MICUSP Version 1.0 - MEC.G0.03.

1 - Mechanical Engineering - Final Year Undergraduate - Male - Native Speaker - Proposal 1

Manufacturing System Design for a Car Assembly Process

Michigan Corpus of Upper-level Student Papers. Version 1.0 Ann Arbor, MI. Copyright (c) 2009 Regents of the University of Michigan
MICUSP Version 1.0 - MEC.G0.03.1 - Mechanical Engineering - Final Year Undergraduate - Male - Native Speaker - Proposal 2

EXECUTIVE SUMMARY

We have been assigned the task of designing the manufacturing system for the assembly of a
simple vehicle. The manufacturer wishes to anticipate any changes in demand, and also wishes to
know the impact of reliability at the system and machine levels (100% system and 95% machine
reliability). As a result, three scenarios of demand and reliability were considered: (1) 50,000
units per year with 100% reliability, (2) 50,000 units per year with 95% machine reliability and
lastly (3) a 75,000 units per year with 95% machine reliability. This task has been completed.

By performing tasks in parallel at the same station, the total number of stations required for
assembly was reduced to 7. In turn, the system reliability was increased in order to meet the
yearly demand. Specifically, for scenario 1, the plant can operate a single 8 hour shift for 51
weeks per year. For scenarios 2 and 3, respectively, two and three 6 hours-a-day shifts for 49
weeks per year can be used to meet the yearly demand. Lastly, a future state map is included to
implement lean manufacturing methods and reduce the costs of operation. The approach, costs
and recommendations regarding the assembly designs are discussed in this report.

Michigan Corpus of Upper-level Student Papers. Version 1.0 Ann Arbor, MI. Copyright (c) 2009 Regents of the University of Michigan
MICUSP Version 1.0 - MEC.G0.03.1 - Mechanical Engineering - Final Year Undergraduate - Male - Native Speaker - Proposal 3




 TABLE OF CONTENTS
1
 PROBLEM
STATEMENT ..............................................................................................................................................4

2
 COMPONENTS
AND
MANUFACTURING
TASK
INFORMATION........................................................................6

3
 ASSUMPTIONS ...............................................................................................................................................................8

4
 TASK
ANALYSIS.............................................................................................................................................................9

4.1
 PRECEDENCE
GRAPH ..................................................................................................................................................9

4.2
 PRODUCTION
LINE
SCHEME .................................................................................................................................. 10

5
 DESIGNS........................................................................................................................................................................ 11

5.1
 APPROACH................................................................................................................................................................... 12

5.1.1
BUFFER
IMPLEMENTATION................................................................................................................................... 12

5.1.2
MODIFICATION
OF
LABOR
HOURS ...................................................................................................................... 13

5.2
 COST
OF
PRODUCTION ............................................................................................................................................ 13

5.3
 1
YEAR
AND
5
YEAR
PROFIT
FORECAST............................................................................................................ 14

6
 CONCLUSIONS
AND
RECOMMENDATIONS ........................................................................................................ 14

6.1
 FUTURE
STATE
MAP................................................................................................................................................. 15


Michigan Corpus of Upper-level Student Papers. Version 1.0 Ann Arbor, MI. Copyright (c) 2009 Regents of the University of Michigan
MICUSP Version 1.0 - MEC.G0.03.1 - Mechanical Engineering - Final Year Undergraduate - Male - Native Speaker - Proposal 4

1 PROBLEM STATEMENT
We have been tasked to design the manufacturing system for the assembly of a simple car. It is
projected that the manufacturer will need to produce at least 50,000 units per year during the first
years of operation. One case is considered for 100% reliability, while another for 95% machine
reliability. The manufacturer wants to anticipate changes in demand which may quickly rise to
75,000 units per year. The manufacturer wants to know what system setup will meet the
manufacturing goal under three discreet conditions, shown in Table 1, below.

Table 1: Demand and Reliability Scenarios

Scenario Units per Year Machine Reliability

1 50,000 100%

2 50,000 95%

3 75,000 95%

Additionally, we have been informed of the following conditions regarding financial and
physical details of the manufacturing system.

Table 2: Variable and Fixed Costs for the System (in USD)
1 Average Cost per Station $100,000

2 Average Cost per Buffer $10,000

8 hours/day (max)
3 Maximum and Minimum Working Hours per Shift
5 hours/day (min)

$40 per hour (normal)

4 Labor Cost and Overtime Labor Cost
$60 per hour (overtime)

5 Car Length 4.5 m

6 Distance Between Station Center Points 9m

7 Line Width (incl. corridors) 13.5 m

Michigan Corpus of Upper-level Student Papers. Version 1.0 Ann Arbor, MI. Copyright (c) 2009 Regents of the University of Michigan
MICUSP Version 1.0 - MEC.G0.03.1 - Mechanical Engineering - Final Year Undergraduate - Male - Native Speaker - Proposal 5

Michigan Corpus of Upper-level Student Papers. Version 1.0 Ann Arbor, MI. Copyright (c) 2009 Regents of the University of Michigan
MICUSP Version 1.0 - MEC.G0.03.1 - Mechanical Engineering - Final Year Undergraduate - Male - Native Speaker - Proposal 6

2 COMPONENTS AND MANUFACTURING TASK INFORMATION

Figure 1: Schematic of car components and sub-assemblies

6

Michigan Corpus of Upper-level Student Papers. Version 1.0 Ann Arbor, MI. Copyright (c) 2009 Regents of the University of Michigan
MICUSP Version 1.0 - MEC.G0.03.1 - Mechanical Engineering - Final Year Undergraduate - Male - Native Speaker - Proposal 7

Table 3: Task breakdown and information

Task Task Predecessor
Operators
Task Name Time
number Required Tasks
(seconds)

1 1.1 Install 2 front suspensions 43 2

2 1.2 Install 2 front wheels 26 2 1.1

3 1.3 Install real axle 42 2

4 1.4 Install 2 rear wheels 26 2 1.3, 2.4

5 2.1 Insert the engine 27 2

6 2.2 Secure the engine 47 2 2.1

7 2.3 Insert /secure transmission 55 2 2.2

8 2.4 Install the drive shaft 31 1 1.3, 2.3

9 3.1 Install the exhaust pipe 44 2 2.2

10 4.1 Install wiring harness* 75 2

11 4.2 Attach dashboard 42 2 4.1

12 4.3 Attach steering wheel 32 1 4.2

13 5.1 Attach front and rear windows 35 2

14 5.2 Seal windows 45 4 5.1

15 6.1 Install 2 front seats 26 2 4.3, 6.3

16 6.2 Install the rear seat 31 2 6.3

17 6.3 Install floor carpet 94 2 4.1

18 7.1 Attach hood 12 2 2.2

19 7.2 Attach deck lid 17 2

20 7.3 Install 2 front doors 43 2 6.1

21 7.4 Install 2 rear doors 44 2 6.2

22 8.1 Fluid fill 102 4 All except 9.1

23 9.1 Inspection 145 2 All others

Michigan Corpus of Upper-level Student Papers. Version 1.0 Ann Arbor, MI. Copyright (c) 2009 Regents of the University of Michigan
MICUSP Version 1.0 - MEC.G0.03.1 - Mechanical Engineering - Final Year Undergraduate - Male - Native Speaker - Proposal 8

Total Task Time 1084 seconds, ~18 minutes

* Wiring harness goes under the floor carpet along the cabin

3 ASSUMPTIONS
Based on the given task information as described in Section 2, several assumptions were made to
facilitate the construction of a liaison diagram. The assumptions made are listed below.

Working Hours

1. Shifts can be 6 to 8* hrs/day, 5 days/week, 49 to 51* weeks/year

* Varies with design scenarios

Manufacturing Operations

2. Inspections done in a different, quiet and clean zone

3. Task time includes all mounting and un-mounting during installation of components

Machines

4. Each task needs 1 machine

5. Each station used has 95% efficiency (Design #2 & Design #3)

6. Failure of one machine does not cause the failure of another machine

Employee

7. Any operator can work on any station, and operators will complete a scheduled
rotation for all stations

8. Workers will not be absent for sick days or vacations

Sub-assembly supply

Michigan Corpus of Upper-level Student Papers. Version 1.0 Ann Arbor, MI. Copyright (c) 2009 Regents of the University of Michigan
MICUSP Version 1.0 - MEC.G0.03.1 - Mechanical Engineering - Final Year Undergraduate - Male - Native Speaker - Proposal 9

9. There is uninterrupted flow of sub-components and parts

10. The engine comes with its own sub-wiring harness

11. The ECU comes with the wiring harness as it is installed

12. The vehicle body is present at each station of the assembly

The assumptions relating to cost and profit calculations are discussed in the Designs section.

4 TASK ANALYSIS
In order to design the manufacturing system for this application, the assembly tasks have been
analyzed and simple heuristics were used to create a precedence graph and a production line
scheme. Task information, task times and number of operators were taken into account, as well
as the components needed to complete a task and the predecessor tasks for each task.

4.1 Precedence Graph

According to task information, a precedence graph was drawn, as shown in Figure 2, below.

Michigan Corpus of Upper-level Student Papers. Version 1.0 Ann Arbor, MI. Copyright (c) 2009 Regents of the University of Michigan
MICUSP Version 1.0 - MEC.G0.03.1 - Mechanical Engineering - Final Year Undergraduate - Male - Native Speaker - Proposal 10

Figure 2: Task precedence graph

4.2 Production Line Scheme

A production line schematic was generated for each design and is shown in Figures 3 and 4,
below. These were drawn in accordance to space requirements listed by the manufacturer. The
space occupied for Design 1 is 972 m2, while Designs 2 and 3 occupy 1215 m2. The approach
used to create the production line schemes is detailed in section 5.1.

10

Michigan Corpus of Upper-level Student Papers. Version 1.0 Ann Arbor, MI. Copyright (c) 2009 Regents of the University of Michigan
MICUSP Version 1.0 - MEC.G0.03.1 - Mechanical Engineering - Final Year Undergraduate - Male - Native Speaker - Proposal 11

Figure 3: Production Line Scheme for Design #1

Figure 4: Production Line Scheme for Design #2 & #3

5 DESIGNS
This section outlines the approach, implementation and outcome of the three demand and
reliability scenarios discussed in Section 1.

11

Michigan Corpus of Upper-level Student Papers. Version 1.0 Ann Arbor, MI. Copyright (c) 2009 Regents of the University of Michigan
MICUSP Version 1.0 - MEC.G0.03.1 - Mechanical Engineering - Final Year Undergraduate - Male - Native Speaker - Proposal 12

5.1 Approach
To obtain a possible task group according to takt time, Ranked Positional Weight (RPW)
heuristics were performed for the manufacturing process design. The results of RPW can be seen
in Table 4, below. When assembly order was determined, we considered the difference in
number of operators for each task along with task precedence. For example, task 5.2 requires 4
operators and must be done after task 5.1. So the four operators of task 5.2 can perform two other
two-operator tasks in parallel, after task 5.1 is completed. With more operators, it was possible to
do tasks in parallel at the same station. These are indicated by tasks that are grouped in
parentheses. Thus, it was possible to use 7 assembly stations, with only one instance of
significant idle time in station 5.

Table 4: RPW Results

Idle Process Required Buffer
Station Assigned tasks time time operators size
1 4.1, 2.1, 4.2 2 144 2
2 6.3, 2.2 5 141 2
3 2.3, (4.3 & 2.4), 5.1, 7.2 7 139 2 1*
4 6.2, 1.1, 6.1, 1.3 4 142 2 1*
5 5.2, (3.1 & 1.2), (1.4 & 7.1) 31 115 4
6 (7.4 & 7.3 ) 8.1 0 146 4
7 9.1 1 145 2

* Buffers apply only to Designs 2 and 3.

5.1.1 Buffer Implementation

Using PAMS software version 4.2 to verify and improve our assembly process, we determined
that we should incorporate two buffers into Designs 2 and 3, between stations 3 and 4, and 4 and
5. The size of each buffer should only be 1 car, and the implementation of buffers would result in
a higher system reliability and throughput. The PAMS approach is further discussed in Appendix
A.

12

Michigan Corpus of Upper-level Student Papers. Version 1.0 Ann Arbor, MI. Copyright (c) 2009 Regents of the University of Michigan
MICUSP Version 1.0 - MEC.G0.03.1 - Mechanical Engineering - Final Year Undergraduate - Male - Native Speaker - Proposal 13

5.1.2 Modification of Labor Hours

With the decreased machine reliability of Designs 2 and 3, it is essential to decrease the number
of stations and maximize the system reliability. Using seven stations in series, the overall
reliability is 0.957 or 69.8%. To counteract the impact of significantly decreased production,
three labor hour modifications were made: (1) a shift was added, (2) the length of the work day
was reduced and (3) the number of weeks of operation per year was decreased. This will allow
the manufacturer to produce more than 50,000 units per year and fulfill the requirements of
Design 2, without implementing a second assembly line and without changing the task
assignment order.

In order to reach the production goal of 75,000 units for Design 3, a third shift can be added and
will work with the same reliability of Design 2. The production outcome and labor hours of each
design can be seen in Table 5, below.

Table 5: Production Results with Modified Labor Hours

Labor
Design # Production # of Shifts Hours/shift Weeks/year Hours
1 50301.37 1 8 51 36720
2 50624.67 2 6 49 52920
3 75937.01 3 6 49 79380

5.2 Cost of Production

The costs of production for each of the proposed designs can be seen in Table 6, below. Each
design requires 7 stations to be constructed, while designs 2 and 3 require two buffers to be
added. The labor costs are determined from the number of hours in production as well as the
number of operators on each station. To implement designs 2 and 3, no machine reconfiguration
is required, since the layout is very similar with the difference being the addition of two buffers.

Table 6: Variable and Fixed Cost Breakdowns

Cost Design 1 Design 2 Design 3
Station $100,000 7 7 7
Buffer $10,000 0 2 2
Labor $40 / hr 36720 hrs/yr 52920 hrs/yr 79380 hrs/yr
Overtime $60 / hr 0 hrs/yr 0 hrs/yr 0 hrs/yr
Total Cost $2,168,800 $2,836,800 $3,895,200
13

Michigan Corpus of Upper-level Student Papers. Version 1.0 Ann Arbor, MI. Copyright (c) 2009 Regents of the University of Michigan
MICUSP Version 1.0 - MEC.G0.03.1 - Mechanical Engineering - Final Year Undergraduate - Male - Native Speaker - Proposal 14

5.3 1 Year and 5 Year Profit Forecast

For the first year of production, scenarios 1 and 2 will incur losses financially. After 5 years the
profit should start to exceed the initial investment cost. Design 3 will be profitable from the first
year. A profit analysis can be seen in Table 7, below. In order to calculate the profit margin after
both 1 and 5 years, we assumed the following:

1. Overhead cost to make each car was $3,200.

2. Value of the material in each car was $10,000.
3. The Manufacturer Suggested Retail Price (MSRP) of the cars is $14,800 for Designs 1
and 2, and $14,480 for Design 3.
4. The cost of marketing the car was $70 million for the first two Designs, but greater for
Design 3 because of the quantity needed to be sold.
5. Administration costs were $5.5 million per year, regardless of number of cars sold.
6. The cost to rent or own the factory was $3 million, regardless of cars sold.
7. All cars that were produced were sold.

Table 7: Design 3 will be profitable after 1 year, Designs 2 and 3 after 5 years
Design 1 Design 2 Design 3
Overhead/car -$3,200 50301.37 50624.67 75937.01
Material/car -$10,000 50301.37 50624.67 75937.01
MSRP of car $14,800* 50301.37 50624.67 74295.13
Marketing/yr $ -70,000,000 1.00 1.00 1.20
Administration/yr $ -5,500,000 1.00 1.00 1.10
Factory rental/yr $ -3,000,000 1.00 1.00 1.00
Total profit for 1 year: $ -186,608.22 $ -337,326.66 $ 254,168.00
Total profit for 5 years: $ 1,866,958.90 $ 1,193,366.68 $ 4,150,840.01
* This price is decreased to $14,480 for Design 3, because of the quantity sold

6 CONCLUSIONS AND RECOMMENDATIONS

We were asked to create a process design for the assembly of a car that composed of 23 different
tasks. Through the implementation of doing tasks in parallel at the same station, we were able to
reduce the total number of stations to 7. Minimizing the number of stations for assembly had a

14

Michigan Corpus of Upper-level Student Papers. Version 1.0 Ann Arbor, MI. Copyright (c) 2009 Regents of the University of Michigan
MICUSP Version 1.0 - MEC.G0.03.1 - Mechanical Engineering - Final Year Undergraduate - Male - Native Speaker - Proposal 15

significant impact on system reliability, given the reduced machine reliability of 95% for
scenarios 2 and 3.

In order to meet the 50,000 unit demand per year for scenario 1, Design 1 can be implemented.
This entails operating the plant for 51 weeks per year, with a single, 8 hour daily shift, using the
production line scheme shown in Figure 3, on page 9. To meet the demand of scenarios 2 and 3
with their respective machine reliability of 95%, we altered the labor hours while keeping the
same task assignment order. Specifically, Design 2 uses two 6 hour daily shifts for 49 weeks per
year, under the production scheme shown in Figure 4, on page 9. Lastly, Design 3 uses three 6
hour daily shifts for 49 weeks per year, under the same production scheme.

6.1 Future State Map

Applying value stream mapping enables us to generate a Future State Map of the manufacturing
process, as seen in Figure 5, below. The following explains the lean manufacturing principles
that are implemented.

Figure 5: Future state mapping uses principles of lean manufacturing to reduce waste

15

Michigan Corpus of Upper-level Student Papers. Version 1.0 Ann Arbor, MI. Copyright (c) 2009 Regents of the University of Michigan
MICUSP Version 1.0 - MEC.G0.03.1 - Mechanical Engineering - Final Year Undergraduate - Male - Native Speaker - Proposal 16

Utilizing a supermarket after the stamping process, we will be able to set a "pull" system through
the manufacturing line. 1 piece flow and first-in-first-out part flow are implemented between
stations, increasing the system's predictability and reduces inventory. The customer also sends
30, 60 and 90 day forecasts, as well as daily orders to the production control, which in turn sends
the aforementioned information to the steel plant. This enables just-in-time manufacturing which
greatly reduces inventory. After each batch of orders produced, the supermarket will also send a
production Kanban to the production control to have constant update of production information.

16

Michigan Corpus of Upper-level Student Papers. Version 1.0 Ann Arbor, MI. Copyright (c) 2009 Regents of the University of Michigan
MICUSP Version 1.0 - MEC.G0.03.1 - Mechanical Engineering - Final Year Undergraduate - Male - Native Speaker - Proposal 17

APPENDIX A - PAMS Analysis

We used PAMS to help verify that our system designs would function as we thought they would
in our initial calculations.

Iterative Approach in PAMS System Design

We initially created a 7 station system using PAMS as shown below, in Figure 6. We gave each
station the processing time shown in Table 4, on page 10. For Design 1 with 100% efficiency, we
achieved our expected throughput as shown in Table 8, on page 15. For Designs 2 and 3, we
assumed each station would break down twice per shift. In an attempt to give each station a 95%
reliability, we gave each station a MTBF of 180 minutes and a MTTR of 9.5 minutes. This yields
approximately 95% reliability since Reliability = MTBF / (MTBF + MTTR). However, in the
PAMS simulation we were only achieving a breakdown rate of about 4%, yielding a 96%
reliability.

Figure 6: PAMS manufacturing system layout

In order to increase the breakdown rate to 5%, we increased the MTTR incrementally until we
achieved a 5% breakdown rate in the PAMS simulation. However, once this goal was
accomplished, we found that our throughput was below the target throughput.

In order to improve our throughput, we decided to add buffers. To determine the best place to
add buffers, we ran a PAMS simulation which improved throughput given a buffer size of 5. It
focused the buffer size at Buf_3 and Buf_4 as shown in Figure 6, above. Using the PAMS output
17

Michigan Corpus of Upper-level Student Papers. Version 1.0 Ann Arbor, MI. Copyright (c) 2009 Regents of the University of Michigan
MICUSP Version 1.0 - MEC.G0.03.1 - Mechanical Engineering - Final Year Undergraduate - Male - Native Speaker - Proposal 18

regarding the best placement and size of buffers, we added one unit to Buf_3 and Buf_4 and left
the other buffers at 0. By doing this, our target demand per year was met. The average work in
progress (WIP) can be seen in Figure 7 below. Throughput for each design can be seen in Table
8, below.

Figure 7: The average WIP in each station for Designs 2 and 3

Table 8: Throughputs using PAMS simulation for each design

Design 1 Design 2 Design 3
Throughput (parts/min) 0.411 0.284 0.284
Shifts 1 2 3
Shift duration (min) 480 360 360
Days/week 5 5 5
Weeks/year 51 49 49
Cars/year 50306 50098 75146
Cars/day 197 204 307

18

Michigan Corpus of Upper-level Student Papers. Version 1.0 Ann Arbor, MI. Copyright (c) 2009 Regents of the University of Michigan