CHAPTER 4: MODEL IMPLEMENTATION AND VALIDATION
4.1.2. Case study 2 – Bloomsdale Residence
The second case study focuses on a substantial finishing process project comprising 51 activities. The aim of this case study is to assess the suitability of the model for a real- world project that involves a substantial number of activities and multiple interdependencies among them. The aim is to determine whether the model can effectively handle the intricacies and challenges posed by such a project scenario.
4.1.2.1. Project information
Project’s information is shown as follow:
Project name: Bloomsdale Residence Project location: Haiphong City, Vietnam Total existing site area: 𝟒𝟖𝟎. 𝟐𝟓𝒎𝟐
Ground floor site area: 𝟐𝟑𝟐. 𝟔𝟏𝒎𝟐
Architectural design: SOCON VIETNAM Finishing sub-contractor: SOCON VIETNAM Finishing work duration: 70 days
Figure 4.30: Some perspective views of the project
Figure 4.31: Project’s ground floor plan
Figure 4.32: Project's first floor plan
In this project, the structure and outside wall of the project are built, the finishing works will be completed by sub-contractor. Table below show the detail finishing work of the residence. Note that the detail works are displayed in the ANNEX A: Case study 2 detail work in MS Project.
Table 4.16: Project's activity, duration, predecessors of case study 2
Activity Duration (days)
Predecessors Activity Duration (days)
Predecessors
1 5 - 27 3 26
2 2 1 28 3 26
3 5 1 29 3 28
4 2 3 30 3 27
5 3 2 31 2 30
6 3 4 32 6 31
7 6 1,3 33 5 30,31
8 4 7 34 2 33
9 3 8 35 4 33
10 4 8 36 2 35
11 2 10 37 2 34
12 3 9 38 2 37
13 3 11 39 7 30
14 4 5 40 7 39
15 4 6 41 5 32
16 1 7 42 5 41
17 6 7 43 14 39,40
18 5 17 44 5 43
19 10 14,15 45 7 37,38
20 7 14,15 46 4 41,42
21 1 7 47 1 32
22 11 12,13 48 2 21,37,38
23 3 16,17,18 49 7 43,44
24 5 19 50 5 49
25 5 24 51 2 50
26 7 3
4.1.2.2. Step 1: Calculate project schedule
The detail values of ES, EF, LS, LF and TF of the second case study are shown in the Appendices. The total project duration is 70 days and the critical path is 1-3-26-27-30- 39-40-43-44-49-50-51. The detail table is illustrated in the ANNEX B: Case study 2 detail project time.
4.1.2.3. Step 2-3: Check for schedule overlapping and calculate SOR value
Similar calculation as part 4.1.1.2 and 4.1.1.3 of this thesis. The summary value of the initial SOR is 208,0134.
4.1.2.4. Step 4: Risk analysis
The value of column and row by summing and dividing by 10 are represented for Probability and Intensity to analyze risk. Besides, by checking clash detection in Navisworks, project manager can get more information to analyze risk and from his/her own experience perspective. The table below shows the top 10 risky activities in the project. Note that the risk degree based on five Euclidean distance values which has been mentioned in part 3.2 of this thesis.
Table 4.17: Top 10 risky activities of the project based on 5 Euclidean distance values
No. Activity Description Duration (days)
Probability Intensity Output (Mamdani)
Risk degree
Risk cost($)
20 Installing doors in
ground floor 7 days 1.0711905 0.6 0.8 VH 800
19
Installing windows system in ground floor
10 days 1.4407143 0.57 0.77931 H 600
39 Wall tiles in 1st floor
(Zone 1) 7 days 0.9202814 0.557143 0.77349 H 600
40 Wall tiles in 1st floor
(Zone 2) 7 days 1.0045887 0.514286 0.72405 H 400
17 Floor tiles in ground
floor (Zone 1) 6 days 0.9121429 0.5 0.7 H 100
22 Lighting system in
ground floor 11 days 1.3595238 0.481818 0.7 H 70
41 Floor tiles in 1st floor
(Zone 1) 5 days 0.7418831 0.52 0.62969 H 200
18 Floor tiles in ground
floor (Zone 2) 5 days 0.7209524 0.6 0.6084 H 200
32 1st floor screeding 6 days 0.6965152 0.533333 0.59767 M 40
24
Outside wall plastering in ground floor
5 days 0.6635498 0.52 0.57452 M 300
4.1.2.5. Step 5: Calculate initial overlapping duration and total risk cost
The risk cost value of each activity will be determined by equation (16). Besides, the initial overlapping duration and the risk cost value of each activity will be calculated as same as part 3.3.6 of this thesis.
Risk cost of each activity is not only based on the Mamdani Fuzzy output but also based on the project manager’s experience and the current activity situation. For example, activity number 20 has the highest risk value and the risk cost is 800, while activity number 24 has the risk cost is 300 even the output is 0.57452. The figure below shows the risk degree by color in NAVISWORKS:
Figure 4.33: Risk degree of activity number 20 and 24, respectively
The value of risk cost between a pairwise of two comparison activities is calculated as equation. After run the calculation in MATLAB, the total overlapping days at initial is 431 days and the total risk cost is $258,058.
4.1.2.6. Step 6, 7: Schedule optimization (SOR optimization) and time cost trade-off GA options in MATLAB:
popSize = 500;
maxGenerations = 500;
crossoverFraction = 0.8;
stallGenLimit = 500;
Figure 4.34: Multi objective optimization in Case study 2
The best solution for objective 1 (Total overlapping days) is 267 days with the following movable duration as shown below:
Table 4.18: Movable duration corresponding with objective 1
Activity Movable days
Activity Movable days
Activity Movable days
1 0 18 32 35 13
2 0 19 23 36 24
3 0 20 9 37 27
4 0 21 10 38 0
5 0 22 8 39 0
6 0 23 40 40 0
7 0 24 31 41 16
8 0 25 4 42 8
9 0 26 0 43 0
10 0 27 0 44 0
11 0 28 32 45 18
12 0 29 24 46 13
13 0 30 0 47 10
14 0 31 10 48 3
15 0 32 0 49 0
16 0 33 7 50 0
17 11 34 7 51 0
The best solution for objective 2 (Total risk cost) is 128,697 and total overlapping with the following movable duration as illustrated below:
Table 4.19: Movable duration corresponding with objective 2
Activity Movable days
Activity Movable days
Activity Movable days
1 0 18 30 35 13
2 0 19 22 36 11
3 0 20 7 37 24
4 0 21 17 38 27
5 0 22 3 39 0
6 0 23 40 40 0
7 0 24 11 41 25
8 0 25 1 42 11
9 0 26 0 43 0
10 0 27 0 44 0
11 0 28 47 45 10
12 0 29 19 46 18
13 0 30 0 47 38
14 0 31 10 48 8
15 0 32 0 49 0
16 0 33 18 50 0
17 19 34 27 51 0
Best solution for objective 3 (SOR optimization). The summary value of the population SOR is 119.2721.
The fitness value is 119.2721
208.0134 = 0.5734
The movable duration of each activity is shown in the table below:
Table 4.20: Movable duration exported from MATLAB
Activity Movable days
Activity Movable days
Activity Movable days
1 0 18 7 35 29
2 0 19 31 36 31
3 0 20 16 37 17
4 0 21 32 38 0
5 0 22 0 39 0
6 0 23 28 40 0
7 0 24 18 41 0
8 0 25 15 42 1
9 0 26 0 43 0
10 0 27 0 44 0
11 0 28 45 45 4
12 0 29 22 46 0
13 0 30 0 47 0
14 0 31 3 48 32
15 0 32 0 49 0
16 0 33 18 50 0
17 23 34 19 51 0
3D MODEL REVIT (.RVT) NAVISWORKS (.NWF)
EXPORT
MS PROJECT (.MPP)
EXPORT CSV. FILE
TIMELINER
Extract activity number/Work Breakdown Structure/Name/ES/EF/
Duration
Input Relationships between activities User
Input?
No
Generate CPM Schedule Data
CPM Analysis
Add CPM Parameters to MS Project to recheck
ParametersCPM
Act. No.
Name ESEF LS LF TF Relationship
Refer to CPM Schedule Parameters
Analyze Schedule Overlap
Schedule Overlapping
Check
Execute Sequential Check Algorithm of Schedule Overlap
Check? No
Generate Initial Schedule Overlap Parameters
Yes
Identify High Overlap Activity with SOR
Update Base Act.
SOR value Yes
Input Fuzzy Parameters based on SOR values
Finished? No
Evaluate Schedule Overlap Risks
Fuzzy Simulink
Fuzzy Mamdani
Update Parameters of Schedule Overlap Risks and calculate Risk
Money (If necessary)
Identify Overlapped Activities with Risk Degree
Recheck High Risky Activities by User s perspective
User Input GA parameters
Execute Schedule Overlap Optimization Genetic
Algorithm
Generate Optimized Schedule Objective (SOR,
Time-Cost), Fitness Function
Update Optimized Schedule Data popSize maxGeneration crossoverFraction stallGenLimit
Update
MATLAB PROCESS USING GENETIC ALGORITHM TO OPTIMIZE PROJECT OVERLAPPING ACTIVITY AND TIME-COST
CPM Analysis Schedule Overlap Check Fuzzy Analysis GA Optimization
Figure 4.35: Process of applying GA in MATLAB for achieving Multi-Objective Optimization