2.2.2. Hanger

The hanger has been developed to enable the lifting of GEWPro from a rooftop in the form shown in Figure 2g, and it can be easily installed, disassembled, and moved horizontally in the rooftop (parapet) without additional equipment. The system was designed to be small and light to increase the ease of work, and it can be applied to apartment complex rooftops (parapet) of various thicknesses with easy installation and disassembling.

#### 2.2.3. Paint and Water Supply Device

The paint and water supply device was designed to be located on the ground to reduce the weight of GEWPro and the convenience of the work, and it was developed as an integrated device for the airless pump and solution tank (Figure 2e).

## 2.2.4. Task Management System

The GEWPro was designed to be able to scan the outer wall in three dimensions and automatically detect obstacles and openings on the exterior walls while moving up from the ground. Based on this scanning information, the system performs painting work by automatically making judgments about the areas to be painted while moving down. As shown in Figure 2i, a task managemen<sup>t</sup> system (controller) was developed to allow an operator to monitor and control the current status of GEWPro and the overall condition of the painting work being conducted in real time using a camera attached to the upper part of the robot (Figure 2h).

#### 2.2.5. Work Process of the GEWPro

According to an analysis based on a field examination regarding the exterior wall painting work that is being performed by GEWPro, the process of the exterior wall painting work by GEWPro can be categorized into the following: (1) preparation work; (2) installation work; (3) painting work; (4) disassembly work; (5) horizontal moving work; and (6) demobilization work. Figure 3 presents the detailed work process.

**Figure 3.** Major composition of GEWPro.

#### **3. Development of Performance Evaluation and LCC Analysis Model of GEWPro**

If the economic validity of automated construction robots were not secured, it would be difficult to introduce them to an actual construction field even if their safety and productivity are far superior to the conventional method. Therefore, proving the advantage of an automated method over the conventional method through economic validity analysis is crucial to the research, development, and commercialization of automated construction robots [14]. Thus, this study aims to develop a performance evaluation and LCC analysis model (Figure 4) that reflects the characteristics of building exterior painting work as well as the specifications and performance of GEWPro and verifies the validity of the proposed model and commercialization possibility of GEWPro through a case study.

#### *3.1. Performance Evaluation Model of GEWPro*

The performance evaluation model (Figure 4-Step 2) developed in this study appraised the performance of GEWPro by analyzing the resources required for exterior wall painting work per day, work time, and work productivity. As shown in Figures 1 and 3, work processes starting from "preparation work" through "horizontal moving work" are basically identical in the conventional method and automated method using GEWPro. The difference in productivity between the two methods stems from the difference in allocated resource amount (equipment, labor, etc.) and the required hours of work. In this study, the performance evaluation of GEWPro was carried out by comparing the productivity between conventional and automated methods resulting from work processes starting from "preparation work" through to "horizontal moving work" excluding the "demobilization work", which typically occurs after the daily working hours (8 hours/day).

#### *3.2. LCC Analysis Model of GEWPro*

The LCC analysis of GEWPro can be carried out by analyzing and calculating the additional cost incurred in exterior wall painting work using GEWPro (compared to that of the conventional method) and the benefits derived from the introduction of this automated method [15]. In this study, reasonable assumptions were made to conduct an economic analysis of GEWPro and the LCC analysis model that reflects the initial cost, maintenance cost, interest rate, number of workable days per year, possible total exterior painting area per year, and net profit per year resulting from the automated method was suggested. In addition, the economic feasibility of GEWPro was verified through a case study. For the means of economic analysis of GEWPro, (1) net present worth; (2) benefit/cost ratio; and (3) break-even point analyses were used; and (4) cost saving effect analysis of the automated method being introduced was conducted.

## *3.3. Sensitivity Analysis Model*

In this study, the major factors that influence the performance evaluation and LCC analysis of GEWPro were identified and sensitivity analysis was performed on those factors to improve the reliability of the performance evaluation and LCC analysis result. Figure 4-Step 4 presents a model for sensitivity analysis.

**Figure 4.** Performance evaluation and LCC analysis model of GEWPro.

#### **4. Performance Evaluation and LCC Analysis of GEWPro**

#### *4.1. Buildings for the Case Study*

For a performance evaluation and LCC analysis of GEWPro, apartment complexes comprised of residences, 99.0~115.5 m<sup>2</sup> in size, which form the highest proportion (24%) among all apartments being built in South Korea [1], were considered for the case study. Two apartment complex construction sites, Chungbuk Jincheon (Site A) and Gangwon Wonju (Site B), where exterior wall painting work was in progress at the time of this study, were selected (Table 1). From the floor plans for both sites, the total painting area of Sites A and B was 8275.0 m<sup>2</sup> and 8013.3 m2, respectively, while the total painting area excluded from the study due to the difficulty of GEWPro access (bold lines of Table 1) was 2446.8 m<sup>2</sup> and 1841.0 m<sup>2</sup> for Sites A and B, respectively (Table 1).


**Table 1.** Floor plan and outline of the buildings for the case study.

\* Inaccessible area to GEWPro can be defined as follows: Places where hangers cannot be installed on the parapet of a rooftop because the parapet is not continuously connected or where the width of the area to be painted is more narrow than 2.6 m, the width of GEWPro.

#### *4.2. Performance Evaluation of GEWPro*

#### 4.2.1. Daily Resource Analysis of Conventional and Automated Methods Using GEWPro

As summarized in Table 2, daily resources (RC, RG) consisting of the equipment, laborer, consumables, material, etc. required for conventional exterior painting work and automated method using GEWPro were identified based on an interview conducted with exterior painting work experts as well as field test results using GEWPro.



#### 4.2.2. Total Work Time Analysis of Conventional and Automated Methods Using GEWPro

The total work time required for conventional exterior painting work and the automated method using GEWPro was analyzed using the work sampling technique according to Step 2.2 of Figure 4. The results show that using the conventional method, the total time required for exterior wall painting work (TC) for a single apartment building on both sites (Site A and B) was 3 days (24 h), whereas the automated method using GEWPro (TG) took 20.9 h for Site A and 20.2 h for Site B for a single apartment building (Table 3). The total work time using GEWPro excludes painting work for areas not accessible by GEWPro (bold lines of Table 1).


**Table 3.** Analysis of the total work time required using GEWPro on a single apartment building.

4.2.3. Work Productivity Analysis of Conventional and Automated Methods Using GEWPro

As summarized in Table 4, the hourly work productivity of the conventional method and automated method using GEWPro (WPhC, WPhG, respectively) and single painter's hourly productivity (LPPC) were analyzed according to Step 2.3 of Figure 4. Regarding the exterior painting area, where GEWPro was not accessible (bold lines of Table 1), two additional painters were deployed to carry out the work. Because the conventional exterior painting work and automated method using GEWPro can be done concurrently without any mutual interference, the total required work time used to calculate GEWPro work productivity was determined by taking the larger value between TG and LPPC × 2, which was the time required for two additional painters deployed to work on areas not accessible by GEWPro. Because the time taken for two additional painters to complete the painting work for areas not accessible by GEWPro was less than TG on both case study buildings, the total work time of GEWPro (TG) was used in GEWPro work productivity analysis (Figure 4-Step 2.3).



#### 4.2.4. Performance Evaluation of GEWPro

According to Step 2.4 in Figure 4, a performance evaluation of GEWPro was conducted using Equation (1). In calculating and analyzing the difference in productivity between the conventional and automated methods, the average hourly work productivities (WPhC, WPhG) obtained from Sites A and B for each method was used for a performance evaluation. Performance evaluation analysis

showed that the productivity of an automated method using GEWPro (WPhG) was 16.8% higher than that of the conventional method (WPhG).

$$\begin{aligned} \text{PerformanceEvaluating Result (PER)}\\ \mathbf{h} = \text{(Work Productivity of the automatically method per hour (WPh}\_{\odot})) \\ \text{( $\text{Work Productivity of the conventional method per hour (WPh}\_{\odot})$ )}\\ = \text{(396.4 } \text{m}^2/\text{h})/\text{(339.4 } \text{m}^2/\text{h}) = 1.168 \end{aligned}$$

*4.3. LCC Analysis of GEWPro*

4.3.1. Establishment of the Assumptions and the Variables for LCC Analysis

Prior to performing LCC analysis, assumptions and variables necessary for LCC analysis were established, as shown in Table 5 below.


**Table 5.** Assumptions and variables for LCC analysis.

#### 4.3.2. Analysis of Expected Total Painting Area per Year

Set as a standard workload in performing the LCC analysis, the expected total painting area per year (TPAy) was calculated by multiplying the number of workable days per year (WDy: day/year) by the daily exterior wall painting work productivity (WPhC × 8 h: m2/day). By establishing the unworkable conditions (i.e., average daily temperature of ≤5 ◦C; highest daily temperature of ≥35 ◦C; daily precipitation of ≥10 mm; daily maximum wind speed of ≥10 m/s; and national holidays) derived based on the Suspended Scaffold Safety Regulations of Korea Occupational Safety and Health Agency (KOSHA) [16] and the Painting Specifications of Korea Land and Housing Corporation [17] and by analyzing the recent ten years of weather data [18], the number of yearly workable days (WDy) was calculated to be 162 days. As shown in Step 3.2 of Figure 4, this yearly workable days (WDy) is multiplied by the smaller of the WPhC and WPhG to obtain the possible total painting area per year (TPAy). In this study, the TPAy required for LCC analysis was calculated to be 439,862.4 m2/year according to Equation (2).

> Expected Total Painting Area per year (TPAy) = Work Productivity of the automated per hour (WPhC) × 8 h/day × the number of Workable Days per year (WDy) = 339.4 m2/h × 8 h/day × (162 day/year) = 439,862.4 m2/year

(2)

=

#### 4.3.3. Annual Benefit Analysis of the Automated Method Using GEWPro

According to Step 3.3 in Figure 4, the annual net profit of introducing the automated method using GEWPro was calculated by multiplying the expected total painting area per year (TPAy) by the difference between the expenses of the conventional method per square meter (Em2C) and the automated method per square meter (Em2G). The expenses per unit area (m2) for conventional and automated methods were taken from the sum of all the necessary expenses of applied resources (laborer, consumables, material) excluding the equipment expense, which is applied directly to the cash-flow diagram. In measuring the expenses for resources, the unit prices actually utilized in the field were applied. As shown in Table 6, which summarizes the resources and expenses for conventional and automated methods, the expenses of the conventional method per unit area (Em2C) were \$0.9351/m2, whereas the expenses for automated method per unit area (Em2G) were calculated to be \$0.7998/m2. Therefore, the annual present worth benefit (AB) of applying GEWPro was calculated to be \$59,520.1/year according to Equation (3).

> Annual Benefit by the automated method (AB)

= Expenses of the conventional method per m<sup>2</sup> Em2C − Expenses of the automated method per m<sup>2</sup> Em2G × Expected Total Painting Area per year (TPAy) \$0.9351/m<sup>2</sup> −\$0.7998/m<sup>2</sup> ×439,862.4m2/year=\$59,520.1/year(3)


**Table 6.** Resources and expenses for the conventional and automated methods.

#### 4.3.4. Cash-Flow Diagram for LCC Analysis of GEWPro

Using the set assumptions and variables established previously and reflecting the expenses calculated for conventional and automated methods, Figure 5 presents the cash-flow diagram of GEWPro for the durability life of 10 years. Table 7 also lists the net present worth (NPW) and equivalent annual worth (EAW) derived from the annual benefit by the automated method (AB), which was reflected in the cash-flow diagram and necessary expenses (i.e., initial cost, maintenance cost) of equipment resources (rEC, rEG) for conventional and automated methods according to Step 3.4 of Figure 4.

**Figure 5.** Cash-flow diagram for GEWPro's LCC analysis.


**Table 7.** Cash-flow analysis (present worth and annual worth).

#### 4.3.5. LCC Analysis of GEWPro


Net present worth analysis is a method for comparing the cash flow of a subject under economic analysis based on its present value. In this study, the present worth was calculated based on the cost difference incurred over ten years of operation (expected service life of GEWPro) using the conventional method versus automated method using GEWPro. According to the economic analysis using net present worth, the present worth of net profit was calculated to be \$430,404, as shown in Equation (4).

## Net Present Worth (NPW)

$$\begin{array}{l} = & \text{Present worth of benefit} - \text{Present worth of cost} \\ = & (\text{EP}\_{\text{C}} - \text{EP}\_{\text{G}}) - \left\{ (\text{IP}\_{\text{G}} - \text{IP}\_{\text{C}}) + (\text{MP}\_{\text{G}} - \text{MP}\_{\text{C}}) \right\} \\ = & (\\$3, 526, 789.7 - \\$3, 016, 470.2) - \left\{ (\\$60, 567.7 - \\$14, 994.5) + (\\$4, \\$994.5) + (\\$70, \\$139.5 + \\$4, \\$359.4) \right\} \\ & (\\$6139.5 + \\$14, 359.4 + \\$12, \\$07.8 + \\$5701) - \\$4665.5 \right\} \\ & = & \\$430, 404 \end{array} \tag{4}$$

=
