Dec 13, 2017 in Management

Subject OS on Operations Management

Con-Way Freight is a transportation company located in Houston. The company deals in the transportation of packages within the state of Texas, particularly within the populous city of Houston. The mode of operation of the company is that customers call into the customer-service section of the company through well-known public hotlines and book for the transportation of their packages. As such, the operators of the hotlines of the company handle a substantially large number of calls every day. Therefore, they have to predict the number of calls that they are likely to receive every day in order to plan ahead of time. They need to plan the number of operators and the number of phones that they should put into use. They need to provide timely service to customers who call in to schedule their pick-ups.

Every single day, there is a forecast of the number of calls that are expected. These forecasts are based on a number of factors. Among these factors are the day of the week, the previous trends and the season. However, these forecasts are never accurate, resulting in a mismatch between the number of calls that are received from customers and the number of operators that are ready to receive these calls. The result is that on some days the operators are overwhelmed by the number of calls received from customers. On other days, some operators sit idle waiting for calls that never materialize. There is, therefore, a need to analyze the data. This should translate into more accurate predictions of the calls that are expected to schedule the pick-ups.

THE MODULES USED FOR ANALYSIS

In order to analyze the data, three modules will be used. These modules will be the moving average, weighted average and trend analysis. These modules are often used in smoothing the forecasting of trends.

The moving average, also known as the rolling average, is used in analyzing several data points. This is done by creating a sequence of averages of diverse subsets within the entire data set. The subsets are modified by shifting the data forward. This is done by continuously excluding the first numbers in the sequences and adding new numbers in new sequences. The moving average is appropriate for this type of data because of its usefulness in analyzing short term fluctuations with a view of getting the long-term trends.

The weighted average is a type of average in which some data points carry more weight than others. This is particularly important in this type of forecasting because we should expect some days to be more useful in prediction than others.

Finally, trend analysis involves examining the underlying trend in a set of data points by eliminating the noise. These are data points that seem to distract attention from the underlying pattern. This would be particularly useful in analyzing this form of data because of its outstanding importance in forecasting.

INTERPRETATION OF COMPUTER OUTPUT

In this particular set of data, there are 36 data points for the calls for scheduled pick-ups. If, for the 36 data points, 4 periods are used to calculate the rolling average, the moving average for each of the subsequent four averages is 247.3636, 247.9697, 248.303 and 246.2424. This shows that the number of calls expected for each day is between 246 and 248. This, however, is a general trend, and it might be a little misleading for individual days. However, for planning and forecasting purposes, these are extremely useful values.

InĀ  the trend analysis, a few numbers are eliminated in order to visualize the general trend of the data. In essence, we could categorize the data into data sets within the range of 210 and 280 in tens. This means that a general trend could be deduced by finding out which set tens between these two numbers has the highest number of entries in these data sets.

In analyzing this data, a few other important parameters have been used, which are the mean absolute deviation (MAD), the Mean square error (MSE), the mean absolute percentage error (MAPE) and the least absolute deviation (LAD). These are all parameters that have an invaluable in the field of forecasting.

The MAD is 24.86806. This means that the difference between the forecasted calls and calls that materialized is averagely 24.86806. Therefore, if, going by the moving average, the management of Con-Way Freight was to expect between 246 and 248 calls every day, it must factor in this average deviation. It must, therefore, expect about 25 calls less or more that their projections every day.

The mean square error is 913.5469. This gives a picture of what we would expect at finding the square root. This would also give a standard deviation of about 30.22. This might be used as an alternative to the MAD, and it would also be used to adjust the expectations into a more accurate range.

The MAPE is 0.102698. This value is also instrumental in calculating the range of error when estimating the number of calls we should expect as a deviation. We could calculate it as a percentage of the absolute number of calls that are forecasted which is about 10 percent of the estimated 246-248. Moreover, this number should be much smaller in order to increase the accuracy of the forecast.

Finally, the LAD is also instrumental in finding out the accuracy of forecasting. In this case, the LAD is 33.5.This implies that on the absolute scale, the deviation from the forecasted number of calls amounts to about 33.5. This value could be used to adjust the forecasted number of calls into a more accurate range.

CONCLUSION

The use of moving mean and trend analysis makes extensive use of the parameters of LAD, MSE, MAD and MAPE in order to adjust the range of the forecasted values into a more acceptable range than if it were not used at all. In the case of Con-way Freight, for instance, the forecasted number of calls for each day could be different depending on various factors. However, given the data points provided for the 36 periods stated, the moving mean puts the number of forecasts between 246 and 248. Beyond this, the numbers could be adjusted into a more realistic range by using the MSE to about 215.78 to about 278.22. Alternatively, the MAD could be used to adjust the figure to 221.13 to 272.87. However, if the range is to be reduced to a narrower, workable range, then the MAPE needs to be reduced below 0.102698

Related essays