Enhancing the Productivity and Assembly Line Balancing Through Takt Time Implementation

doi:10.21203/rs.3.rs-4275556/v1

Download PDF

Research Article

Enhancing the Productivity and Assembly Line Balancing Through Takt Time Implementation

https://doi.org/10.21203/rs.3.rs-4275556/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

In Pakistan, automotive sector is setting leaps in term of expansion, and achieved a phenomenal growth of 171% between the years 2014 and 2018. Keeping in view, the high demand and production of the sector, it is crucial to generate fruitful revenue from it. For this purpose, the production lines efficiency and productivity should be significantly optimized. Nowadays majority of industries are shifting towards green manufacturing with the application of lean manufacturing tools such as implementation of Takt Time. The proposed study is carried out at Honda Pakistan to reduce the cycle time of KIRA machine through line balancing techniques (Largest Candidate Rule, Longest-Task-Time Heuristic and Mathematical Modelling), capacity calculation, and validated the results via Rockwell Arena simulation software. Moreover, cost benefit analysis is committed to check economic effect of proposed model. Further, after implementation of proposed model, the cycle time of KIRA is reduced from 68.8 sec to 59.08 sec (takt time), and production rate is enhanced from 2800 parts to 3263 parts per day. Lastly, bottleneck is resolved, ultimately leading to increase in line efficiency and station utilization, while on other hand balance delay and line smoothing index are decreased.

takt time

line balancing

cycle time

balance delay

lean manufacturing

The automobile industry is among the most competitive business sectors. Because of the ever-increasing worldwide competitiveness, one of the primary goals of the automobile business is low-cost production [1]. Industries in the automotive sector have looked for different strategies to keep their market share and to assure a prominent position in comparison to the competitors Invalid source specified.. Keeping in view, the high demand and production of the sector, it is crucial to generate prolific revenue from it. For this purpose, the production lines efficiency and productivity should be significantly optimized. In order to achieve these goals different techniques of continuous improvement are applied like required capacity, line balancing, Simulation, Heuristic and mathematical modeling which helps in increasing the production capacity of line, balance of workload among workstations and reduce waste. These are the most commonly used line balancing techniques for balancing product type layout having mass production [2].

Assembly line balancing problems are used to optimize assembly lines, distribute the tasks evenly between workstations, and decrease cycle time. SALBP are further divided into three types with respect to optimization. SALBP I is related to minimize the number of workstations given the cycle time, SALBP II is related to minimize cycle time given the number of workstations, and SALBP E is concerned with line efficiency maximization [3]. This study deals with SALBP II considering multi performance criteria which includes cycle time, workload smoothness and balance-delay optimization. First criteria indicate the minimization of cycle time for bottleneck machine and is reduced to calculated takt time or below takt time in order to maximize the production rate then, moving to second criteria the workload is evenly distributed among the workstations this can be done by using LCR and longest task time heuristic techniques thus setting up the equality of workload among workstations resulting in achieving the minimum level of line smoothing index between workstations. Third criteria aim to reduce the idle time, this can also be addressed by using heuristics and the unutilized capacity is minimized resulting in reducing the balance-delay and increased utilization.

Mathematical model is also developed for line balancing with the objective of minimization of cycle time in first stage and then in second stage number of machines are minimized, and number of workers are minimized in third stage [4]. Arena Simulation software makes it possible to see the real time scenario of production parameter achieved through assembly line balance as it is newly adopted technique by the researchers in related research to evaluate the production performance of a layout. Another crucial performance factor is the cost of adopting a certain line balance solution. The major components of this cost are labor and equipment. Because various line balances will involve a varying number of people and pieces of equipment (where paralleling is feasible), this information must be considered alongside output performance. As a result, while comparing line balancing systems, the number of people and quantity of equipment required will be considered as input performance parameter [5].

This study is carried out in production department of automotive industry on CD-70 cylinder production lines, where daily production goals are not fulfilled. Cylinder daily demand is 3200 units but line is producing 2800 units per day. The proposed study is carried out to reduce wastes and increase value of product through lean manufacturing [6] and line balancing techniques Largest Candidate Rule, Longest-Task-Time, Heuristic [7], mathematical modeling [8], simulation [9], and capacity calculation. Cost benefit analysis is also committed to check economic effect of proposed model.

Problem Statement

Atlas Honda is a Pakistani motorcycle giant since 1962 in Pakistan. It manufactures 8 distinct types of products in different departments. In this study the CD-70 Cylinder production line of the production department is targeted. An average daily production is 2800 units while average daily demand is 3263 units with an average 1.95% Shrinkage which means 14.18% orders are not timely fulfilled. The lead time (L.T) of the production line is on average 319 sec while cycle time (C.T) on average of its bottleneck (KIRA Machine) is 68.8 sec that is almost 21.5% of the lead time. Owing to stated scenario, other machines on the line are underutilized. Proper debottlenecking and line balancing approaches can assist in cycle time adjustment with takt time (T.T) and machine efficiency enhancement.

The lean concept emerged from Japanese industry, particularly Toyota. Lean manufacturing is based on optimizing resource usage through waste reduction [10]. Another concept related to this is lean line balancing. In order to maintain a balance between a worker and workload to achieve required production rate/takt time line balancing technique is used [5]. In terms of the optimization approach, the SALBPs may be classified into three types. Type I simple assembly line balancing problem (SALBP-I) aims to reduce the number of workstations given the cycle time. Type II simple assembly line balancing problem (SALBP-II) aims to reduce cycle time given the number of workstations. The Type E Simple Assembly Line Balancing Problem (SALBP-E) aims to maximize the line's efficiency [3]. To solve SALBP- I numerous heuristics based on a tabu search, branch-and-bound procedure, and genetic algorithms (GAs) have already been suggested in the relevant studies [11]. The basic SALBP-I (type-I) and SALBP-II (type-II) problems have been extensively studied in the literature the key goal of balancing an assembly line is to maximize the effectiveness of the line whichever by SALBP-I or by SALBP-II [12]. In assembly line balancing manufacturers are concerned about the effect of uncertainties on assembly line [13]. In a study to enhance and to identify wastes and provide solutions, analysis of the electronics production process will be conducted. Line balancing index, productivity, and resource effectiveness will be demonstrated to demonstrate the quality of the manufacturing line. As a result, the assembly process, workstation layout, and workstation cycle time should all be examined [14].

Computer simulation is a new method that has captured the attention of certain experts in recent decades for system modelling and simulation. Simulation, according to Pedgen et al. [15], is "the process of constructing a model of a real system and performing tests with this model to understand the behavior of the system and evaluate alternative operating system techniques". Research is conducted in garment industry; the ARENA simulation software is used. The model is authenticated via hypothesis test that contains about eight different system variables. During model validation, five potential 'what-if' situations for assembly line reconfiguration are investigated [9]. Recognition of bottlenecks is a critical step in enlightening the output of a production organization [16]. A process is said to be bottleneck if it reaches its extreme production target but fails to meet a business goal [3]. A new debottlenecking approach is adopted based on principal component analysis (PCA) that can be utilized to recognize and remove sustainability bottlenecks during the planning stage. Researchers can use this context as a tool to quickly identify the significant bottlenecks impacting the efficiency of a specified process, structure, or design [12]. Hong-Sen yan et al, [11] proposes a net-like model for KMS shifting bottleneck identification that is particularly suited to extremely supple and intelligent manufacturing systems. The model is used to explain the idea of a whole production net as well as a way for spotting bottlenecks in such nets.

The mathematical model is used for small and medium-sized problems, and it tries to minimize total amount of exceptional operations whereas taking machine volumes and substitute machines into account [17]. When the problem becomes too complex, such as when trying to minimize cycle time (lessening the cycle time may increase the number of workstations), lessening the number of workstations, reducing workstation idle time, and minimizing workload variance. A genetic algorithm model is projected to efficiently resolve the problem [18]. S.I. Satoglu et al.[17] anticipated a mathematical model for a medium sized line balancing problem in which a sequential methodology of integer mathematical model is developed grounded on AD philosophies for HMS design. It minimizes the inter cell arrival and gives the alternative to satisfy machine demands considering machine capacities. Deckro and Rangachari[19] created a mathematical program design model for the basic assembly line balance problem with the purpose of reducing the number of workstations while seeing cycle time, zoning, idle time, sequencing, and costs.

The heuristic approaches for solving the ALB problem have received the greatest attention and discussion in the literature [20]. The author proposed a unique population heuristic method grounded on the generalized differential evolution (DE) approach. Nearchou [21] investigated the bi-criteria single-model deterministic assembly line balancing issue. First criteria indicate the cycle time of the assembly line and the balance interruption time of the workstations is minimized while second criteria focus on maximization of smoothness index of the assembly line's workload. McMullen et al. [20] applied the heuristic to six distinct line-balancing issues for every stated rule. The designs that develop are simulated, and the performance outcomes are investigated. The study shows, through simulation, how the proposed heuristic can be applied to resolve different line-balancing problems through various techniques, and how these approaches may be assessed transversely several performance parameters.

Cycle time is one of the key characteristics that must be optimized as plentiful as feasible when the production line is attempting to increase productivity, cost base, and customer receptiveness [22]. Cycle time for each station is already fixed by vendor during the production operation in traditional manufacturing systems, but the expansion of smart manufacturing techniques empowers the cycle time of each machine to be made changeable to rapidly respond to marketplace demand fluctuations and to more improve system productivity [23]. Increasing or decreasing the system's cycle times to match takt time will upsurge general efficiency of the system [24]. Takt time is utilized to stride production by linking consumer demand to accessible production time [25]. All feasible cycle times for any scheduling method may be determined using the SPPM (sequential component production matrix). The effective adoption of the takt production model demands a further constructive method, begin with production planning. Furthermore, clarity of production planning is required for successful production controlling and monitoring [26]. Takt-time examination is widely utilized and well recognized in various businesses, such as automotive, equipment and plant manufacturing [27].

Machine, Labor, power supply, Material, warm-up time, production pace, and energy consumption are all elements to consider in cost analysis [8]. The break-even analysis is an effective method for calculating the return on investment. Its purpose is to identify the point, at which investment costs equal to earnings [13]. Break-even analysis often used in those conditions where a Managerial decision need a thorough examination of cost and profit behaviour [2].

For reducing cycle time proper techniques of debottlenecking and line balancing has applied and by this productivity has been increased and for cost reduction a cost analysis is done to check whether the cost is increasing or decreasing by reduction of time and enhancement of productivity and if cost is increasing then 3rd objective needs to fulfill else everything is okay.

Following techniques are implemented.

Required Capacity

Line balancing

Largest Candidate Rule
Heuristic Method
Line Balancing through Linear Programming
Kilbridge Wester Method

Required Capacity

The necessary capacity is determined by the processes allocated to the machine and the required production rate for each item.

Number of Machines Needed

Once the needed capacity and available productive time per machine have been determined, calculating the required number of machines is simple. If the facility operates on one shift at 95 percent efficiency, the available time per machine is 456 minutes (480 0.95). The total capacity required is 1308.72 minutes, which necessitates the use of 2.87 (1308.72/456) or 3 machines. Table 1 summarizes the number of machines requirement.

Table 1

Number of machines required calculation
Operations	Demand	Shrinkage (%)	Total Demand	Processing Time	Loading/ Unloading time	Total Time	Personnel Allowances (Pa)%	Pa Time	Total Time/ Operation	Time Taken for Demand	Available Time	No. of Machines Required	Discrete No Of Machines
HSF	3200	1.95	3263.6	40.8	10.76	51.54	15	0.1	51.6	168447	64260	2.6	3
BSF	3200	1.95	3263.6	44.8	10.42	55.26	15	0.1	55.3	180634	64260	2.8	3
KIRA	3200	1.95	3263.6	70.6	10.32	80.92	15	0.1	81.0	264483	64260	4.1	4
SPM	3200	1.95	3263.6	43.6	10.06	53.68	15	0.1	53.8	175458	64260	2.7	3
MT	3200	1.95	3263.6	35.6	14.22	49.82	15	0.1	49.9	162835	64260	2.5	3
FB	3200	1.95	3263.6	38.0	15.82	53.77	15	0.1	53.9	175749	64260	2.7	3
RH	3200	1.95	3263.6	12.5	20.88	33.38	15	0.1	33.4	109104	64260	1.7	2
FH	3200	1.95	3263.6	14.1	20.17	34.22	15	0.1	34.3	111843	64260	1.7	2
LT	3200	1.95	3263.6	22.6	10.19	32.79	15	0.0	32.8	107177	64260	1.7	2

There are 9 different machines in each line and 3200 parts daily demand, but with 1.95% shrinkage rate demand will be 3263.6 parts per day. According to company policy 85% efficiency is needed while other 15% is supposed as personal allowance, and at the end we total processing time of each machine per part is calculated by adding processing, handling time and the time of personal allowance with the reference of total of processing and handling time. As through known daily demand the total cycle time of each machine is multiplied with daily demand of each machine’s and divided by total available time per day. And answer is required number of machines needed to fulfil demand with given time.

Comparing with real scenario all machines are same in number except KIRA, Rough Honing & Fine Honing machine, in real system there are 3 KIRA, 3 RH & 3 FH while after calculation 4 KIRA, 2 RH & 2 FH are needed, its means 1 KIRA should be added while 1 RH & 1 FH should be removed to fulfil demand on time according to takt time.

Table 2

Machine comparison of existing and proposed model
S. No	Existing System	Recommended System
1	KIRA = 3	KIRA = 4
2	RH = 3	RH = 2
3	FH = 3	FH = 2

Line Balancing

The single-model deterministic straight-type assembly line balancing problem is the modest amongst all the ALB problems. This problem deliberates single model which is to be collected in the assembly line in which the workstations are organized in a straight-line system. The implementation period of every task of the single model (product) that is assembled on the assembly line is deterministic. The machine comparison has been shown in Table 2 and in Table 3 all the tasks’ breakdowns have been depicted. This section presents a detailed review of SM_D_S assembly line balancing problem while referring to;

Mathematical models
Heuristics
Largest Candidate Rule
Kilbridge wester methods
Simulation Method

Table 3

Task breakdown
Operation	Name	task time
1	HSF	13.5
2	BSF	14.8
3	KIRA	17.0
4	SPM	14.8
5	MT	11.7
6	FB	12.5
7	RH	5.2
8	FH	5.2
9	LT	14.2
		108.9

There are some parameters that need to be calculated before and after line balancing to observe the impact of different line balancing techniques.

Takt Time

To calculate takt time we use Eq. 2

$$Takt Time=\frac{Available Time}{Demand}$$

$$Takt Time=\frac{21420}{1087}$$

$$Takt Time=19.7 Sec$$

Min Number of Work Stations

To calculate min no. of workstations we use Eq. 3

$$Min no. of WS=\frac{Sum of Task Time}{Takt Time}$$

$$Min no. of WS=\frac{108.9}{19.7}$$

$$Min no. of WS=5.46=6$$

Line Efficiency

To calculate line efficiency we use Eq. 4

$$Efficiency=\frac{\sum Task Time}{WS*Takt Time}$$

$$Efficiency=\frac{108.9}{9*19.7}$$

$$Efficiency=61.4\%$$

Balance Delay (Idle)

To calculate balance delay we use Eq. 5

$$Balance Delay=(100-Efficiency)\%$$

$$Balance Delay=(100-61.4)\%$$

$$Balance Delay=38.6\%$$

Line Smoothing Index

To calculate line smoothing index we use Eq. 6

$$SI=\sqrt{{\sum _{j=1}^{k}\left(\text{S}\text{m}\text{a}\text{x}- \text{S}\text{j}\right)}^{2} }$$

$$SI=\sqrt{{\sum _{j=1}^{9}\left(17.0- \text{S}\text{j}\right)}^{2} }$$

$SI=$ 18.99050913.

Table 4

Task times
WS	S_j
1	13.5
2	14.8
3	17.0
4	14.8
5	11.7
6	12.5
7	5.2
8	5.2
9	14.2

Table 4 represents the task times for each operation.

Mathematical Model

The work is divided into 9 fundamental jobs and all jobs take 322.5 sec to complete. The cycle period, or how long the work item is accessible at each work station, is 59 sec. Thus, the least number of stations necessary is 322.5/59, or six, and the maximum number of work stations is 9, or the number of jobs involved. The challenge is determining the precise number of workstations, machines and workers required and which duties will be assigned to particular station. The precedence diagram is depicted in Fig. 1.

To tackle these problems, there are 6 types of constraints equations are developed:

Assignment constraint

Takt time constraints,

Precedence constraints.

Workstation constraint

Resource availability constraint

Worker constraint

Also, these all constraints are using for 3 types of objectives depicted below

Minimize Cycle time

Minimize number of machines

Minimize number of workers

Symbols

To write these all constraints and objective functions following symbols are being used

Ws: 1, 2, 9

m: machine, m = 1, 2, ….9

w: worker, w = 1, 2, ….9

i: task, i = 1,2, …9

T: Takt time

N: total number of ws, m & w

Aws: number of workstations

Bm: number of machines

Xis: if task i is allocated to workstation ws = 1, otherwise = 0

Ymws: if machine m is assigned to workstation ws = 1, otherwise = 0

Zwws: 1 if staff w is assigned to workstation s 0 otherwise

ti: time required to assign task i

Ea: initial task in precedence diagram

La: latest activity in precedence diagram

Xaws: if initial task k is assigned to workstation ws = 1, otherwise = 0

Objective Function

There are 3 objectives of this mathematical model depicted below

$$\sum _{i=1 }^{9 }ti------------\left(1\right)$$

$$\sum _{m=1 }^{9 }Ymws-----------\left(2\right)$$

$$\sum _{w=1 }^{9 }Zw.ws------------\left(3\right)$$

The objectives are to minimize (1) the cycle time, (2) the number of machines and (3) the total number of workers in a production line which are 9 for all thrice line.

Constraints

As per objective functions there are three types of constraints

Objective 1 Constraints

$$\sum _{ws=\text{E}\text{i} }^{\text{L}\text{i}}Xiws=1------------\left(4\right)$$

$$\sum _{i=1 }^{9 }ti Xiws\le T------------\left(5\right)$$

$$\sum _{ws=\text{E}\text{a} }^{La}wsXaws\le \sum _{ws=Eb}^{Lb}wsXbws---------\left(6\right) \forall (\text{a}, \text{b})\in \mathbf{P}$$

Constraint (4) is task restraint which guarantees that every task is allocated only one workstation. (5) is a takt time limitation, which confirm that the whole times in each workstation not surpass the provided takt time which is 59 sec. (6) is a precedence relative constraint, which assure that precedence relationship amongst tasks is not disrupted in Fig. 11.

Objective 2 Constraints

$$\sum _{m=1 }^{9 }Ymws\le Bm------------\left(7\right)$$

Constraint (7) is a resource/machine obtainability constraint, which confirm that the total number of resources in workstation is less than or equal to the number of accessible machines.

Objective 3 Constraints

$$\sum _{w=1 }^{9 }Zwws\le 1------------\left(8\right)$$

Constraint (9) is to limit only single worker to be allocated to just one workstation reliant upon his/her services.

Heuristic Method of Line Balancing

The heuristic technique entails drawing a precedence illustration in a specific manner that shows the flexibility offered for shifting tasks across from one column to alternative in order to accomplish the maximum promising balance. When Wester and Kilbridge applied the heuristic approach to TV assembly line problems, they got very good results (with 133 steps). Table 6 depicts the times of each machine and the precedence diagram shown in Fig. 2.

The heuristic method entails the subsequent steps:

Define the work (job).

Break the work down into simple tasks.

List the various steps as shown in Table 5

Table 5

List of various steps
Operation	Name	task time	Predecessor
1	HSF	13.5	----
2	BSF	14.8	----
3	KIRA	17.0	1, 2
4	SPM	14.8	3
5	MT	11.7	4
6	FB	12.5	3
7	RH	5.2	6
8	FH	5.2	7
9	LT	14.2	4, 7
		108.9027

Table 6

Applications of Heuristic Technique
WS	Machines	Task Time	Total Time	Idle Time
1	1	13.5	13.5	6.2
2	2	14.8	14.8	4.9
3	3	17.0	17.0	2.7
4	4	14.8	14.8	4.9
5	5	11.7	11.7	8.0
6	6	12.5	12.5	7.2
6	7	5.2	17.7	2.0
7	8	5.2	5.2	14.5
7	9	14.2	19.4	0.3

Largest Candidate Rule

The Largest Candidate Rule is a typical approach for line balancing that is applied to equally split workload amongst workstations. It enables a constant movement of work in progress (WIP) over the line with slight or no buffering among workstations. Though, bottlenecks rise often as the assembly is hard to balance appropriately. In line design, LCR takes into account the cycle time and precedence linking. The work foundations are allocated to workstations through this technique built on the size of the elements time, Te (work foundations time) standards. Using the supplied formulas, the cycle time, least number of workstations, balance delays, line productivity, and line smoothness directory of an assembly line are computed [13]. Table 7 lists down the times associated with each station and the precedence diagram is shown in Fig. 3.

Table 7

Largest candidate rules technique applications
Longest Operation Time Heuristic Method
	Station	Task Time	Time Left	Task Ready
1	2	14.8	5	1,3
2	3	17.0	3	1, 4. 6
3	4	14.8	5	1, 5, 6
4	1	13.5	6	5, 6
5	6	12.5	7	5, 7
5	7	5.2	2	5, 8
6	5	11.7	8	8, 9
6	8	5.2	3	9
7	9	14.2	5

Kilbridge wester methods

Table 8 represents the various steps taken in the Kilbridge wester method and based on the information; a precedence diagram shown in Fig. 4 is created.

Table 8

Kilbridge wester technique
WS	Time Remaining	Eligible	will Fit	Average Task Time	Remaining Time	Idle Time
1	19.69	1, 2	1, 2	2(14.8)	4.94
1	4.94	1, 3	None		4.94	4.94
2	19.69	1, 3	1	1(13.5)	6.16
2	6.16	3	None		6.16	6.16
3	19.69	3	3	3(17.0)	2.65
3	2.65	4, 6	None		2.65	2.65
4	19.69	4, 6	4, 6	6(12.5)	7.19
	7.19	4, 7	7	7(5.2)	2.00
	2.00	4, 8	None		2.00	2.00
5	19.69	4, 8	4, 8	4(14.8)	4.91
5	4.91	5, 8	None		4.91	4.91
6	19.69	5, 8	5, 8	5(11.7)	8.04
	8.04	8, 9	8	8(5.2)	2.84
	2.84	9	None		2.84	2.84
7	19.69	9	9	9(14.2)	5.49
7	5.49	None	None		5.49	5.49

Artificial bee colony algorithm ABC

Artificial Bee Colony Algorithm is one of the more modern population-based optimization algorithms for resolving the multidimensional optimization problems. It was stated by Karaboga in 2005 [28]. An intellectual behavior of honey bee colony that hunts new food cradles around their squirrel was considered to comprise the algorithm. This colony of artificial bees involves of three groups of bees called employed bees, onlookers and scouts. While a semi colony consists of the employed reproduction bees, the other half comprises the onlookers. There is only one employed bee for each food source. It means that the number of employed bees is equal to the number of food cradles around the garner. The main stages of the algorithm are given by:

ABC algorithm:

• Send the lookouts onto the initial food sources Recurrence Send the employed bees onto the foodstuff sources and define their juice amounts.
• Calculate the possibility value of the cradles with which they are favored by the onlooker bees.
• Stop the misuse process of the sources uncontrolled by the bees.
• Send the lookouts into the hunt area for discovering new food sources, arbitrarily.
• Memorize the best food source found so far. Until (requirements are met)

Simulation Method

Rockwell Arena

Arena is a comprehensible software that is built on the SIMAN simulation linguistic and contains it. Arena offers substitutable patterns of graphical simulation-modeling and investigation elements that may be used to create a variety of simulation representations. Panels are often used to organize modules. One can access a novel set of demonstrating structures and abilities by switching panels. Modules from numerous panels may usually be utilized in the same simulation model.

Simulation Model

A prominent approach for understanding multifaceted systems is simulation modelling. This pattern, in a nutshell, generates a basic depiction of a system under study. The example then gates to experimentation with the system, showed by a predefined set of areas like better system plan, cost examination, design constraint sensitivity, and so on. Investigation involves generating system histories and following system behavior and numbers throughout. Consequently, the illustration created defines system structure, while the histories produced define system behavior [29].

There are two simulation models

The model of prevailing production system

The model of proposed production system

The model of existing production system

In the existing system there are 9 workstations total, the detail of workstations is given in Table 9 and the simulation model is shown in Fig. 5.

Table 9

Details of Rockwell Arena Simulation Software existing model input
S. No	Workstation Workstation	No of Resources	Processing Time Expressions	Handling Time
1	Head Side Facing	3	38.4 + 4.51 * BETA (1.36, 1.44)	10.9
2	Bottom Side Facing	3	42.1 + 4.49 * BETA (1.66, 1.85)	10.0
3	KIRA	3	54 + 19 * BETA (5.65, 1.98)	15.4
4	Special Purpose Machine	3	42.1 + 4.44 * BETA (1.6, 1.56)	10.9
5	Multi Tapping	3	UNIF (32, 38)	11.9
6	Fine Boring	3	10 + 31 * BETA (3.22, 0.518)	13.1
7	Rough Honing	3	7 + 7.9 * BETA (1.21, 1.51)	11.9
8	Fine Honing	3	7 + 7 * BETA (1.45, 1.49)	22.6
9	Leak Tester	2	NORM (28, 4.78)	11.0

Arena Report of Existing Model

Key Performance Indicators

Number of Parts out of Existing Model

Replications: 500

System Average

Number Out 2,814

Schedule Utilization of Existing Model: The scheduled utilization is calculated by dividing the time average number of busy resources by the time average number of scheduled resources. In Fig. 6 the results of existing model are shown below.

The model of proposed production system

In the existing system there are 9 workstations total, the detail of workstations is given in Table 10.

Table 10

Details of Rockwell Arena Simulation Software proposed model input
S. No	Workstation Name	No of Machines/Resources	Processing Time Expressions	Handling Time
1	HSF	3	38.4 + 4.51 * BETA (1.36, 1.44)	10.9
2	BSF	3	42.1 + 4.49 * BETA (1.66, 1.85)	10.0
3	KIRA	4	54 + 19 * BETA (5.65, 1.98)	15.4
4	SPM	3	42.1 + 4.44 * BETA (1.6, 1.56)	10.9
5	MT	3	UNIF (32, 38)	11.9
6	FB	2	10 + 31 * BETA (3.22, 0.518)	13.1
7	RH	2	7 + 7.9 * BETA (1.21, 1.51)	11.9
8	FH	3	7 + 7 * BETA (1.45, 1.49)	22.6
9	LT	2	NORM (28, 4.78)	11.0

Arena Report of Proposed Model

Key Performance Indicators

Number of Parts out of Proposed Model

Replications: 500

System Average

Number Out 3,751

Schedule Utilization of Proposed Model

The scheduled utilization is calculated by dividing the time average number of busy resources by the time average number of scheduled resources, shown in Fig. 7.

Cost analysis by Break-Even Analysis

Proposed Model

Variable Cost of Proposed Model

Cost per labor = $ 155.99 per month

Cost of labor = (24) ($ 155.99) = $ 3743.76

Units Production / shift / month = 900*24 = 21600

Labor Cost per unit = 3743.76/21600 = $ 0.17

From Appendix 1;

Fixed Cost = $ 1,412,092.00

Labor Cost per unit = $ 0.18

Electricity Cost per unit = $ 0.04460

Selling Price per unit = $ 3.57

Revenue = 3.57*3200 = $ 11,424

Table 11 shows all the costs for proposed model while Fig. 8 represents the Breakeven points.

Table 11

All costs for proposed model
Labor Cost	Cost per day/person	Material Cost	Price	V.C	F.C	C.M	BEP x	BEP c
$ 156.25	$ 4,062.50	$ 2.90	$ 3.57	$ 2.90	$ 942,254.50	$ 0.63	1507507	$ 5383952

$$\text{B}\text{r}\text{e}\text{a}\text{k}-\text{e}\text{v}\text{e}\text{n} \text{p}\text{o}\text{i}\text{n}\text{t} \text{i}\text{n} \text{u}\text{n}\text{i}\text{t}\text{s} \left(BEPx\right)=\frac{Total Fixed Cost}{Price-Variable Cost} = \frac{F}{P-v}$$

= 1507507 units

Break-even point in dollars (BEP$) = BEP x P = $\frac{Total Fixed Cost}{\frac{1-variable Cost}{Price}}$ = $\frac{F}{1-\frac{v}{p}}$

= $ 5383952

By changing number of machines, a cost benefit analysis needed to check either company will be in profit or loss and for this purpose a break-even analysis had done and results are given in Table 12

Table 12

Comparison of Break-Even analysis results
	Existing System	Proposed System
BEP ($)	$ 5435022	$ 5383952
BEP (x)	1521806 units	1507507 units

In existing model, we can achieve break even on $5435022 while in proposed model $53,83,952, By applying new model $51,070 can be saved. As per cost benefit analysis t is clear the proposed model is more effective than existing because it saved money and fulfil all requirements.

Also, the existing system was not fulfilling the demand and for this purpose system is being optimized, but after optimization the production lines unbalanced,

Hence, therefore some line balancing techniques were applied to balance and smooth the flow through production lines.

Here in Table 13, some results comparison of every technique used are summarized.

Table 13

Results Comparison
Parameter	Existing System	Heuristic	LCR	KW
Line Efficiency	61.4%	78.99819%	79%	79%
Balance delay	38.6%	21.00181%	21%	21%
Line Smoothing Index	18.9	7.608173	3.23	11.6496242

According to efficiency and Balance delay Largest Candidate Rule, Heuristic & Kelbridge Wester Methods are about same but according to line smoothing index Largest Candidate Rule is optimum.

Manufacturing industries tend to reduce their wastes and increase productivity via reckoned techniques, some of which are Line balancing, Mathematical modeling, Rockwell arena simulation and no. of machine calculation. This study is carried out to reduce cycle time of bottleneck and optimize number of workstations to meet customer demand and fulfil customer satisfaction through high quality. As we know that CD-70 Cylinder is essential part of motor bike and cylinder production dependent upon all the 9 machines through which it is processing, but the KIRA machine is limiting its production which means the single machine is cause of limited production or under production of whole Atlas Honda LTD industry. However, through proper line balancing and capacity calculation techniques this bottleneck is resolved and production met to demand of the industry.

Before implementation of Takt time techniques, the efficiency of production line was 60%. After applying Capacity requirement and line balancing techniques the efficiency of production line improved up to 79%. Therefore, the application of lean manufacturing tools like Takt Time is one of the most effective and less costly ways of enhancing productivity because cost analysis proved that after applying techniques cost is reduced and customer satisfaction is improved.

SALBP Assembly Line Balancing Problem

CT Cycle Time

GA Genetic Algorithm

CD Capacitance Discharge

Availability of data and materials

The data used in this research is a sole and confidential property of Atla Honda Pakistan. We were refrained from using the collected data with any third party outside of the Atlas Honda Pakistan.

Code availability

Not applicable.

Competing interests

There is no competing or conflict of interests among the authors.

Funding

This research did not receive any fundings.

Author’s contribution

Data curation. Analysis, methodology, Abdul Basit Ali. Resource, software, writing – original draft, Muhammad Khalil. Formal analysis, project administration, validation and writing – review and editing, Muhammad Shihab Ud Din and supervision, general guidelines, technical support and visualization, Muhammad Tahir.

Acknowledgements

Not applicable.

Tisza and Czinege, “Comparative study of the application of steels and aluminium in lightweight production of automotive parts.,” International Journal of Lightweight Materials and Manufacture, vol. 1, no. 14, pp. 229–238., 2018.
Bongomin et al., “Improvement of garment assembly line efficiency using line balancing technique.,” Engineering Reports, vol. 2, no. 4, p. e12157, 2020.
Li, Kucukkoc, and Tang,, “A comparative study of exact methods for the simple assembly line balancing problem,” Soft Computing, vol. 24, no. 15, pp. 11459–11475, 2020.
Sivasankaran and Shahabudeen, “Literature review of assembly line balancing problems,” The International Journal of Advanced Manufacturing Technology, vol. 73, no. 9, pp. 1665–1694, 2018.
McMullen and Frazier, “Using Simulation and Data Envelopment Analysis to Compare Assembly Line Balancing Solutions,” Journal of Productivity Analysis, vol. 11, no. 2, pp. 149–168, 1999.
Qattawi and Madathil, “Assembly line design using a hybrid approach of lean manufacturing and balancing models.,” Production & Manufacturing Research, vol. 7, no. 1, pp. 125–142, 2019.
Türkmen, Yesil and Kayar, “Heuristic production line balancing problem solution with MATLAB software programming,” International Journal of Clothing Science and Technology, 2016.
Kharuddin et al., “Line balancing using heuristic procedure and simulation of assembly line,” Indonesian J Elect Eng Comput Sci, vol. 17, no. 2, pp. 774–782, 2020.
Tarigan and Ambarita, “Production layout improvement by using line balancing and Systematic Layout Planning (SLP) at PT. XYZ,” in IOP Conference Series: Materials Science and Engineering. 2018., IOP Publishing, 2018.
Sundar et al., “A review on lean manufacturing implementation techniques,” Procedia Engineering, vol. 97, pp. 1875–1885, 2016.
Kilincci and Bayhan, “A Petri net approach for simple assembly line balancing problems,” The International Journal of Advanced Manufacturing Technology, vol. 30, no. 11, pp. 1165–1173, 2016.
Cao et al., “An Optimization Model for Assembly Line Balancing Problem with Uncertain Cycle Time,” Mathematical Problems in Engineering, pp. 1–13, 2020.
Kucukkoc, et al., “A mathematical model and artificial bee colony algorithm for the lexicographic bottleneck mixed-model assembly line balancing problem,” Journal of Intelligent Manufacturing, vol. 30, no. 8, pp. 2913–2925, 2019.
Lam et al., “Lean Line Balancing for an Electronics Assembly Line,” Procedia CIRP, vol. 40, pp. 437–442, 2016.
Pegden, Sadowski and Shannon, Introduction to simulation using SIMAN, McGraw-Hill, Inc., 1995.
Yan et al., “A new bottleneck detecting approach to productivity improvement of knowledgeable manufacturing system,” Journal of Intelligent Manufacturing, vol. 21, no. 6, pp. 665–680, 2020.
Satoglu Durmusoglu and Ertay, “A mathematical model and a heuristic approach for design of the hybrid manufacturing systems to facilitate one-piece flow,” International Journal of Production Research - INT J PROD RES, vol. 48, pp. 5195–5220, 2020.
Lee et al., “Multi-Objective Assembly Line Balancing Problem with Setup Times Using Fuzzy Goal Programming and Genetic Algorithm,” Symmetry, vol. 13, no. 2, p. 333, 2021.
Gökçen and Agˇpak, “A goal programming approach to simple U-line balancing problem,” European journal of operational research, vol. 171, no. 2, pp. 577–585, 2006.
McMullen and Frazier, “A heuristic for solving mixed-model line balancing problems with stochastic task durations and parallel stations,” International Journal of Production Economics, vol. 51, no. 3, pp. 177–190, 2017.
Nearchou, “Multi-objective balancing of assembly lines by population heuristics,” International Journal of Production Research, vol. 46, no. 8, pp. 2275–2297, 2018.
Taifa and Vhora, “Cycle time reduction for productivity improvement in the manufacturing industry,” Journal of Industrial Engineering and Management Studies, vol. 6, no. 2, pp. 147–164, 2019.
Li, Huang and Chang, “Data-Enabled Real-Time Modeling for Production Systems with Variable Cycle Time,” Procedia Manufacturing, vol. 53, pp. 561–567, 2021.
Ali and Deif, ““Dynamic Lean Assessment for Takt Time Implementation”.,” Procedia CIRP, vol. 17, pp. 577–581, 2014.
Linck and Cochran, The Importance of Takt Time in Manufacturing System Design, 1999.
Keskiniva, Saari and Junnonen, “Takt Planning in Apartment Building Renovation Projects.,” Buildings, vol. 10, no. 12, p. 226, 2010.
Schneider et al., “Operational Excellence in Practice—the Application of a Takt-Time Analysis in Pharmaceutical Manufacturing.,” Journal of Pharmaceutical Innovation, vol. 10, no. 2, pp. 99–108, 2015.
Karaboga, ““An idea based on honey bee swarm for numerical optimization,” - Technical report-tr06,” Computer Engineering Department, Turkey, 2005.
Hammann and Markovitch, “"Introduction to Arena [simulation software],",” in Winter Simulation Conference Proceedings, 1995.

No competing interests reported.

AppendixA.docx

Download PDF

Version 1

posted

You are reading this latest preprint version

Enhancing the Productivity and Assembly Line Balancing Through Takt Time Implementation

Status:

Version 1

Abstract

Figures

INTRODUCTION

Problem Statement

LITERATURE REVIEW

METHODOLOGY

Required Capacity

Number of Machines Needed

Line Balancing

Mathematical Model

Symbols

Objective Function

Constraints

Objective 1 Constraints

Objective 2 Constraints

Objective 3 Constraints

Heuristic Method of Line Balancing

Largest Candidate Rule

Kilbridge wester methods

Artificial bee colony algorithm ABC

Simulation Method

Rockwell Arena

Simulation Model

The model of existing production system

Arena Report of Existing Model

Key Performance Indicators

The model of proposed production system

Arena Report of Proposed Model

Key Performance Indicators

Cost analysis by Break-Even Analysis

Proposed Model

Variable Cost of Proposed Model

Results

Conclusion

Abbreviations

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1