4.

REVELANT THEORY In order to obtain a complete and an accurate picture of the productive time and idle time of the workers in a specific production area it will be necessary to observe continuously all the workers in the area. This method will require a lot of man-hours do and will cost a lot of money to implement. All the time study methods has being compared and found out that the best method to be used is activity sampling. This method was chosen in the sense that: It less costly to use than other time studies techniques. Simultaneous work sampling study of several operators or workers can be made by a single observer resulting in the decrease of study cost. Work sampling measurement may be made with a pre-assigned degree of reliability. The workers are not observed for the long period so the system would not be interrupted. Heiland and Richard have developed a rather clear-cut definition of work sampling: Work sampling study consists of large no of observations take at random interval, in taking the observation the state or condition of the objective of the study is noted, and this state is classified into predetermined categories of activity pertinent to the particular work situation. From the proportions of observations in each category, interference is drawn concerning the total work activity under study. Since 1934 when L.H.C Tippet developed this method it has being used throughout the world from generations to generations. A lot of Industrial engineers recommend this method as one the best method because it is based on the laws of probabilities.

5. FACTORS INFLUENCING THE WORK SAMPLING 1. The activities to be studied: are actually determined by the purpose of the study. What does management what to know?

2. The representative period: Investigation should cover a representative period of time to ensure that all normal factors which may influence the work being investigated will have a chance to occur. 3. The confidence level: can be defined as the probability that the interval estimated contains the population parameter. It can also be father explained as the degree of probability that the result of an investigation will be a true reflection of the fact. Most Industrial engineers prefers to use a confidence level of 95%. 4. Accuracy: The purpose of the study will determine the degree of accuracy. If management require high accuracy, large number of observation will be needed therefore the cost of implementing to study will be expensive. The accuracy is a means of limit of error allowed .Limit of error can be defined as the greatest possible distance between the point estimate and the value of the parameter it is estimated. 5. The number of observation: to be made will depend on the confidence level, the limit of error and the percentage of occurrence of the activity .This can be obtained by using the formulae on the appendix. 6. Stratification: it the means of ensuring that the observation will cover all the levels of the population. 7. The observation times: should the times which the observation are to be made be selected according to the preference or system, the result would obviously be influenced by that preference or system and the reliability of the results would decrease .It is therefore essential that all the moments should have equal chance without bias to be included in the sample of observation times. 6. Application 6.1 Procedure followed Determine the percentage of occurrence (P) of the activities to being measured. P=X/N Determine the sample size (N) for a confidence level of 95% for any desired accuracy N=
4 P(100 P) L2

Determine the limit of error 2

L= 2x

P (100 P ) (for 2 ) N

alternative formula: L= Sa X P%

6.2 Preliminary Study It is essential to make a preliminary estimate of the percentage occurrence of the delay to be measured. After a day of preliminary study the following percentages of occurrence were obtained: Piling-15% Cooking-33% Packing-27% Non-productive work-13% Others-12%
4 P(100 P) 4 * 33(100 33) 354 Observatio ns = L2 52

Therefore number of observations can be calculated by using the formula: N=

6.3 Random Observations In work sampling the time of observation for any given sample must be selected without bias. For that to be accomplished the calculator can be used to obtain the random numbers. Because the cafeteria is such a busy place during the day the management can not avoid to loss time. The management allocated a 1hour lunch break for the worker in the main shop from 12h00 to 13h00.By the time they go to lunch the worker from the coffee shop closes and work at the main shop for that hour. The manager has advised the Industrial engineer that it would be unnecessary to the work between 12h00 and 13h00 because few customers come. The length of the job being studied is 13hrs (8:30 am to 9:30pm).Therefore there are 780min per working day which can be divide into 78 ten minutes increments Random numbers Usable random numbers 49 53 25 45 20 73 49 53 25 45 20 73 Arranged in numerical order 1 2 12 13 15 20 Corresponding observation time 8:30 8:40 10:20 10:30 10:50 11:40(12:40) 3

12 74 37 51 46 13 53 1 25 15 2 77

12 74 37 51 46 13 1 15 20 77

26 37 45 46 49 51 53 73 74 77

12:40 14:30 15:50 16:00 16:30 16:50 17:10 20:30 20:40 21:10

16 Observation will be done per day Number of observation required per day = 16 observatio ns / day 22.13 =
354 observatio ns

23days

6.4 Interim Calculation

Activity sampling analysis sheet Cafeteria Mrs. S.C Heizer PILLING COOKING PACKING Activities First study: 09 March 2009 Analyst: Mr. A.H Van Zyl NONPRODUCTIVE OTHERS

cumulative

TIME 8:30 8:40 10:20 10:30 10:50 11:40(12:40) 12:40 14:30 15:50 16:00 16:30 16:50 17:10 20:30 20:40 21:10 Total Cum total Percentages

6 6 37.5 7 7 43.75 2 2 12.5 1 1 6.25 0 0 16

WORK

4 P (100 P) 4 * 43 .75 (100 43 .75 ) = =394 L2 52 N required = 394-16 = 378

N=

L=2

P (100 P ) =2 N

43 .75 (100 43 .75 ) = 24.8 16

7. FINDINGS

In all calculation made above it was found that only 15days are necessary to make the full study. This will result in the difference of 8 days from the 23 initial estimated days .The desired limit of error was found on the 24th March 2009, when 239 observations were made. The desired limit of error occurred when 120 days were still required.

8. CONCLUSION The conclusion was made from all the findings obtained and all the calculations made. It can conclude by stating the following percentage of delays: Teller-20.08% Cooking-19.67% Packing-20.08 Non productive work-20.08 Others-20.08

9. RECOMMENDATION Worker are the ones whom are exposed to work more than any one, they should be asked for any recommendation of solving problems. Short breaks should be given to workers after working a busy period e.g. lunch time. Standard times should be introduced to the system. The control charts should be regularly used to test the production system. The 10. BIBLIOGRAPHY Morar B, 2008 .Engineering work-study 2. Unpublished lecture notes Cape Peninsula University of technology: Cape Town