Winco Manufacturing Simulation Report



Winco Manufacturing Simulation Report

Course: Industrial Systems Simulation

Instructor: Carlos Oliveira

Sponsor: Winco Manufacturing, Inc.

]

Team: John Domeier

Brent LeBlanc

Raymond Lee

Jason Spranza

Date Submitted: April 15, 2004

Executive Summary

Our team’s assessment and improvement project at the Winco manufacturing facility will undoubtedly prove to be very beneficial to both the group’s knowledge and experience and to Winco’s production and efficiency. We assessed the treatment table production line and found several problems and areas for improvement. We drew upon our combined teachings and ideas to analyze the situation and to generate solutions to each of the problems. Though many learned techniques we then analyzed the probable outcomes of these changes.

Changes are to be proposed in almost all areas of the table manufacturing line. One of the more drastic changes suggested by our team is the reduction of the length of the line from approximately 130 feet to only about 80 feet. Along with this change will come many others in areas such as layouts, material handling routs, workstation design, and ergonomics, all of which are aimed at the goal of improving production and efficiency numbers on the assembly line.

Through the analysis conducted, we estimate that production times will be significantly reduced and production numbers increased. An increase of as much as 22.72% could be expected in numbers of tables produced in a month, which jumped from 213 to 276. These changes will also hopefully translate to increased savings in costs and increased profit for Winco.

Table of Contents

I. Project Formulation………………………………………….

II. Simulation Model Building…………………………………..

III. Model Verification and Validation…………………………..

IV. Input Data Collection and Analysis………………………….

V. Simulation Experiments………………………………………

VI. Simulation Output Analysis…………………………………..

VII. Conclusions and Recommendations………………………….

VIII. Reference List…………………………………………………

IX. Appendix……………………………………………………….

Project Description

Assembly Line

Winco is a leading manufacturer of medical furniture based in Ocala, Florida. Plant manager,

Jim Ankoviak, assigned us to the assembly line for the model-856 table (Treatment Table with

Adjustable Back). This particular model is one of Winco's most popular products. The 856-

model table, along with all other Winco products, is hand assembled.

Current Layout

On our first visit to the Winco facility our team was surprised at a number of the observations we were able to make. Some of the processes we saw we looked on favorably, like the impressiveness of the amount of activities going on in this relatively small manufacturing facility. There were also many possibilities for improvements. At first glance of the production floor it was hard to observe the entire process due to its complexity. We were given a brief tour of the whole facility by Mr. Ankoviak then shown to the table assembly line where we were to focus for upgrade implementation. We were taken down the assembly line station by station, and had the production procedure explained.

Initial Upholstery Area

At this first station we spotted a very long, almost continuous roller line right away. It stretched well over 100 feet down facility to the shipping area and even 40 feet down to the next station. On the sides of the roller line were laid out materials, tables, shelves, racks and bins, all of which served to create an isle on each side of the rollers. At the end of the line was a suction lifting crane and wooden table backs. Behind the line were large stacks of varying size foam and boxes. On one side of the line were storage racks and bins, and a table and shelf for hardware and documentation storage.

Upholstery Press Area

The first thing we noticed at the upholstery pressing station was the metal framed pressing machine with the roller conveyer underneath used to push the tables down and hold them tight while workers stretch out and attach the vinyl upholstery. Also, air hoses were hanging down here for the air tools that were shared at this station and at the previous one. Under the rollers at the press there was a platform which was being used to store some hand tools, unused air tools, and other various things.

Table Assembly Area

Most noticeable at this station were the large deep storage bins on one side of the line, and the cluttered hardware storage shelf and table on the other. This assembly area was mostly just a long continuous stretch of rollers with metal rail and leg storage underneath. There were hanging air hoses in this area also, with the actual air tools placed in seemingly random spots in the station. Also, on the far side of the line at this station were large stacks of unformed cardboard boxes.

Packaging and Shipping Area

This area was also simply a expanse of rollers with rail storage underneath. This station was pointed out as being where the product was boxed and prepped for shipping. The area was mostly open except for a short table on the other side of the line where various pieces of sorted hardware and instructions sat. This, we were told, was the area where self assembly hardware packs were bagged and included with the final product.

Problems

Teamwork

One of the biggest problems that our sponsor faces each day is the teamwork within the

company. Many of the employees are not willing to try to work together with their assemblies

and this drastically decreases their productivity. With our simulation, we will be looking at how

much productivity will increase with our suggestions and how the number of workers required

doing a certain task affects the productivity.

Inventory Placement

After looking at the facility, we also are looking at how much we can improve the manufacturing

process by changing where the company stores the inventory. Currently, the inventory is stored

under the production line and the workers must bend under the line and pick it up. Also, the

current process does not stock the inventory regularly, causing the employees to stop production

in the middle of a shift and restock.

Standard Assembly Techniques

After decreasing the time of the process by increasing teamwork and changing the location of the

inventory, it may be in the best interest of the company to regulate the techniques used during

assembly. Currently, each employee uses their own personal technique, and this may not be the

best possible one to use. The company can easily standardize these techniques which will also

allow new employees better training.

Tools

There was a great deal of time lost in the system due to the workers not having the correct tools.

The manager of the company estimated that approximately 40% of each shift was “wasted” on

non-value add tasks, such as changing a tool to a different air hose or walking to another part of

the line to get screws, nuts, or tools.

Planned Solutions

One of the biggest advantages of using Arena to model our manufacturing facility is the easy

way that many different layouts and work methods can be explored. The problems that our

sponsor faces can be explored using simulation by creating many different scenarios, and

comparing the results.

For example, the issue involving teamwork that our sponsor faces can be studied by creating two

simulations that vary by the number of workers that are available to do each task. The placement

of inventory can also be addressed by creating many different material layouts, with each

variation being analyzed. Our team’s goal is to address each of our sponsor’s problems using

simulation.

Proposed Layout

We observed the treatment table production line and found several areas for improvement. We

proposed these improvements to analyze the situation and to generate solutions to each of the

problems. Changes are to be proposed in almost all areas of the table manufacturing line. One

major change suggested by our team is the reduction of the length of the line from approximately

130 feet to only about 80 feet. Along with this change will come many others in areas such as

layouts, material handling routes, workstation design, and ergonomics, all of which are aimed at

the goal of improving production and efficiency numbers on the assembly line.

Original Layout

[pic]

Simulation Model Building

Simulation of Assembly Line

The purpose of this project is to use the ARENA simulation software package to build an

accurate model of the Model #856 table. We will model the simulation using data obtained for

the line balancing problem as well as new data still to be obtained. The main goal is to observe

how the changes we proposed in the line balancing problem will affect the assembly line.

Importance of Project

Using ARENA to simulate the proposed assembly line will better show how the line will be

affected by or proposals. Initially we stated there would be a reduction in chair production time

by 8.44 minutes. This conclusion was a rough estimate based on all the data we had available. By

modeling the production in ARENA we can take into account more real-life variability and see a

more accurate representation of our proposals. ARENA will also allow us to experiment with

different line arrangements and see what other changes we could make that have not yet been

proposed.

Initial Process Modeling

The following are some of the entities and processes we used in the simulation:

• Create Module: This module was used to create the entities which modeled the

initial parts of the table

• Process Module: This module was used to simulate each process in the assembly of the #856 table

• Assign Module: This module was used to assign the properties of each entity, such as process times

• Statistics: This is where we specified what data we wanted to keep track of, such as queue lengths, average time in system, and total number of units produced per shift

• Resources: This is where the different resources of our model were defined: for this simulation the resources were the workers on the line

• Dispose module: This was used to remove the finished entities (tables) from the system

Possible Simplifications

Due to the large variance in the manufacturing times of the tables, there was a need to

simplify the number of operations that we are including in our simulation in order to fit certain

times into an accurate distribution. In order to reduce variance in our entire system, we

blocked a large number of operations together. The idea of the preceding is that the total

completion time of all of the tables contains less variance in production time than each of the

individual operations. This was systemic from the methods of production involved in table

production. Individual workers may have different techniques performing certain operations, but

on average with the new proposed layout, the total table production will stay constant.

Expected Findings

By simulating the assembly line system, we hope to solidify the projected time savings of the

initial project. Also by modeling this system, our team hopes to find new inefficiencies in the

assembly line that can be explored further in our report. The model may need to be scaled down

based on the variance of the data that we take. Data with little variance is necessary so we will be

able to have the data fit a given probability distribution.

Detailed Modeling Procedures

The following is a detailed description of the modeling process. We will also describe how the

assembly line processes, resources, and entities will be replicated in the simulation.

The assembly line we are modeling is linear with most operations being performed by hand. This

eliminated the need to factor machine downtime into our simulation.

Schedule: Workers at the WINCO factory work a 6:30 a.m. to 3:30pm shift, with a 1/2 hour break

for lunch at 11:30am and a 10 min break at 9:00am and 1:30pm

Create: Entity arrivals (tables) were based on a schedule and forecasted demand (our sponsor

gave this information to us). The entities arrived one at a time (no batching).

Resources: Our sponsor wanted us to model the assembly line with 3, 4 and 5 workers. The actual number of workers can vary from 3 and 8. Most important to him, however, was balancing the line with 4 workers. This was the most common configuration for this particular assembly line. The resource schedule is described above.

Processes: There are roughly 55 separate processes performed by the workers to complete each model 856 table. We logically grouped these processes into about 10-12 steps. This greatly simplified our modeling procedure.

Station: Each set of about 5 processes were performed at a different workstation. Each workstation was manned by a different worker.

Transportation: The entire assembly line is made up 3 separate roller conveyers. The distance between each workstation is spanned by a piece of the conveyor. The time it takes to get from station to station is modeled in the transportation spreadsheet.

Statistics: The statistics spreadsheet contains all of the user defined variables and attributes we needed to statistically analyze. This included queue lengths, value added time per unit, total cycle time, resource utilization, average number of tables produced per shift, and the time each unit spends at each workstation. This last variable allowed us to better identify bottlenecks.

Record: This in conjunction with our statistics spreadsheet helped us keep track of all the variables we needed to present accurate data analysis to our sponsor.

Model Verification and Validation

Our simulation project entailed the inputs of real world factors to facilitate this system flow. It was mainly based on the work schedule of employees, including breaks, along with the number of workers operating at each station. We tested our correctness of our data by calculating how long it would take to complete one 856 table, which included processing and traveling times. We then used this value and checked our correctness by using the input analyzer to fit our time study values of each operation and traveling times into a distribution. This allowed us to check the correctness of our data.

Input Data Collection and Analysis

On March 15, 2004 our simulation team took time studies on the 856 table production line at

Winco. The line was broken down into different processes. Each block of color in the tables

above represents a station in the production line. The stations are then broken down into detail

of steps during that process. Data was recorded from two runs of each process having either an

operation or retrieval time (in seconds). We then averaged each process and delay time to come

up with a value to be used for our input analyzer.

A video camera was used to record the processes. Members of the team went back to the video

footage to record the times of each step in the assembly line. After compiling the data, we were

considering fitting this data with either an exponential or triangular distribution. Using the input

analyzer ,we fit this data to determine which distribution gave us a better fit.

Another important thing to note from the table is the number of workers in each process. There

are no more than 2 employees working together at the same time. We observed that some

employees wait around for others to finish before starting their process. Improving this queue

time can produce a larger output. Adding or subtracting workers from each process can also

make the model reach its optimal output.

Simulation Experiments

The assembly is based on the number of orders that come in each day. Out team objective was to

increase the output of the 856 table. This was done by analyzing the number of additional

workers involved at each station in our simulation model. We calculated the number of runs we

needed in our model by using the half width of the total assembly time. We used the equation

n= (n0*h0^2) /h^2. The parameters were the following: initial replications were n0=10

replications, half-width= , and 95% confidence interval we wanted to change it to was. The run

lengths were determined by the operating hours of the manufacturing facility, which was for 8 ½

hours Monday- Thursday, and 6 hours on Friday.

Simulation Output Analysis

Conclusions and Recommendations

Balancing Assembly Line

The first step to improving the current production was to balance the assembly line. Balancing an assembly line primarily consists of assigning equal work to all workstations. The basic principle is to make sure each workstation takes the same amount of time to finish their work. The maximum value this time can have, and still reach production goals is called the cycle time. For this facility we did not have a particular production goal so the current cycle time was used as a baseline value. The cycle time for our proposed layout is slightly less.

For our application, the line is balanced by assigning processes to a specific workstation and adding and subtracting workers to each of these workstations. In other words, for all applicable situations we assume operation times have an inverse relationship to the number of employees working on that operation. For example, if an operation takes 4 seconds for one worker it will only take 2 seconds for 2 workers.

The current line has 53 operations completed in four workstations with times ranging from 7.56 minutes to 9.11 minutes (Appendix II). This large variation in times can potentially cause bottlenecks and idle time, both of which reduce the efficiency of the line. The total time it takes to complete the operations on the table (worker movement times not included) is currently 33.28 minutes. The production is completed entirely in series. This means that one operation is started only when all the previous steps are finished.

The proposed balanced line has the same 53 operations; however, 51 of them are done in series, while 2 of them are done parallel to the others. This reduces the operation time from 33.28 minutes to 26.54 minutes (Appendix II). Make assembly pack, and assemble box top are the two parallel operations. Both of these actions can be performed off the line. This ensures they do not disrupt the flow along the line. The proposed line has workstation times ranging only from 6.54 minutes to 6.70 minutes. The numbers of workers used originally and in the new line are shown in Appendix II.

Standardization

Another major issue is standardization of the production line. Each workstation should have a clear definition of tasks, along with the number of workers assigned to each task.

Currently, there is variation in number and order of operations during production, depending on the worker. This is a major source of the disparity of times between workstations.

Most operations on the proposed line require two workers. This reduces much, but not all worker idle time. During operations where only one worker is sufficient, the idle worker should perform preparation or maintenance tasks. This will keep the flow of the line moving efficiently.

Material Handling

Material handling was one of the major issues concerning our new proposed layout. We compiled the traveling time for each operation by taking time studies for each run. After analyzing video footage from each run, data showed that there were a total of 34 different traveling times. The layout of the current production line had an average of 3.84 minutes of traveling time. To reduce these excess trips to gather materials, we cut down their travel times by minimizing the distances from material storage areas to the assembly line.

Our new proposed layout for our facility has significantly reduced the total average time for the material handling routes. We have placed certain materials adjacent to the assembly line to reduce turning distances for the employees. Examples would include moving the foam and cardboard boxes to the assembly line. We subtracted one second from the traveling times of these operations to account for the difference of turning. We have also eliminated the traveling distances to retrieve screws by placing bins on the assembly line. Rather than retrieving necessary equipment for an operation, we proposed that there should be multiple sets of blocks and staple guns to be placed in slots beside the bins on next to the assembly line. We also accounted one second for each traveling time for the new improvements. During certain operations of the production line, employees make use of tools such as utility knives, wrenches, and screwdrivers. They constantly have to travel back and forth to the shelves to pick up these tools. We proposed having employees carry a light tool belt around their waist to eliminate these movements to and from the shelves.

References

1. Jim Anchoviak, Plant Manager,

5516 S.W.First Lane

 Ocala FL 34474-9307 1 904 854-2929

Appendix I

|Work Schedule |  |  |

|  |  | |  |

|  |Mon-Thurs |Fri | |

|Operating Hours |6:00- 3:30 |6:00-2:00 | |

|# Hours/day |8 |6 | |

|Days |4 |1 | |

| Overtime | 40 minutes |70 minutes | |

|Breaks |2(10 Minute) |2(10 Minute) | |

| | | | |

| | | | |

| | | | |

| | | | |

| | | | |

| | | | |

| | | | |

| | | | |

Description: This table depicts the total time savings for the proposed over the original layout.

|Appendix II (Page 1 of | | | |

|2) | | | |

|Step |Operation |Time(sec) |Total Station Time(Min) |

|Step |Operation |Time(sec) |Total Station |Workers | |

| | | |Time(Min) | | |

|Operation |Description |Run 1 |Run 2 |Avg time | | |

|Description: This table describes the individual operations that have traveling times. |

|There are a total of two runs and an average time for each operation. | |

Appendix V

[pic]

Description: Original Table Production Process

................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download