Perilloux - IE 2400 - Lab #3 (Time Studies, Part 2 Allowances, Performance Ratings, and Learning Curves) PDF

Preview

Summary

LAB REPORT #3...

Description

Lab #3 Time Studies, Part 2 Allowances, Performance Ratings, and Learning Curves Name: Blake Perilloux Course: IE 2400 Instructor: Dr. Isabelina Nahmens and Dr. Roberto Champney Date: 2-26-18

Table 1: Performance Rating using the Westinghouse Rating System Skill: Effort: Conditions: Consistency: Performance Rating: Total +1

0.06 0.1 0.02 0.03 0.21 1.21

1.) In order to determine a performance rating, we used the Westinghouse Rating System. We rated the worker’s skill as “Good” which is a +.06. The worker was not extremely skilled because it was his first time assembling the helicopter, but because the task entailed connecting Legos, it was not a very hard job, and he was a little above average in assembling the helicopter. For effort, we rated him as “Excellent” which is a +.10. Throughout all six assemblies, he stayed focused and showed optimal effort. We gave a +.02 rating, which is “Good,” for conditions. We had good lighting and air conditioning, but the working space was a little crowded with all the computers on the tables. An “Excellent” +.03 rating was given for consistency because throughout all 6 cycles of assembling, he maintained the same focus and worked at a relatively constant pace. Adding these ratings up resulted in a performance rating of .21, which was added to 1 to provide the rating factor of 1.21.

Table 2: PFD (Personal Needs, Fatigue, and Delay) PFD: Personal Needs Basic Fatigue: Special: Total: (Total +1 ):

0.05 0.04 0.04 0.13 1.13

2.) Personal needs and basic fatigue are what make up constant allowances. On average, workers are usually given a 4% - 7% allowance of the total work time for any personal needs and basic fatigue. Daily personal needs may include factors such as going to the bathroom, using the water fountain, making phone calls, etc. We therefore provided a 5% allowance for personal needs, which is the average percentage usually allowed under normal working conditions. Because we were in quite comfortable working conditions with proper lighting, air condition,

chairs, and tables, there was not too much stress or fatigue the worker had to overcome. Therefore, a 4 % allowance was given for any basic fatigue a worker would encounter throughout the entire work day. According to the book, 9% of the total time is usually given for constant allowances, which is what we allowed for. Variable fatigue is defined as T −t × 100 T where T is the 6th observed cycle time and t is the 1st observed cycle time. In our case, the 6th observed cycle time was 364.4 seconds and the first observed cycle time was 547.8 seconds. When calculating, the variable fatigue equaled -50.33%. Because we were still in the learning phase during these 6 cycles, our time actually improved as we continued to build the helicopter, which is why the variable fatigue comes out to be a very high negative number. Because our time improved in these short 6 observations, there was essentially no variable fatigue. As a result, we did not include this negative percentage when calculating our overall PFD allowance in order to refrain from disrupting our data. By adding the 9 % for constant allowance and the given 4% for special allowance, we obtained our 13% PFD allowance which was added to 1 for further calculation of normal time. Table 3: Time Studies Chart (in seconds)

3.) First, the average observed times for each element was taken which is showed under the “OT (xbar)” column. Normal time = (performance rating)

× (observed time), so we

multiplied the average observed element times by the rating factor to get a normal time for each

element. This data is shown under the column, “NT.” Standard time = (normal time)

× (1 +

PFD allowance), so we multiplied the normal times for each element by 1.13 to get a standard time for each element, which is shown under the column, “ST.” By adding all the element standard times, a standard cycle time was found which equals 581.16 seconds indicated by “CT” on the chart. 4.) In order to find a learning percent for the complete task, we first had to calculate n which is defined as

n=

(ln ( 6 thobserved cycle time ) −ln ( first observed cycle time )) ln(number of cycles)

where our 6th observed cycle time was 364.4 seconds, our 1st observed cycle time was 547.8 seconds, and our number of cycles was 6. When plugging these numbers into the formula, n equals -0.227549. This n calculation is then plugged into the learning percent equation which is defined as

Learning percent=

k (2 x)n n kx

where k is the first observed cycle time, x is the number of cycles produced, and n is the exponent representing the slope. The first observed cycle time, k, was 547.8 seconds, x was 6 cycles, and n was equal to -0.227549. When plugging these values into the equation, the learning percent equaled 0.854085.

Table 4: Predicted Times for cycles 1 through 10

#4 continued.) In order to find these predicted cycle times, we used the equation y=k x

n

where k is the first observed cycles time (547.8 seconds), x is the cycle number, and n is -0.227549. When plugging in these numbers to the equation, we were able to find predicted cycle times for cycles 1 through 10 as listed in the table above. 600.0 500.0

Time (s)

400.0 Predicted Exponential (Predicted) Exponential (Predicted) Actual Exponential (Actual )

300.0 200.0 100.0 -

0

2

4

6

8

10

12

14

16

Cycles

Figure 1: Predicted cycle times (1-10) vs. Actual Cycle Times (1-6) This graph displays the predicted times for cycles 1 through 10 in blue as well as the actual times for cycles 1 through 6 in orange that we observed in lab. My predicted time for the 10th cycle was 324.4 seconds, and my average observed time of the actual 6 cycles was 425.04

seconds. The predicted 10th cycle time was less than the average observed time of the actual 6 cycles because in the 10th cycle, you should be getting closer to moving out of the learning curve. Once you observe a worker who is outside the learning curve, you should be able to complete less observations to get an accurate standard time than if you were observing someone less experienced. A very skilled worker has very little fluctuation in their cycle times, and it is therefore easier to get a correct standard time. Because the standard cycle time of 581.16 seconds was found by observing an assembler in the learning phase, the standard time should be revised. We would have to take many more observations to get a more accurate standard time of a skilled worker. The number of cycles needed to be taken to get a standard time with +/- 5% error and 90% confidence is much greater than 6 for each element. Therefore, the standard cycle time would probably be much less if the observations were taken outside of the learning curve where times do not fluctuate.

Discussion: 1.) The average observed cycle time to assemble the entire product was 425.04 seconds. To find the accuracy needed to assemble the helicopter in an average time of 425.04 seconds with 90% confidence, we first found the error using the equation for determining the number of cycles to time:

[

(z)( s ) n= E

]

2

Where z equals 1.65 based on 90% confidence, s is the standard deviation of the 6 observed cycle times, n is the number of cycles needed to calculate a correct standard time, and E is the error allowed. When plugging in 6 for n, 78.13 for s (found from the chart in Lab 2), and 1.65 for z, we get an error equal to +/- 52.629. Dividing this number by 425.04 seconds, the average observed time, gives us an accuracy of +/- 12.38%. Therefore, in order to assemble the helicopter 6 times in an average time of 425.04 seconds with 90% confidence, a +/- 12.38% accuracy is needed.

2.) In setting up a time study, I would ensure I complete the necessary steps before diving into actual time activities. First, I would specify the specific task or job being evaluated as well as the reason for the time study in writing. Laying out specific reasons for conducting a time study helps set a specific goal and allows for better communication with employees. It is important for employees to understand the actual reasoning for the time study and to not have any miscommunications on why a specific time study is being done. Prior to my time study, I would have to ensure that all managers and employees working together in the company are properly engaged in the time-study. They should all know the goals of the time study and the reasons for actively participating in order to get effective results. Obtaining useful data is impossible without

manager, union, and employee buy-in. Before the actual time study, I would also verify the specific worker being analyzed and that he or she is well-trained. It is important to ensure the worker you are observing is properly skilled. Observing a worker who is not properly trained will result in useless standard times that will be much higher than if you were observing someone who is very skilled at what they do. It is also very important to verify the resources and tools used by the workers during the time study in order to ensure most efficient use of these resources. It is crucial to separate the entire operation into individual, small elements that are to be evaluated during the time study. Having elements that take too little or too much time will result in data that is not specific. Having detailed notes on the overall layout of the workplace, the working conditions, and machine parameters are important for providing correct PFD allowances and understanding how to effectively fix an element of work that seems to take too long. Overall, it is very important to complete these necessary steps in order to ensure that the time study runs as smoothly and effectively as possible.

3.) Based on the time study taken in Lab 2, about 37 observations would need to be taken to obtain a proper standard with +/- 5% accuracy and 90% confidence. The number of observations needed to be timed was found using the equation in discussion #1. The average observed cycle time was 425.04 seconds and the z value was 1.65 based on the 90% confidence. The standard deviation of the 6 observed cycle times was 78.13, which was also found in lab 2. Based on this data, we did not perform enough observations in the lab to obtain a good standard time. We only took 6 observations in lab which is much less than 37 observations. During these 6 observations, the assembler was also in the learning phase, so many more observations would need to be taken to ensure a useful standard time.

Time Studies data from Lab 2...