代做Principles of Operations Management代写留学生Matlab程序

Principles of Operations Management

Coursework: Understanding Bottlenecks

For this assignment, you are going to use a MS Excel spreadsheet to simulate a series of sequential processes in an industrial setting. The structure is similar to the one developed during our tutorials, where we simulated the matchsticks exercise from Goldratt’s book.

You will work on the case of a company in the miniature manufacturing sector. More specifically, they produce die-cast scale models of commercial airplanes and sportscars for collectors.

Die casting is a metal casting process in which molten metal is forced into a steel mould under high pressure. The steel moulds, known as dies, are designed to produce castings with intricate shapes in a manner that insures both accuracy and repeatability. In the miniature industry, the castings are then painted, welded (or glued), and finally polished to provide an accurate scale-representation of the original object.

In  a  simplified  view,  the  process  consists  of  the  following  stages:  order  preparation,  where  the different materials (led, zinc alloy, plastic, rubber, etc.) are prepared and sorted for processing. These materials are then taken to the Casting workshop, where the different parts of the miniature are moulded. The finished pars are then taken to the Painting station, and from there to the Assembly workshop, where the different parts and components are put together, welded, andor glued to build the scale models. The Finishing station takes care of the inspection, polishing, deburring, washing, and finally varnishing of the finished replicas. Finally, all the miniatures are taken to the Sorting and Boxing desk, where orders are sorted together and put in boxes ready for distribution.

In one hour of work, the Order Setup team can get ready materials for up to 250 units. This is a very simple procedure, with a standard deviation of only 16 units per hour. The next station, Casting, can only process 100 units per hour, but due to the nature of the process and the high temperatures required, it has a higher variability, with a standard deviation of 15 units per hour. The capacity of the Painting station is in average 175 units, with a minimum of 100 (for more detailed miniatures) and a maximum of 250 (for models demanding less detail). The Assembly unit also works relatively fast and are capable of processing up to 200 units per hour, with a standard deviation of 20 units per hour. The Finishing station oversees the quality control of the company and, therefore, tend to be slightly faster than the preceding units. They can process, in average, 250 units, with a standard deviation of 10. Finally, the Sorting  and  Boxing  team  can  process  up  to  300  units  per  hour,  also  with  an  almost negligible standard deviation of 15 miniatures per hour.

We  assume  that  all  the   indicated   capacities,  apart  from  the   painting  station,  follow   uniform distributions with the indicated parameters (mean, standard deviation, maximum and/or minimum). The  processing  times  at  the   painting  station  follows   a   normal  distribution  with  the   indicated parameters.

The factory works two uninterrupted 8-hour shifts per day. Breaks and other needs of the operators are already considered in the productivity data.

Part A

You will use random numbers to run a simulation of the 6-station manufacturing process described above, in the same way that you ran the simulation of the game described by Goldratt in chapter 14 of  The Goal. You are expected to simulate two 8-hour shifts  in  a working day  (16  hours)  and to compute and observe the average throughput at each station. Assume there are no set up times between shifts.

For each intermediate processing unit, other than Order Setup, you will have the following columns: “WIP Inventory”, “Productivity”, and “Throughput” (WIP is the acronym for Work in Process). For the Order Setup process, you will only have “Productivity” and “Throughput” . Notice that, when using the metaphor  of  the   boy  scouts,  the  “WIP   Inventory”  column   is  equivalent  to  the  “bowl”  of  the corresponding player in the classroom case.

To   simulate  the   outcome   or  “Productivity”   using   a   normal   distribution,   use   the   command =CEILING(NORM.INV(rand(),Mean,StDev),1).   When    using   the    input   [rand()],   the command  NORM.INV generates  a  random  observation  of  a  normal  random  variable  with  the indicated mean and standard deviation. As we are interested in getting a whole (integer) number instead of a real number (we are talking about production units), you can use the command CEILING, which rounds any rational number up to the closest integer. A uniform distribution can be simulated using =RANDBETWEEN(Min,Max).

Your table will have 16 rows. At the bottom of your table, compute the average throughput and the  average WIP Inventory for each station, together with the corresponding standard deviation (these  will be stored in two additional rows). Two sums are of particular importance, the total “Throughput” of the Setup station, which represents the Total Use of Materials and the total “Throughput” of the  Sorting unit, which represents the Total Output of the system. Compute these two values in a different  row. Finally highlight the “end of the day” WIP inventory (the last cell in the WIP column of each  station).

Repeat the experiment 500 times. Include in your table the following columns: Total Use of Materials; Average Throughput, Throughput Standard Deviation, Average  WIP of each station, and the Total Output. Include an additional column for the Total Waste, which is defined as the difference between the Total Use of Materials and the Total Output. Record at the bottom of the simulation table the average of all the columns. Your analysis will be based on these results.

QA1. Is there a station that accumulates significantly more inventory than the others? Con you suggest away to reduce this inventory without incurring in any investment in equipment or personnel?

QA2. Assume that the company has a weekly demand of 20,000 units, are they able to satisfy their demand every week if they work 6 days per week? Are they facing delays? What would you suggest?

Deliverables for Part A

1.           One  sheet  in your  MS Excel file, containing the table of one working day and the 500 simulations.

2.           In your written report, you should include detailed answers to questions in QA1 and QA2.

Part B

Build another table representing a working day (16 hours) and adjust the “ Productivity” of the Setup station, in this case we will replace the randomly generated number with a fixed value.

Introduce a fixed value in the first cell of the “Productivity” of the Setup Station (this may be cell B2). We  refer to this as the reference value.  Replace every other cell of this column  (32 cells) with a reference to that cell.

Build now a table for a Two-Way Simulation that will be filled-in using the  What-If function in Excel. Your table must have one column for the simulation round number, and one more column for each of the values you will give to the “ reference value” . Starting from the minimum production possible in this station, introduce  new columns, each of them will will  have a value equal to the one in the previous column plus one standard deviation. Stop when you reach 300. As before, the table should have 500 rows.

1.    In  an  empty  cell  in  the simulation table,  include  a  reference to the  cell where the  Total Production value is stored. This is indicated in orange in the figure below (the reference may be different in your table).

2.    Open the What-If-Analysis dialog window. In the row input cell enter the location of the fixed value that we defined for the “target variable” (in my case B2). In the column input choose any empty cell nearby the table.

3.    Once you press OK, you will have a table of 500 simulations for each possible value of the "target variable". At the bottom of the table compute all the column averages.

Identify which is the best throughput rate for the Setup station. Remember that you should be able to satisfy the demand.

Let us now analyse the waste. All the materials that are prepared at the Setup station that are not used at the end of the day, goto waste. Build a new simulation table, the best position is to the right of the last one, leaving a couple of empty columns.

Repeat all the steps above, but now we will analyse the total waste of the company. To do so, the formula that appears in blue in the table above should now point to the cell where the Total Waste is stored. Make sure you include the right reference.

Identify the throughput rate for the Setup station that minimises waste.

QB. Which is the best fixed value for the productivity of the Sorting station? What happens to the Total production and Total Waste values as you increase the reference value? Discuss the trade-off between  product  and  waste.  Is  it  eliminating  waste  completely  a  feasible  target?   Will  it  be economically viable for the company? What would you suggest to the manager?

Deliverables for Part B

1.           One  sheet  in  your  MS  Excel  file,  containing  the  table  of  one  working  day,  and  the simulation tables for the two cases described above.

2.           In your written report, you should include detailed answers to questions in QB.

Part C

Let us go back to our initial setting, the one described in Part A. In that section you were asked to suggest a solution that did not involve any investment. Now, consider that it may be worth investing in increasing the production capacity of certain process, either by expanding its capabilities, or by introducing some parallel processing.

To do this in a more formal way, we first must identify the system’s bottleneck(s). This can be done by finding the processor processes that cause more accumulation of WIP Inventory.

Consider the earliest process identified in the previous paragraph.  After some market research, our Operations team has found two possible improvements:

a)    Introduce parallel processing in that station, buying a new machine of the same characteristics of the first one; or

b)   Consider a 100% increase the capacity of the current machine, with the disadvantage that this increases the process variability. This is obtained by multiplying the station’soriginal average output times 2 and its standard deviation times 3.

QC1. Test these two alternatives independently based on the initial configuration. Compute Total Output and Total Waste using the What-If tool as we did in Part A. You will need one sheet for each of the two alternatives, (a) and (b). Is there an alternative that clearly outperforms the other? Can you identify a new bottleneck?

After applying the best solution, you may notice that there is now another station that is accumulating large amounts of inventory (always moving rightwards from the current station). Which station is it? The company has the following possibilities:

a)   Allow extra time in this station. In particular, the proposal is to pay volunteers for 4 hours more of work. This can be simulated by adding 4 more rows in the workday simulation, leaving the rows and columns corresponding to other stations blank, and allowing the bottleneck and subsequent stations to continue processing during the 4 additional hours (rows). Be careful and think thoroughly the formulas you need to use.

b)   To semi-automatise some of the processes in the station, increasing the average output of the station by 50%, and reducing its standard deviation by 30%.

QC2. Test these two alternatives independently based on the best configuration reported in QC1. Use a spreadsheet for each alternative. In all cases, compute Total Output and Total Waste using the What-If tool. Is there an alternative that clearly outperforms the other?

Can you identify a new bottleneck?

If you identify a new bottleneck. Suggest an improvement of your own inspiration (this can be a variant of the solutions proposed in this section) . Remember that the analysis may be conducted after you have implemented the best options in QC1 and QC2.

QC3. Test your suggested solution based on the best configuration reported in QC2. Compute Total Output and Total Waste using the What-If tool. Is your proposal good? Did you manage to reduce inventory accumulation and waste? If not, what you think is the reason it did not work?

Deliverables for Part C

1.           As many sheets in your MS Excel file for the different alternatives that you tested to find the best set combination of processes for the bottlenecks.

2.           In your written report, you should include detailed answers to questions in QC1 to QC3.

Part D

Assume that a temporary (seasonal) increase in demand is forecasted for the next three months. The expected increase is about 20% over the average weekly demand of 20,000 units.

QD. As the operations manager of this company, what would you do to adjust your plant’scapacity to guarantee fulfilment of the increased demand?

REPORT

You are expected to prepare a 1500 words report based in your simulation experiments. Your report must include an introduction where you explain the aims of the experiment and describe what a bottleneck is, a methodology section with a short description of the simulation experiment and its aims, and main section where you address the questions in this document. Notice that the answer to the questions should not be provided in a question/answer manner but, instead, embedded in the document in a single narrative where you discuss your findings and the implications for the company. If  necessary,  use  textbooks  and  other  academic  literature  to  support  your  claims  or  to  provide theoretical grounds to your assertions.

Your submission will consist of one MS Excel File and one Report (MS Word or PDF).