Assembly Line Balancing - Yaşar Üniversitesi · Assembly Line Balancing Gonca Karaarslan Ali Galip Matlar PROBLEM DEFINITION Line Balancing Problem at Indesit Industrial and Trade

Page 1

Assembly Line Balancing

Gonca Karaarslan

Ali Galip Matlar

PROBLEM DEFINITION

Line Balancing Problem at Indesit Industrial and Trade Inc. Company

in refrigerator factory for Lines 6 & 7

Two Sided

AssemblyPre-Assembly

Takt Time: C = 86 sec (Polyurethane Machine)

MM _ D _ S2 (I)*

Straight-2 Sided Line

Multi Model DeterministicType-I: Min # Stations given C

CLASSIFICATION

GOLDEN & STRIKE ZONEAnalysis of work operations in the area

that favors the handling in order to

minimize movement and reduce fatigue.

Mechanic Shop Plastic Shop Paint Shop

Door FoamingPre-Assembly

Assembly

Quality Control Packaging Delivery

MANUFACTURING PROCESSES

SETST: Set of tasks

S: Set of stations

SL: Left side of stations; SL

SR: Right side of stations; SR

SE: Set of stations that can be used for operations

which required zoning restriction; SE

TL: Set of tasks that can be assigned to left side; TL

TR: Set of tasks that can be assigned to right side; TR

β : Set of concurrent task groups

TC : Pair of tasks belonging to concurrent task group c

that must be performed simultaneously, c ∈ β

ℸ : Set of marriage tasks groups

Tα : Set of tasks belonging to group α that must be

assigned to same station, α ∈ ℸ

Pi : Set of immediate predecessors of task i, i ∈ T

⊂S

⊂ S

⊂S

⊆

⊆ T

T

TB : Set of tasks that can be assigned to both sides; TB

TE : Set of tasks belonging to operations

which required zoning restriction; TE T

RightLeft

S5

S3

S1

S6

S2

S4

T1.1 T1.2

T1

T2

T3

Reducing Non-Value AddedOperations

Decreasingthe number of

stations

Improvementof total workload

Minimizing theIdle Time =

Maximizing the station utilization

Decreasing the number of workers =

Reduction of labor cost

KEY PERFORMANCE INDICATORS

Min 𝒋∈𝑺𝒚𝒋

𝒋∈𝑺𝑳

𝒙𝒊𝒋 = 1 , ∀𝒊 ∈ 𝑻𝑳

𝒋∈𝑺𝑹

𝒙𝒊𝒋 = 1 , ∀𝒊 ∈ 𝑻𝑹

𝒋∈𝑺𝒙𝒊𝒋 = 1 , ∀𝒊 ∈ 𝑻𝑩

𝒊∈𝑻𝒕𝒊 𝒙𝒊𝒋 ≤ 𝑪 𝒚𝒋 ∀𝒋 ∈ 𝑺

1+ 𝒋∈𝑺𝑳

𝒋 𝒙𝒊𝒋 = 𝒋∈𝑺𝑹

𝒋 𝒙𝒌𝒋

∀𝒊 ∈ 𝑻𝑳 ∩ 𝑻𝑪 , 𝒌 ∈ 𝑻𝑹 ∩ 𝑻𝑪, 𝒄 ∈ 𝜷

𝒋∈𝑺𝑬

𝒙𝒊𝒋 = 1 ∀𝒊 ∈ 𝑻𝑬

𝒙𝒊𝒋 ∈ {0,1}

𝒚𝒋 ∈ {0,1}

ASSIGNMENT

Each task must be assigned

to an eligible station

CYCLE TIME

Total duration of tasks assigned

to a station cannot exceed cycle time

PRECEDENCE

A task can be assigned to a station

if all predecessors are assigned

previously

SIGN RESTRICTIONS

Binary decision variables

CONCURRENT TASKS

A concurrent task must be performed

simultaneously in different sides of

same physical station

ZONING RESTRICTIONS

Certain tasks can only be

assigned

to specific stations

OBJECTIVE FUNCTION

Minimizing number of stations}

}

}

}

}

}

𝒋∈𝑺

𝒋 𝒙𝒊𝒋 ≤

𝒋∈𝑺𝑹

𝒋 𝒙𝒌𝒋 +

𝒋∈𝑺𝑳

𝒋 + 𝟏 𝒙𝒌𝒋

∀𝒌 ∈ 𝑻 , 𝒊 ∈ 𝑷𝒌

}

Decreasing the number of stations

Decreasing non-valueadded operations

⊆T

⊆

CURRENT SYSTEM IMPROVED SYSTEM IMPROVEMENTS BY NVA OPERATIONS

Decision Variables

xij=1 if task i is assigned to station j, 0 otherwise

yi=1 if station j is used, 0 otherwise

Parameters

C: Cycle Time

ti: Duration of tasks i; i∈ T

LINECURRENT

SYSTEM

MODEL

RESULT

HEURISTIC

RESULT

ALGORITHM

RESULT

IMPROVEMENT

(%)

LINE 6

PRE-ASSEMBLY14 11 11 11 21,43%

LINE 6

ASSEMBLY14 9 9 9 35,71%

LINE 6

TOTAL28 20 20 20 28,57%

LINE 7

PRE-ASSEMBLY13 10 10 10 23,08%

LINE 7

ASSEMBLY18 14 14 14 22,22%

LINE 7

TOTAL31 24 24 24 22,58%

LINE 6 PRE-ASSEMBLY

OLD NEW %

Station 1 81.72 79.90 2.23%

Station 2 79.20 73.30 7.45%

Station 3 43.40 43.40 0.00%

Station 4 76.68 73.60 4.02%

Station 5 60.50 58.60 3.14%

Station 6 70.30 67.60 3.84%

Station 7 68.70 63.10 8.15%

Station 8 64.44 62.60 2.86%

TOTAL 544.94 522.10 4.19%

LINE 6 ASSEMBLY

OLD NEW %

Station 1 78.08 69.90 10.48%

Station 2 75.60 70.20 7.14%

Station 3 84.06 79.20 5.78%

Station 4 86.00 86.00 0.00%

Station 5 85.77 85.77 0.00%

Station 6 34.56 34.00 1.62%

Station 7 51.66 47.00 9.02%

Station 8 65.34 61.38 6.06%

Station 9 64.14 63.50 1.00%

TOTAL 625.21 596.95 4.52%

LINE 7 PRE-ASSEMBLY

OLD NEW %

Station 1 61.92 59.76 3.49%

Station 2 63.18 62.80 0.60%

Station 3 75.24 75.06 0.24%

Station 4 76.86 72.20 6.06%

Station 5 60.48 58.70 2.94%

Station 6 63.00 59.60 5.40%

Station 7 82.84 76.40 7.77%

Station 8 78.12 74.32 4.86%

TOTAL 561.64 538.84 4.06%

LINE 7 ASSEMBLY

OLD NEW %

Station 1 76.10 69.44 8.75%

Station 2 71.82 71.82 0.00%

Station 3 67.32 65.34 2.94%

Station 4 64.98 64.98 0.00%

Station 5 68.58 66.80 2.60%

Station 6 63.90 63.90 0.00%

Station 7 61.20 61.20 0.00%

Station 8 32.50 32.50 0.00%

Station 9 67.86 67.86 0.00%

Station 10 63.40 63.40 0.00%

Station 11 65.52 63.70 2.78%

Station 12 64.55 64.55 0.00%

Station 13 50.76 49.90 1.69%

Station 14 48.60 47.50 2.26%

TOTAL 867.09 852..89 1.64%

0

20

40

60

80

100

S 1 S 2 S 3 S 4 S 5 S 6 S 7 S 8 S 9 S 10 S 11 PU S 13 S 14

Station Time(sec) Idle Time

0

20

40

60

80

100

S 1 S 2 S 3 S 4 S 5 S 6 S 7 S 8 PU S 10 S 11

VA NVA idle

LINE 6 PRE-ASSEMBLY WORKLOAD & IDLE TIMELINE 6 PRE-ASSEMBLY WORKLOAD & IDLE TIME

(sec) (sec)

(sec) (sec)

0

20

40

60

80

100

S 1 S 2 S 3 S 4 S 5 S 6 S 7 S 8 S 9

0

20

40

60

80

100

S 1 S 2 S 3 S 4 S 5 S 6 S 7 S 8 S 9 S 10 S 11 S 12 S 13 S 14

LINE 6 ASSEMBLY WORKLOAD & IDLE TIMELINE 6 ASSEMBLY WORKLOAD & IDLE TIME

LINE 7 PRE-ASSEMBLY WORKLOAD & IDLE TIME LINE 7 PRE-ASSEMBLY WORKLOAD & IDLE TIME

LINE 7 ASSEMBLY WORKLOAD & IDLE TIME LINE 7 ASSEMBLY WORKLOAD & IDLE TIME

S 11S 14

S 14

0

20

40

60

80

100

S 1 S 2 S 3 S 4 S 5 S 6 S 7 S 8 PU S 9 S 10

S 9

0

20

40

60

80

100

S 1 S 2 S 3 S 4 S 5 S 6 S 7 S 8 S 9 S 10 S 11 PU S 12 S 13S 13

36%

IMPROVEMENT

0

20

40

60

80

100

S 1 S 2 S 3 S 4 S 5 S 6 S 7 S 8 S 9 S 10 S 11 S 12 S 13 S 140

20

40

60

80

100

S1 S2 S3 S4 S5 S6 S7 S8 S9 S10 S11 S12 S13 S14 S15 S16 S17 S18S18

LINE 6

TOTAL

LINE 7

TOTAL

M MT

MTM is a method for determination of duration.Main purpose of MTM-UAS:

Standardization of every movements

and process time in operations.

Characteristics of MTM-UAS:• Analyzing by UAS

• No appreciation of tempo

• Hints for time improvement

27 Stations

123 Operations

1,124 MTM Codes

Pre-Assembly Assembly

32 Stations

118 Operations

1,006 MTM Codes

2,130Total

Coding

5 Days MTM Training

Main Operations of UAS:

A Get and Place

P Place

B Operate V Visual Control

Z Motion CyclesK Body Motions

A Approximate

B Loose

C Tight

Changeable Factors of UAS:

Distance Range of UAS:

1 0 cm ≤--≤ 20 cm

2 20 cm <--≤ 50 cm

3 50 cm <--≤ 80 cm

Distance Range

Main Operation

BA 3

Handle Tool, Get Place and Place AsideH

LINE 6 & 7

S18

S 10

S14

76%

24%YES

NO

81%

19%

71%

29%

81%

19%

IMPROVEMENTS

BY STRIKE ZONELINE 6 LINE 7

IMPROVED S. IMPROVED S.

CURRENT S. CURRENT S.

.

Head Count: Blue Collar:927, White Collar:73

Industrial Site: Manisa Organized Industrial District

Products

GENERAL INFORMATIONIncorporation: 1987

Commercial Office: Istanbul

Merged at Whirpool Corpoaration Group: 2014

Consolidated Net Sales: $265Million

Levent Kandiller

Hande Öztop

GOLDEN ZONE

AA

60

A

B

C

D

Not Ergonomic

Ergonomic

Not Ergonomic

1500mm

600mm

43 2 1

Arm Forearm Wrist Hand

STRIKE ZONE

22%

IMPROVEMENT

23%

IMPROVEMENT

21%

IMPROVEMENT

(%)

Machine

400 cm

700 cm

21.43%

35.71%

28.57%

23.08%

22.22%

22.58%

IMPROVEMENT

MATHEMATICAL MODEL

5% 10%

(OPTIMAL)(%)

𝒙𝒊𝒋 𝒙𝒌𝒋 ∀𝒊 ∈ 𝑻𝜶 , α ∈ ℸ, 𝒋 ∈ 𝑺

(single variable)

𝒊∈𝑻𝜶 𝒕𝒊 𝒕𝒌 ∀𝒊 ∈ Tα , α ∈ ℸ

(single duration parameter)

TASK MARRIAGE

Certain tasks must be performed

together at the same station.

}

63%38%

STRIKE ZONE

YESNO

41%

24%

18% 18%

0%0%

10%

20%

30%

40%

50%

AA A B C D

GOLDEN ZONE

Indesit Company.(2003,May 29).

Boysen, N., Fliedner, M., Scholl, A. (2007).

European Journal of Operational Research, 183, pp.674-693.

International MTM Directorate. Introduction to MTM-UAS.Retrieved from the International MTM Directorate website:

“MTM UAS Üniversal Analiz-Sistemi El Kitabı”,2008

http://mtm-international.org/introduction-to-mtm-uas/

Kim, Y.K., Song, W.S., Kim, J.H. (2009).A mathematical model and a genetic algorithm

Computers and Operations Research, 36 (3), pp.853-865.

REFERENCES

1.

5.

2.

4.

3.

ALGORITHM

for two-sided assembly line balancing.

A classification of assembly line balancing problems.

Indesit History,Retrieved from August 10

Changable Factors

INPUTSDECISION SUPPORT SYSTEM

Initial scheduling according to model output.

Improvements on the basis of station respect to side restrictions.

Operations which exceed takt time(86 sec.) are assigned to next station.

FEASIBLE

1st combination

Both operations

2nd combination

Both operations

Whichever gives the minumum number of stations.

Right RightLeft Left

Right Right

Left Left

1.

2.

1.

2.

(To make use of existing idle times)

Successful, continuous and applicable to other factories’ assembly lines

Can be implemented to different industrial areas such as automotive, electronic

Feasible and applicable for firms which aims to be active against to industrial revolution

APPLICABILITY OF DSS

Qualified for adapting to Industry 4.0 revolution easily

OUTPUTS

If two stations are assigned to the same side and summation of

their operations’ duration does not exceed cycle time,

these stations can be combined regarding to

precedence relations.