Unit 2

Management of Conversion System

Chapter 8: Workforce Measurement

Lesson 25 - work measurement.

Learning Objectives After reading this lesson you would be able to understand Performance dimension Role of accuracy Work measurement techniques

What is the need of work measurement? The fundamental purpose of work measurement is to set time standards for a job. Such standards are necessary for four reasons:

1. To schedule work and allocate capacity All scheduling approaches require some estimate of how much time it takes to do the work being scheduled. 2. To motivate the workforce and measuring workers’ performance Measured standards are particularly critical where output based incentive plans are employed. 3. To evaluate existing performance and bid for new contracts Question such as “Can we do it?” and “how are we doing?” presume the existence of standards. 4. To use for benchmarking Benchmarking teams regularly compare work standards in their company with those of similar jobs in other organizations. Work measurement and its resulting work standards have been controversial since Taylor’s time. Much of this criticism has come from unions, which argue that management often sets standards that cannot be

regularly achieved. There is also the argument that workers who find a better way of doing the job get penalized by having a revised rate set. Despite criticisms, work measurement and standards have proved effective. Work Measurement Techniques There are two common techniques for measuring work and setting standards; time study and work sampling. Highly detailed, repetitive work usually calls for time study analysis. When work is infrequent or entails a long cycle time, work sampling is used. Time study A time study is generally made with a stopwatch, either on the spot or by analysing a videotape for the job. The job or task to be studied is separated into measurable parts or elements, and each element is timed individually. Some general rules for breaking down the elements are: 1. Define each work element to be short in duration but long enough so

that it can be timed with a stopwatch and the time can be written down.

2. If the operator works with equipment that runs separately, separate the actions of the operator and of the equipment into different elements.

3. Define any delays by the operator or equipment into separate elements.

After a number of repetitions, the collected times are averaged. (The standard deviation may be computed to give a measure of variance in the performance times.) The averaged times for each element are added, yielding the performance time for the operator. However, to make this operator’s time usable for all workers, a measure of speed or performance rating must be included to “normalize” the job. The application of a rating factor gives what is called normal time. Normal time = observed performance time per unit x Performance rating When an operator is observed for a period of time, the number of units produced during this time, along with the performance rating, gives NT =

ditsproduceNumberofunTimeworked x Performance rating

For example, if an operator performs a task in two minutes and the time-study analyst estimates her to be performing about 20 percent faster than

normal, the operator’s performance rating would be 1.2 or 120 percent of normal. The normal time would be computed as 2 minutes x 1.2 , or 2.4 minutes. Standard time It is derived by adding to normal time allowances for personal needs (such as washroom and tea breaks), unavoidable work delays (such as equipment breakdown or lack of materials), and worker fatigue (physical or mental). Two such equations are - Standard time = Normal time + (Allowances x Normal time) Or, ST = NT + (Allowances x NT) Or, ST = NT (1 + Allowances) -------- (1) And ST =

AllowancesNT

−1 ------------------ (2)

Equation (1) is most often used in practice. If one presumes that allowances should be applied to the total work period, then equation (2) is the correct one. To illustrate, suppose that the normal time to perform a task is one minute and that allowances for personal needs, delays, and fatigue total 15 percent; then by equation (1) ST = 1 (1 +0.15) = 1.15 minutes In an eight-hour day, a worker would produce 8 x 60/1.15, or 417 units. This implies 417 minutes of working and 480 – 417 = 63 minutes for allowances. With equation (2), St = 1/ (1-0.15) = 1.18 minutes. In the same eight-hour day, 8 X 60/1.18 (or 408) units are produced with 408 working minutes and 72 minutes for allowances. Depending on which equation is used, there is a difference of nine minutes in the daily allowance time.

Example 2: If Premabai wants to be confident at 95 percent level that the ratios for disentangling and knitting are within + 2 percent of the real value, what is the total number of observations, which Hema should perform? Taking the knitting and disentangling activities only, to what extent are the time standards precise? Solution: n = 4p (1 – p)/ E2

= 4 (0.25) (0.75) / (0.02)2 where, p = 0.25 for disentangling However, looking at knitting we have the following requirement n = 4 (0.4) (0.6) / (0.02)2 = 2400 where, p = 0.40 for knitting. Therefore, the total number of observations should have been 2400. The number of observations performed by Hema are much lower than those required. The answer to the second part of the question is derived as N = 300 = 4 (0.40) (1 – 0.40) / E2 Therefore, E = √ (4 (0.40) (0.60) / 300) = + 0.0565 = + 5.65% zthe time standard for knitting will b precisely only to that extent, i.e. + 5.65% Making similar calculations for disentangling, E = √ (4 (0.25) (1-0.25)/ 300) = + 0.05 = + 5%

The time standard for disentangling activity will be precise to the extent of + 5%. It may be noted that the above figures reflect the looseness of the time standards more than what they show. The error is relative to the activity’s own fractional ratio. In other words the disentangling ratio is 25 + 5 percent, i.e. 20 to 30 percent. This is a margin of as much as + 20 percent error compared to itself. Work Sampling Work sampling involves observing a portion or sample of the work activity. Then based on the findings in this sample, statements can be made about the activity. The time it takes to make an observation depends on what is being observed. Many times, only a glance is needed to determine the activity, and the majority of studies require only several seconds’ observation. Three primary applications for work sampling are

1. Ratio delay to determine the activity-time percentage for personnel or equipment. For example, management may be interested in the amount of time a machine is running or idle.

2. Performance measurement to develop a performance index for workers. When the amount of work time is related to the quantity of output, a measure of performance is developed. This is useful for periodic performance evaluation

3. Time standards to obtain the standard time for a task. When work sampling is used for this purpose, however, the observer must be experienced because he or she must attach a performance rating to the observations.

The number of observations required in a work-sampling study can be fairly large, ranging from several hundred to several thousand, depending on the activity and desired degree of accuracy. Normally the desired level of accuracy and the desired level of statistical confidence is specified in advance, such as “the observed (estimated) percentage of time spent waiting on customers should be within 3% of the actual value with 95% confidence.” The minimum number of observations necessary to achieve this accuracy and confidence an be computed by the equation (3) Number of observations required = N = [z2p(1-p)]/e2 ---------- (3) Where p is the proportion of observations during which the target activity is performed, e is the maximum absolute error desired, and z = 1.65 for a 90% confidence level, 1.96 for a 95% confidence level, and 2.33 for a 99%

confidence level. Notice that the required sample size is affected by the actual proportion of time, p, that the target activity is performed. Because this is not known in advance, the sample size may be determined iteratively. An initial estimate of the proportion of time spent on the target activity is made, and then a preliminary sample size is derived. In addition to sample size, we need to determine when and how often to observe. Suppose we decide to take 1100 observations over the course of 22 days; this means we will take 50 observations per day. Within each day we do not usually take the observations at uniform increments of time for at least two reasons: (1) There may be some cyclic pattern of work, so that the worker may do one activity for 10 minutes, then change to another for 10 minutes, and then change back to the original activity for 10 minutes. If the observations are timed to be done every 20 minutes, the observer would always see the same activity being done and conclude that 100% of the worker’s time was spent on that activity. (2) A random observation pattern makes it more difficult for workers to mislead the observer. For example, if a worker has deduced that he is being observed every 20 minutes, he may change what he is doing just before he expects an observation, so the observer will see what the worker wants the observer to see. Five steps are involved in making a work-sampling study:

1. Identify the specific activity or activities that are the main purpose for the study. For example, determine the percentage of time that equipment is working, idle, or under repair.

2. Estimate the proportion of time of the activity of interest to the total time. These estimates can be made from the analyst’s knowledge, past data, reliable guesses from others, or a pilot work sampling study.

3. State the desired accuracy in the study results 4. Determine the specific times when each observation is to be made. 5. At two or three intervals during the study period, recompute the

required sample size by using the data collected thus far. Adjust the number of observation if appropriate.

The number of observations to be taken in a work-sampling study is usually divided equally over the study period. Thus, if 500 observations are to be made over a 10-day period, observations are usually scheduled at 500/10, or 50 per day. Each day’s observations are then assigned a specific time by using a random number table. Work sampling compared to time study Work sampling offers several advantages:

1. One observer may conduct several work-sampling studies

simultaneously. 2. The observer need not be a trained analyst unless the purpose of the

study is to determine a time standard. 3. No timing devices are required. 4. Work of a long cycle time may be studied with fewer observer hours. 5. The duration of the study is longer, which minimizes effects of short-

period variations. 6. The study may be temporarily delayed at any time with little effect. 7. Because work sampling needs only instantaneous observations, the

operator has less chance to influence the findings by changing his or her work method.

Example 1. Premabai undertakes knitting of sweaters for various shops. She has several helping hands who, besides knitting also carry out cleaning, disentangling woolen thread, measuring and cutting, seeing and customer contact activities. Hema, an enthusiastic industrial engineer, did an activities sampling (work sampling) study and came up with the following data Activity Number of Observations Knitting 120 Cleaning 40 Disentangling 75 Measuring and Cutting 20 Sewing 20 Speaking to Customers 25 Total number of observations 300 Hema rated the help she had observed at 95 for the disentangling activity and 100 for the knitting activity. If at the end of the four-day(36 work hours) study, Hema found that the helping hand had disentangled 2.3 kg of woolen thread and knitted a two meter length equivalent, what are the standard times for these activities? Take total allowances at 25 per cent. If Premabai gives the work of disentangling woolen thread to a helper for four hours, how much wool should be disentangled? Solution: Assuming that the total number of observations are adequate, the ratios are:-

Disentangling: (75/330) = 0.25 Knitting: (120/300) = 0.40 Which means, (0.25 x 36 h) = 9 working hours have been spent on disentangling with an output of 2.3 kg at a pace of 95 percent rating. Therefore, Normal time = (9 hr x 0.95)/2.4 kg = 3.72 hr per kg Standard time = (Normal time) / (1 – allowances) = (3.72)/ (1 – 0.25) = 4.96 hr per kg In 4 hours, the output should be, as per the standard, (4/4.96)kg = 0.806 kg The ratio for knitting being 0.40 and rating being a 100: Normal time = (36 x 0.40) h/2 meter = 7.2 hr per meter Standard time for knitting = 7.2/(1-0.25) = 9.6 hr per meter Activity A retail store manager wants to use work sampling to study the activities of her workers in the sports equipment department. The manager wants to randomize both the days selected for observation and the times during the days when observations occur. (a) During the next 50 days the manager would like to select 10 days for observation. Use the random number to determine which 10 of the 50 days should be selected for observation. (b) On days when observation is performed, the manager would like to take 12 observations of the department during each two-hour period. Use the random numbers to determine when the observations should occur for one of these periods. Express the randomized observation times in terms of minutes after the beginning of the period. In each part, arrange the observation days and times in increasing order. With that, we have come to the end of today’s discussions. I hope it has been an enriching and satisfying experience. See you around in the next lecture. Take care. Bye.