Grátis 62 pág.
- Denunciar
Pré-visualização | Página 15 de 20
benchmarks against which to measure a range of variables such as cost of the product and share of work between team members. However the work measurement technique has been criticised for being misused by management in determining worker compensation. The time needed to perform each work element can be determined by the use of historical data, work sampling or most usually time study. 10.6.1 Time Study The purpose of Time Study is through the use of statistical techniques to arrive at a standard time for performing one cycle of a repetitive job. This is arrived at by observing a task a number of times. The standard time refers to the time allowed for the job under specific circumstances, taking into account allowances for rest and relaxation. The basic steps in a time study are indicated below: 1. Establish the standard job method It is essential that the best method of undertaking the job is determined using method study before a time study is undertaken. If a better method for the job is found then the time study analysis will need to be repeated. 2. Break down the job into elements The job should be broken down into a number of easily measurable tasks. This will permit a more accurate calculation of standard time as varying proficiencies at different parts of the whole job can be taken into account. 3. Study the job This has traditionally been undertaken with a stopwatch, or electronic timer, by observation of the task. Each time element is recorded on an observation sheet. A Video camera can be used for observation, which permits study away from the workplace, and in slow motion which permits a higher degree of accuracy of measurement. 4. Rate the worker’s performance As the time study is being conducted a rating of the worker’s performance is also taken in order to achieve a true time rating for the task. Rating factors are usually between 80% and 120% of normal. This is an important but subjective element in the procedure and is best done if the observer is familiar with the job itself. Download free eBooks at bookboon.com Operations Management 47 Work Systems Design 5. Compute the average time Once a sufficient sample of job cycles have been undertaken an average is taken of the observed times called the cycle time. The sample size can be determined statistically, but is often around five to fifteen due to cost restrictions. 6. Compute the normal time Adjust the cycle time for the efficiency and speed of the worker who was observed. The normal time is calculated by multiplying the cycle time by the performance rating factors. Normal Time (NT) = cycle time (CT) x rating factor (RF) 7. Compute the standard time The standard time is computed by adjusting the normal time by an allowance factor to take account of unavoidable delays such as machine breakdown and rest periods. The standard time is calculated as Standard Time (ST) = Normal Time (NT) x allowance 10.6.2 Predetermined Motion Times One problem with time studies is that workers will not always co-operate with their use, especially if they know the results will be used to set wage rates. Combined with the costs of undertaking a time study, a company may use historical data in the form of time files to construct a new standard job time from previous job element. This has the disadvantage however of the reliability and applicability of old data. Another method for calculating standard times without a time study is to use predetermined motion time system (PMTS) which provides generic times for standard micromotions such as reach, move and release which are common to many jobs. The standard item for the job is then constructed by breaking down the job into micromotions that can then be assigned a time from the motion time database. The standard time for the job is the sum of these micromotion times. Factors such as load weight for move operations are included in the time motion database. The advantages of this approach are that standard times can be developed for jobs before they are introduced to the workplace without causing disruption and needing worker compliance. Also performance ratings are factored in to the motion times and so the subjective part of the study is eliminated. The timings should also be much more consistent than historical data for instance. Disadvantages include the fact that these times ignore the context of the job in which they are undertaken i.e. the timings are provided for the micromotion in isolation and not part of a range of movement. The sample is from a broad range of workers in different industries with different skill levels, which may lead to an unrepresentative time. Also the timings are only available for simple repetitious work which is becoming less common in industry. 10.6.3 Work Sampling Work Sampling is useful for analysing the increasing proportion of non-repetitive tasks that are performed in most jobs. It is a method for determining the proportion of time a worker or machine spends on various activities and as such can be very useful in job redesign and estimating levels of worker output. The basic steps in work sampling are indicated below: Download free eBooks at bookboon.com Operations Management 48 Work Systems Design 1. Define the job activities All possible activities must be categorised for a particular job. e.g. “worker idle” and “worker busy” states could be used to define all possible activities. 2. Determine the number of observations in the work sample The accuracy of the proportion of time the worker is in a particular state is determined by the observation sample size. Assuming the sample is approximately normally distributed the sample size can be estimated using the following formula. n = (z/e)2 * p(1 - p) where n = sample size z = number of standard deviation from the mean for the desired level of confidence e = the degree of allowable error in the sample estimate p = the estimated proportion of time spent on a work activity The accuracy of the estimated proportion p is usually expressed in terms of an allowable degree of error e (e.g. for a 2% degree of error, e = 0.02). The degree of confidence would normally be 95% (giving a z value of 1.96) or 99% (giving a z value of 2.58). Download free eBooks at bookboon.com Click on the ad to read moreClick on the ad to read moreClick on the ad to read moreClick on the ad to read moreClick on the ad to read moreClick on the ad to read moreClick on the ad to read moreClick on the ad to read moreClick on the ad to read moreClick on the ad to read moreClick on the ad to read moreClick on the ad to read moreClick on the ad to read moreClick on the ad to read moreClick on the ad to read moreClick on the ad to read moreClick on the ad to read more Operations Management 49 Work Systems Design 3. Determine the length of the sampling period There must be sufficient time in order for a random sample of the number of observations given by the equation in 2 to be collected. A random number generate can be used to generate the time between observations in order to achieve a random sample. 4. Conduct the work sampling study and record the observations Calculate the sample and calculate the proportion (p) by dividing the number of observations for a particular activity by the total number of observations. 5.Periodically re-compute the sample size required It may be that the actual proportion for an activity is different from the proportion used to calculate the sample size in step 2. Therefore as sampling progresses it is useful to re-compute the sample size based on the proportions actually observed. 10.7 Learning Curves Organisations have often used learning curves to predict the improvement in productivity that can occur as experience is gained of a process. Thus learning curves can give an organisation