourism destination. Consequently, without tourism security, the tourism industry may not pursue sustainable development.

In recent years, concerns about tourism security have increased dramatically due to a series of terrorist attack in Europe, United States, East Asia and Australia. Even tourism destinations and tourists have always been ‘soft targets’ for terrorist activities [5]. One of the most heartbreaking terrorist attacks was in Nice, France in 2016. On the evening of 14 July 2016, a 19-tonne cargo truck was deliberately driven into crowds of people celebrating Bastille Day on the Promenade des Anglais in Nice, France, resulting in the deaths of 86 people and the injury of 458 others. It was a tragedy not only for the victims and their families, but also for the tourism industry. Due to terrorism around the world, travel to tourists became a source of fear; and the tourist’s fear brought adverse effects on the travel industry such as the cancellation of the reservation and avoidance of tourism destinations. Furthermore, such adverse effects are not limited to immediate and short-term periods. The World Travel and Tourism Council (WTTC) estimates that rebuilding the tourism sectors of countries hit by terrorism takes on average 13 months, from two to 22 months [6]. In other words, tourism security has a significant impact on short-term and long-term tourism sustainability, and the failure of security is directly linked to the survival of tourism destination and local residents. Thus, preemptive precautions are absolutely required to prevent and respond to the terrible threats for short-term and long-term tourism sustainability.

Recently, government agencies and researchers have proposed various policies and frameworks to enhance the security of tourist destinations. For example, the French government announced a plan to roll out more CCTVs in sensitive areas where tourists have been targeted, notably hotels on the edge of Paris [7]. However, such policies and frameworks require a specific method to efficiently install and operate security devices. In other words, quantitative guidelines for the specific installation and operation of security devices should be provided with consideration of the characteristics of the tourism destinations. Policies and frameworks without specific installation and operation guidelines are hard to obtain actual effectiveness. Herein lies the goal of this study. This study quantitatively addresses the installation and operation issues of security devices to prevent and respond to the physical threats in tourism destination. In this study, the complementary use of traditional and emerging security devices is addressed. To efficiently mitigate the threats to tourism destinations, the installation and operation issues of CCTV and UAV are quantitatively addressed. Such devices will cooperate to monitor the key locations in tourism destinations that change over time depending on tourist flows. Mathematical optimization models are developed and tested to derive specific guidelines on the system installation and operation. The result of this study is expected to provide actual contributions to enhance tourism security by suggesting specific methods for the implementation of security devices in tourism industry. The proposed approach can create synergy with previous studies on the policies and frameworks for tourism security by quantitatively supporting the implementation of actual security activities in tourism industry.

2. Related Studies

The security failure of tourism destinations directly causes many kinds of risks for the tourism destinations. Unfortunately, perceived travel risks directly affect the tourism industry. A variety of studies were conducted to identify and analyze the effect of security failure on the tourism industry based on the survey and impact analysis. Reference [8] investigated key factors of international tourism decision making processes. Via mail survey and telephone interview, risk perception levels were found to directly influence international destination choice. Reference [9] conducted a survey against 1180 international travelers visiting Hong Kong in the fall of 2003 to investigate the impact of the perception of risk on international travelers. Findings of the survey suggested that international travelers appear to be sensitive towards the occurrence of any type of risk in their evoked destinations, but travelers may have varying degrees of the perceived risk based on the national cultures. The perceived risk of a particular tourist destination can have a positive effect on the competitive tourist destinations. Reference [10] focused on the short-run impacts of the September 11 attacks in New York on tourist preferences for competing destinations in the Mediterranean and the Canary Islands. A stated preference model was developed using survey results with two different samples taken at different points in time (pre-and post-September 2001). Results of the model pointed out that competing destinations were upgraded on their image and attractiveness because of terror events. Reference [11] investigated perceived risks of diseases such as SARS and bird flu as well as terrorism on Thailand’s tourism industry. Interviews and surveys were conducted, and the results showed that people did not discontinue travelling completely but did select different options from alternatives. Reference [12] quantitatively analyzed major terrorism events around the world during 1985–1998. Tragic terrorist attacks in Paris, Tel Aviv, Egypt and the Kenya and Tanzania were investigated and their effect on tourism demand was analyzed. Reference [13] compared the tourism trend of the United States and Hawaii to ascertain how the terrorist attacks of 9/11 and subsequent terror events affected tourism flows. It was found that the United States tourism has not fully recovered from 9/11 and other international attacks while Hawaiian tourism is enjoying robust growth in the aftermath of 9/11. Reference [14] analyzed the impact of terrorism on the Turkish tourism industry between 1986 and 2006. Relationship between the number of people killed and injured due to terrorist incidents and total international arrivals was analyzed by the development of autoregressive distributed lag bounds test. Results of the test indicated the existence of a negative causal effect of terrorism on the Turkish tourism industry. In addition, Reference [15] analyzed safety and security as one of the seven key factors affecting the satisfaction of tourists.

As we have seen, the negative impact of security failures and perceived travel risks of tourism destinations is obvious, not tentative. Many governments and authorities are devising security measures, policies and frameworks to secure the security of the tourism destination. For example, threat of terrorism continues occurring in Paris and Nice. Terrorism threatens not only the French people but also the lives of tourists, which has greatly affected the French tourism industry. According to the study in [8], France will spend 42 million euros to boost security around tourism in the hope it will encourage visitors to return. Among the security measures include a plan to roll out more video surveillance or CCTV in sensitive areas where tourists have been targeted, notably hotels on the edge of Paris. Singapore’s civil aviation authority announced its support for the commercial use of drones (UAVs) in delivering parcels, inspecting buildings and providing security. Over the past six months, the government has been experimenting with security drones [16]. The Singapore Civil Defense Force (SCDF) deployed them to protect the hundreds of ministers and delegates attending the ASEAN Summit, which began on 11 November. India and Korea also use UAVs for the security enhancement of national park [17]. In additions, researchers theoretically support such efforts. Reference [18] suggested a generic model for analyzing and developing tourism disaster management strategies. Reference [19] evaluated the Australian federal government’s response to a series of tourism disasters and crises that affect the Australian tourism industry using the model of [19]. Reference [20] reviewed previous models that have been developed to help strategic planning for crisis management of international tourism industry. The terrorist attacks of 9/11 are used as an exemplar of external shock, and a crisis management process model was compared and contrasted with the steps actually undertaken at the company. Reference [5] suggested a strategic framework for terrorism prevention and mitigation in tourism destinations. Based on the analysis of interviews with 16 experts on terrorism and tourism, the authors offer a framework for the development and implementation of a destination-specific anti-terrorism strategy.

However, such policies and frameworks require a specific method to efficiently install and operate security systems and devices. In other words, quantitative guideline on the specific installation and operation of security devices should be provided with the considerations on the characteristics of tourism destinations. Policies and frameworks without specific installation and operation guidelines are hard to obtain actual effectiveness. The installation and operation issues of CCTV and UAV systems were widely investigated in a variety of industries. However, the applications for tourism industry were rare. Reference [21] addressed the issue of UAV depot installation and task assignment for safety and security of tourism attraction. A mathematical model was developed and tested with the example of Children’s Grand Park in Korea. Reference [22] suggested a management scientific approach to select proper UAVs and derive UAV schedules for security safety missions in tourism attractions. Reference [23] suggested the use of UAV for wireless service of crowded tourism areas. Optimal number of UAV base stations and their altitude were derived to handle network congestion issues in peaking seasons. On the other hands, some studies investigate the use of RFID system for tourism security. Reference [24] suggested an ant colony based optimization approach for RFID reader deployment in theme park. More RFID applications in hospitality industry can be found in [25].

The means of crime are evolving day by day and various emerging technologies are utilized to combat this. Therefore, the security system for preventing and responding to these problems should also use emerging technologies efficiently. In the tourism industry, security has a significant impact on long-term sustainability. To cope with the threats, many policies and strategies are suggested. However, the quantitative approach to deal with the system installation and operation has been rare addressed. The contribution of this study lies here. This study firstly addresses the complementary use of traditional (CCTV) and emerging (UAV) security devices for tourism destination. Mathematical models are developed to optimize the installation and operation of the systems to quantitatively improve the tourism security and sustainability. The results of this study can provide answers on how to implement and operate security system to embody macroscopic level of policies and strategies for tourism security and sustainability.

3. CCTV and UAV Security System in Tourism Destination

3.1. Characteristics of CCTV and UAV Systems

3.1.1. CCTV Security Systems

CCTV security systems are traditional but powerful equipment for security. It is easy to install with low setup cost. The existence of CCTV systems itself can prevent and control crime, and it is possible to arrange an effective criminal arrest through image analysis in case of crime. In addition, when using CCTV system, it can monitor key areas instead of security manpower, so that security manpower can be utilized efficiently. However, CCTV system can help prevent crime and facilitate effective arrest through analysis of past images, but it is impossible to respond promptly to the current crime. As CCTV itself is installed in a fixed location, it cannot track moving objects such as criminals or criminal vehicles.

3.1.2. UAV Security System

The installation of UAV security systems are more costly than CCTV security system. Furthermore, commercial UAV has limited flight time due to the small size of battery. However, the operation cost of UAV system is very low [26]. Moreover, UAV’s operating range is not as stationary as CCTV. UAV provides flexible mobility, enabling immediate response and continuous tracking. According to [27], UAV (or drone) security system gives following benefits for surveillance:

• Mixing tethered drones and deployable drones help security respond to all situations quickly and efficiently.
• Intruder Responses.
• Using drones help security forces get closer to intruders without alerting them.
• UAV’s can run a perimeter route faster and more easily.
• Drones can be used to search blind spots and corners, minimizing risk to people.
• Tracking the intruders just became easier.
• With drones, response time is much faster.
• Drone can access to the inaccessible area and provide surveillance task

That is, CCTV and UAV systems have their advantages and disadvantages come from location-based characteristic and mobility. Thus, there is a huge possibility for the complementary cooperation of CCTV and UAV systems for the tourism security. In the complementary cooperation, two systems will be efficiently used to maximize their benefits while minimizing their weaknesses. This complementary cooperation can help secure the long-term sustainability of the tourist destination by enhancing the security of the destination and helping tourists to travel comfortably.

3.2. Tourist Flow of Tourists

The flow of tourists’ movements at tourism destinations varies depending on the characteristics of the destination and time. For example, in beach destination, tourists will be crowded in the downtown area during mealtimes. However, in the afternoon, more tourists will visit the beach area, and the beach area will be less visited by tourists after dinner. In other words, the security priority of key locations will vary according to the characteristics and the time of the tourism destination. Therefore, the security system of tourism destination should be installed and operated to efficiently monitor key locations by considering their time-specific priorities using limited resources. Figure 1 shows the example of relative security priorities (maximum of 10) of downtown area and beach area in Nice, France over 24 h.

3.3. Installation and Operation Issues

Now we will focus our attention on issues related to the installation and operation of CCTV and UAV security system in tourism destination. With the considerations on the characteristics of CCTV, UAV and tourist flow, quantitative methodology is required to efficiently install and operate system component. We can install and operate the system without any quantitative approaches. Without the approach, however, we will not be able to efficiently use our resources, leading to failure of the security system. Therefore, to efficiently utilize limited resources and achieve optimal security system installation and operation, a mathematical optimization approach has been developed. The mathematical optimization model consists of an objective function and a set of constraints. A set of constraints will compose a feasible region. Optimal solutions will stay inside the feasible region while trying to maximize the value of objective function.

In the proposed system, there are |I| key security locations in tourism destinations. The security priority of the key location changes over time throughout the day are as discussed previously. Near the key locations, there are |J| and |S |number of candidate locations for CCTV and UAV station installations. The role of UAV stations is to replenish consumable UAVs such as UAV battery. Each candidate location for CCTV and UAV station has their service range. Therefore, only key locations within the service range of each candidate can be monitored. A key security location may need to be monitored from one or more security devices. Additionally, certain key security locations cannot be monitored by CCTV due to locational issues, and only UAV monitoring can be available. In this situation, the objective of the mathematical optimization model is to maximize the total sum of priority of the monitored key locations over a day (24 h) by installing and operating CCTV and UAV stations efficiently within a limited budget. Two mathematical models are developed for (1) a CCTV-only system, and (2) a CCTV and UAV system. The following section will provide developed notations and mathematical models.

4. Mathematical Models

A UAV security system has tremendous merits, however, it belongs to a new technology and there are many difficulties in installing and utilizing the system. So far, most security systems are based on the CCTV system, and some of the state-of-the-art systems are operating UAV based security systems. Therefore, in this study, two mathematical optimization models were developed for (1) a CCTV-only system and (2) a CCTV and UAV system to quantitatively install and operate both current and future security systems (please refer to [28] and [29] for the diverse use of mathematical optimization approach for CCTV and UAV systems).

4.1. Mathematical Model 1: CCTV Security System

4.1.1. Notations of Model 1

Table 1. Model 1 notations.

System Variables
I	Set of key locations for monitoring
J	Set of candidate CCTV locations
T	Set of time periods
pit	Security priority of key location i at time t
dij	Distance between location i and j
ccctv	Purchase cost of CCTV
rcctv	Service range of a CCTV
kit	Service requirement of key location i and time t. It represents the number of required security equipment to monitor key location i at time t
b	Available budget (USD) for system installation
Decision Variables
Xijt	Binary decision variable, it is equal to 1 if key location i is monitored by CCTV at j in time t; otherwise, 0.
Yj	Binary decision variable, it is equal to 1 if CCTV is located at candidate location j; otherwise, 0.
Wit	Binary decision variable, it is equal to 1 if key location i is fully monitored by CCTV with the number of required security equipment at time t; otherwise, 0.

shows the system and decision variables for mathematical model 1.

4.1.2. Mathematical Formulations of Model 1

Equation (1) represents the objective function of mathematical model 1. The goal of mathematical model 1 is to maximize the total sum of monitored security priority of key location i at time t. By the development of Equation (1), CCTV security system can be installed and operated in a way to efficiently observe key locations in time and maximize the benefits of the CCTV security system. Please note that only fully covered key locations are perceived as being perfectly monitored by the security system. Partial monitoring can also improve security, but it includes the risk of accidents due to the imperfections. In order to eliminate this incompleteness and increase the case of perfect monitoring, only fully covered key locations were considered. Thus, Equation (1) maximizes the total priority sum of the perfectly monitored key locations during T.

$$Maximize{\displaystyle \sum _{i\in I}{\displaystyle \sum _{t\in T}{p}_i^t\cdot W_i{}^t}}$$

(1)

Equations (2)–(6) show the constraints of the mathematical model 1. Equations (2) and (3) link two decision variables of the mathematical model 1 and represent service availability. Key location i can be observed if the CCTV is installed in the CCTV candidate location j and is within the service range. Equation (4) enforce mathematical model 1 to satisfy the service requirement of key location i and time t. The number of installed CCTVs is limited by the available budget by Equation (5). Lastly, Equation (6) describes the variables of mathematical model 1.

$${\displaystyle \sum _{i\in I}{X}_{ij}{}^t}\le M\cdot Y_j,\forall j\in J,\forall t\in T$$

(2)

$$d_{ij}\cdot X_{ij}{}^t\le r_{cctv}\cdot Y_j,\forall i\in I,\forall j\in J,\forall t\in T$$

(3)

$${\displaystyle \sum _{j\in J}{X}_{ij}{}^t}\ge k_i{}^t\cdot {\mathrm{W}}_i^t,\forall i\in I,\forall t\in T$$

(4)

$$c_{cctv}\cdot {\displaystyle \sum _{j\in J}{Y}_j}\le b$$

(5)

$$X_{ij}{}^t,Y_j,W_i^t\in \{0,1\},\forall i\in I,\forall j\in J,\forall t\in T$$

(6)

4.2. Mathematical Model 2: CCTV and UAV Security System

Mathematical model 2 addresses the simultaneous use of CCTV and UAV security systems. Therefore, it requires new and additional notations and mathematical formulations. The previously discussed UAV security system requires UAV stations and affiliated UAVs for persistent replenishment of its consumables. Additionally, due to the limited flight duration of UAVs, a single UAV may not be able to monitor the entire time period t. In this case, two or more UAVs can alternately monitor the key location. At last, some key locations, such as sea and port areas cannot be monitored by CCTV due to locational issues and only UAVs can be monitored. In this case, only UAVs can monitor the point and proposed mathematical model 2 support such decisions. New and additional notations and mathematical formulations take into account the characteristics of the UAV security system and derive the installation and operation guidelines of the security system using CCTV and UAV.

4.2.1. New Notations of Model 2

Table 2. Model 2 notations.

System Variables
S	Set of candidate UAV stations
dis	Distance between location i and s
uit	Mandatory UAV service request of key location i (usually, area where installation of CCTV is limited)at time t. It is equal to 1 if key location i should be monitored by UAV at time t; otherwise, 0.
cstation	Installation cost of UAV station
cuav	Purchase cost of UAV
nuav	Number of UAVs that requires to monitor a key location using UAVs
ruav	Service range of a UAV station
Decision Variables
Zs	Binary decision variable, it is equal to 1 if UAV station is installed at s; otherwise, 0.
Uist	Binary decision variable, it is equal to 1 if key location i is served by UAV station s at time t; otherwise, 0.

represents the additional system and decision variables for mathematical model 2.

4.2.2. Mathematical Formulations of Model 2

Objective function of mathematical model 2 is same with that of model 1. The goal of mathematical model 2 is to maximize total sum of monitored security priority of key location i at time t by the both CCTVs and UAVs like mathematical model 1.

$$Maximize{\displaystyle \sum _{i\in I}{\displaystyle \sum _{t\in T}{p}_i^t\cdot W_i{}^t}}$$

Equations (2), (3), (7) and (8) depict service availability of CCTV and UAV security system. Equations (7) and (8) work in the same way as Equations (2) and (3) to guarantee the service availability of the UAV security system.

$${\displaystyle \sum _{i\in I}{X}_{ij}{}^t}\le M\cdot Y_j,\forall j\in J,\forall t\in T$$

$$d_{ij}\cdot X_{ij}{}^t\le r_{cctv}\cdot Y_j,\forall i\in I,\forall j\in J,\forall t\in T$$

$${\displaystyle \sum _{i\in I}{U}_{is}{}^t}\le M\cdot Z_s,\forall s\in S,\forall t\in T$$

(7)

$$d_{is}\cdot U_{is}{}^t\le r_{uav}\cdot Z_s,\forall i\in I,\forall s\in S,\forall t\in T$$

(8)

Equations (9) and (10) simultaneously work to satisfy the mandatory UAV service requests. Equation (9) guarantees that if a key location i requests mandatory UAV security service at time t, then UAVs should serve the task. The task assignment to CCTVs becomes impossible due to the development of Equation (10). Equation (11) enforces mathematical model 2 to satisfy the service requirement of key location i and time t by both CCTV and UAV systems. Equation (12) shows the budget limitation of the system. The purchase cost of CCTVs, UAV stations and UAVs are considered together. At last, five decision variables of mathematical model 2 are shown in Equation (13). Please note that in this study, one UAV station can monitor only one location per each time period.

$${\displaystyle \sum _{s\in S}{U}_{is}^t}\ge u_i{}^t,\forall i\in I,\forall t\in T$$

(9)

$$X_{ij}^t\le M(1-u_i{}^t),\forall i\in I,\forall j\in J,\forall t\in T$$

(10)

$${\displaystyle \sum _{j\in J}{X}_{ij}{}^t}+{\displaystyle \sum _{s\in S}{U}_{is}{}^t}\ge k_i{}^t\cdot W_i^t,\forall i\in I,\forall t\in T$$

(11)

$$c_{cctv}\cdot {\displaystyle \sum _{j\in J}{Y}_j}+c_{station}\cdot {\displaystyle \sum _{s\in S}{Z}_s}+c_{uav}\cdot n_{uav}\cdot {\displaystyle \sum _{s\in S}\underset{t\in T}{\max }({\displaystyle \sum _{i\in I}{U}_{is}^t})}\le b$$

(12)

$$X_{ij}{}^t,Y_j,U_{is}{}^t,Z_s,W_i^t\in \{0,1\},\forall i\in I,\forall j\in J,\forall s\in S,\forall t\in T$$

(13)

4.3. Solution Approach

Proposed mathematical optimization models belong to the mixed integer linear programming (MILP) and we can obtain optimal solution of MILP using commercial optimization software, CPLEX [30]. Therefore, in this study, CPLEX is adopted as solution approach to obtain optimal guidelines for the installation and operation of CCTV and UAV security system.

5. Case Study

5.1. Case Study Details

The validity of the proposed mathematical models was tested via a realistic case study. As mentioned above, this study proposed two different models: (1) CCTV-only, and (2) CCTV and UAV security system. CCTV is able to monitor ground areas, while UAV is capable of monitoring ground and coastal areas. Therefore, two situations are simulated to test mathematical model 1 and 2. In situation 1, the installation of CCTVs is investigated using mathematical model 1. On the other hand, situation 2 considers the simultaneous use of CCTV and UAV for monitoring.

Nice, France was selected as the case study spot. The location of the key monitoring points has been set up as areas with a large floating population, such as tourist attractions and intersections. Nice is also a beach city, and many tourists enjoy the sea. Therefore, the coastal area was selected as having key points. In Figure 2, yellow and green boxes stand for such key locations. Yellow boxes are usually located at ground points while green boxes are located at coastal area and high land. This distinguishes the usability of CCTV and UAV. UAV has wider monitoring range that CCTV. Thus, it is able to monitor yellow and green key locations. However, CCTV is able to monitor only yellow key locations due to the limited range and installation limits. Consequently, monitoring yellow key points is available via both CCTV and UAV. However, green key locations can be monitored using only UAV. At the edge of yellow boxes, there are yellow spots, which indicate the candidate locations for CCTV installation. Due to the limited monitoring range or CCTV, CCTV should be installed around key locations. On the other hand, a UAV system requires the UAV station for replenishment of UAV consumables such as battery, camera, etc. Green spots in Figure 2 indicated the candidate locations for UAV stations.

As previously mentioned, installation of CCTVs is investigated in situation 1 while use of both CCTV and UAV is considered in situation 2. Consequently, in situation 1, the monitoring of only yellow key locations is investigated via mathematical model 1. In addition, in situation 2, monitoring of both yellow and green key locations is analyzed using mathematical model 2. Furthermore, simulation (situation) 1 and 2 were conducted using the personal computer of Intel(R) Core (TM) i7-10510U CPU 1.80 GHz and 16.00 GB of RAM, and CPLEX 12.9.0 is applied to derive solutions.

5.1.1. Cases Study Parameters for Situation 1

As this research derives the location of CCTVs that can increase tourism security, the range that a CCTV can monitor is set as 1 m. It means that CCTV can almost only monitor the area that is located nearby them. The security priority of each key location is decided by considering the area importance, floating population, etc. The key locations that should be mandatorily monitored are set as 100 to be recognized as high priority area. The locations with the medium priority are set between 5, 10. Lastly, the areas with low security priority are decided between 1 and 2 considering the time. For medium and low priority areas, the difference in the priority value by time depends on the density of tourists. Such a relative priority value will provide a more efficient and effective security system by monitoring higher value areas preferentially.

The purchase cost of each CCTV is decided as USD 365. The price of each CCTV varies from a few dollars to more than hundred dollars depending on the manufacturing company. Though cheap CCTVs can be applied indoors, their usage may be extremely limited outdoors. Therefore, the price of each CCTV is decided as USD 365, which is the price revealed at online market for the high outdoor usage. Lastly, budget is decided based on the price of CCTV. In this case study, there are 62 CCTV candidate slots, and if all the CCTVs are installed, USD 22,630 is needed. However, the purpose of this research is to derive optimal decision with the limited budget. Therefore, the budget is set that allows about 35%, 50%, and 65% of candidate slots to be fulfilled. The budget to satisfy each percentage is USD 8000, 11,500, and 15,000.

5.1.2. Cases Study Parameters for Situation 2

Situation 2 considers the installation and use of CCTV and UAV systems to monitor both ground and coastal areas. The parameters related to the CCTV are same as the parameters at Section 5.1.1. To apply the UAV for monitoring, the installation of UAV stations is essential for battery replenishment. Therefore, the cost to install UAV station and each UAV cost are set to USD 3000 and 1199. First, the price of each UAV is decided by comparing the price uploaded at the DJI Store. As the price of each UAV gets higher, flight time, flight speed, service time, service range, and camera resolution get better. This study applied UAV systems to monitor coastal and high-altitude areas. From the DJI store, the UAV with at least USD 1199 of purchase cost will guarantee reasonable service range and flight time. For persistent and continuous UAV monitoring system, UAVs are required to shift their mission appropriately. UAV station will support this shift. Usually, the flight duration of commercial UAV is around 30 min. Thus, to provide persistent monitoring service, at least two or three UAVs are required for swift shifts. A station will maintain such multiple UAVs and efficiently support the mission shift. Based on that idea, the price installation of a UAV station is decided as USD 3000, which corresponds to the price for about two to three UAVs.

The security priority of each key location is decided among values of 1, 2, 5, 10, 20, 40, and 100. The security priority values of 1, 2, 5, 10, 100 are set the same as the situation 1. However, the coastal areas are newly considered in situation 2. Therefore, by considering coastal areas as more important than medium priority but not mandatory, their priorities are decided as 20 or 40 depending on the time period. The range that a UAV can monitor is set as 600 m. Compared to the CCTV, a UAV can monitor the key points while moving around the area. Moreover, the actual distance between locations is calculated based on the latitude and longitude of each location.

5.2. Results of the Case Study

5.2.1. Results of the Situation 1

Using the parameters settings in Section 5.1.1, case study with situation 1 was conducted. The result of situation 1 with the USD 8000 budget was depicted Figure 3. Most key points are monitored by two CCTVs. With the budget of USD 8000, all the mandatory and medium priority key points are monitored. Therefore, with the current condition, USD 8000 is the minimum budget to guarantee the least security. In case of the low priority key points, key point 19 is monitored by one CCTV during the time between 7 o’clock to 18 o’clock. Therefore, it is derived that a total 21 CCTVs are needed to keep the least security. However, keeping the higher security can guarantee the safety of tourists. As the safety level gets higher, more tourists might prefer visiting the place, and it will lead to the true sustainability of tourism. Therefore, additional analysis with larger budgets is conducted. The run time of simulation 1 with USD 8000 budget was 08.13 s.

Figure 4 graphically shows the result of simulation (situation) 1 with USD 11,500 and 15,000 budgets. Obviously, the number of CCTVs increases when the budget becomes USD 11,500. Compared to the budget of USD 8000, key point 19 is fully monitored with two CCTVs. Most key points that are newly monitored by installing 2 CCTVs. In case of key point 5, they need 2 CCTVs to be monitored during 7 o’clock to 18 o’clock. Therefore, only one CCTV is installed.

When the budget increases to USD 15,000, most key points are monitored by the two CCTVs. Compared to the USD 11,500, some monitored key points changed. For example, the key point 25 is monitored with the budget of USD 11,500, but they are not monitored when budget is USD 15,000. However, the monitored key points can change if the reward is the same. Therefore, though the monitored key points change, the total reward can be the same. Such results are obtained within 07.44 and 04.24 s for USD 11,500 and 15,000 budget cases.

5.2.2. Results of the Situation 2

Based on the parameters set at 5.1.2, the case study with both CCTV and UAV was conducted. The result of situation 2 with the USD 8000 budget was depicted as Figure 5, and the result was derived in 05.93 s. Compared to situation 1, the monitored key points do not change. Therefore, it can be found that it is impossible to apply UAVs with the budget that can barely install CCTVs for mandatory key points (please remind that the installation of UAV station is USD 3000).

Limited budget does not allow monitoring of coastal and ground areas using UAVs. Therefore, additional analysis with increased budget is conducted, and the result is shown in Figure 6. When the budget increases, UAV is applied for coastal and the ground areas. When the budget is USD 11,500, coastal areas with high priority are monitored. Instead, some of the ground areas with medium priority are not monitored. By applying one UAV with increased budget to USD 15,000, it is founded that two coastal areas and one ground area are monitored. The number of ground areas that are monitored by two CCTVs can decrease. Consequently, it can be inferred that the slight increase of the budget can make it able to monitor a diverse area, but difficult to monitor all mandatory, high and medium priority areas. When the budget become USD 15,000, which is almost double compared to the original USD 8500 budget, monitored coastal areas increases. However, the number of ground areas that are monitored decreases. Such results are obtained within 06.78 and 09.01 s for USD 11,500 and 15,000 budget cases.

Adoption of both CCTV and UAV systems can increase the monetary burden for the tourism spots. Therefore, without the proper plan to set the priority of tourism spots, tourism spots might not be monitored even with a reasonable budget. Therefore, by referring to this research, tourism authorities can set the economic and effective strategy for the optimal monitoring system of tourism areas. In addition, the proposed methodology in this study was also able to find the amount of required money to install perfect monitoring system, which means the monitoring of all key locations with service requirements. To obtain this analysis, the proposed mathematical model can be run with a sufficient budget to obtain the total priority of the full monitoring system. By reducing the budget little by little, the proposed model will run repeatedly until the total priority decreases. In this manner, we can obtain an appropriate budget level to install perfect monitoring system. The result of mathematical model indicates that the amount of USD 58,069 is required to monitor all ground and coastal areas using CCTV and UAV systems.

6. Concluding Remarks

Tourism security is drawing more attention, as it can directly affect the satisfaction and protection of tourists. The countries that have the threat of misdemeanors and terrorism experienced decreasing tourists and collapse of tourism industry. Therefore, the system that can provide sustainable tourism should be devised. For this, investigation on the tourism safety would contribute providing the sustainable tourism. Tourism safety is essential to the regions that can be exposed to the threat of terrorism. To prevent such a terrible case, various efforts are being made. One of the representative methods is the installation of CCTV to monitor a crime-ridden district. However, the monitoring range of each CCTV is limited, and installation of CCTV is also limited especially in the case of tourism attractions. To supplement the relatively limitations of CCTV, UAV system can be adopted. UAVs can monitor the coastal area and provide wide security views as they monitor from a location in the sky. However, installation of UAV asks for expensive price to apply compared to CCTV. Thus, CCTV and UAV systems have trade-offs, and the goal of this study is to provide the optimal guideline for installation and operation of CCTVs and UAVs for monitoring considering such trade-offs.

This research considers the situation to determine the proper type of security system. Two different type of security system (CCTV and UAV) are considered. CCTV can only monitor the ground area while UAV can monitor both ground and coastal areas. The installation and operation of a perfect security system is highly needed; however, tourism authorities have limited security budgets in general. Therefore, the plan to install and operate optimal security system within a limited budget should be considered. For this, two mathematical models have been developed and numerical experiments are conducted to investigate the use of CCTV and UAV security system with different budget limits. With the basic budget USD 8500), it is difficult to apply both CCTV and UAV. When about 40% of budget increases USD 11,500) some UAVs can be applied. With the UAVs, coastal areas are monitored, and some lands are monitored. When the budget increases up to almost 100% USD 15,000), enough UAVs to monitor sea can be retained. However, the number of monitored ground locations decreased. In addition, it is derived that total of USD 58,069 is needed to install and operate perfect security system.

The contribution of this research quantitatively investigates the installation and operation of rational security system. Optimal guidelines can be derived with the considerations on the system trade-offs and limited budgets. Of course, the parameters applied in this research are hypothetically generated by referring to the market price. However, it is obvious that the actual cost to apply UAV monitoring system asks for big budget compared to the CCTV monitoring system. Instead, UAV system provides wider monitoring range and movability. At the practical point of view, countries or regions that plan to install security system can adopt actual data, analyze many case studies, and finally derive an optimal security guideline with their budget range using the methodology proposed in this study. They will be able to check the plan that can provide maximum security with their budget. Then, they can modify the priority of key points too. The data-based decision-making system devised in this research will prevent the waste of limited budgets. Moreover, the usage of UAV is getting bigger in different industries, and the way to handle them is becoming more specified. Therefore, though this research considers wide scanning of the tourism spots, additional studies that consider the roles and features of UAVs are expected to be conducted as further research.