R. Radharamanan

Sales forecasting using time series and neural networks

Abstract This paper presents the use of time series ARIMA model with interventions, and neural network backpropagation model in analyzing the behavior of sales in a medium size enterprise located in Santa Maria (RS), Brazil for the period January 1979 - December 1989. The forecasts obtained using the backpropagation model were found to be more accurate than those of ARIMA model with interventions.

Quality function deployment as applied to a health care system

Quality Function Deployment (QFD) is used in a health care system to deploy the voices of the customers in understanding their requirements and to include them for continuous improvement of quality in services provided.

A heuristic algorithm for group scheduling

Abstract Group technology is a rapidly developing productivity improvement tool that can have a significant impact on the development of totally integrated manufacturing facilities and flexible manufacturing systems. Production scheduling associated with group technology is called “Group Scheduling”. There are many heuristic algorithms developed for general job shop applications based on unrealistic hypothesis, complicated computations etc., which are not addressed to group scheduling. In this paper, from the existing algorithms for group scheduling, a heuristic algorithm has been developed and programmed for computer/microcomputer applications. The developed algorithm has been used to determine the optimal group and the optimal job sequence for a batch type production process with functional layout. The developed algorithm is far simpler and easier to compute, compared to the other similar heuristic algorithms and certainly in comparison to other optimization methods such as branch and bound method.

Quality and productivity improvement in a custom-made furniture industry using Kaizen

Kaizen philosophy is applied in a small-sized custom-made furniture industry for continuous improvement and to develop the products with higher quality, lower cost, and higher productivity in meeting the customer requirements.

A branch and bound algorithm for the traveling salesman and the transportation routing problems

Abstract In this paper, a branch and bound model with penalty tour building is developed for solving travelling salesman and transportation routing problems. The algorithm for determining the optimal solution of the problem is of the general form which can solve symmetric and asymmetric single and multiple travelling salesman problems (STS and MTS), and the transportation routing problems with capacity restrictions for the vehicles of same or of different capacities. The behavior of the developed algorithm is tested with randomly generated data. As an application, the routes for gas distribution is planned and carried out in several interior towns of a state, with the objective of making an efficient distribution in order to minimize the total cost of delivery with an optimal solution.

Sensitivity analysis on the CUSUM method

Abstract The cumulative sum (CUSUM) control chart is considered to be an alternative or complementary to Shewhart control charts in statistical process control (SPC) applications, owing to its higher sensitivity to small shifts in the process mean. It utilizes all the available data rather than the last few ones used in Shewhart control charts for quick decision making. The CUSUM method assumes that the process is stable and under control. In addition, the variability of the process should be well below when compared to the specification limits, i.e. the process must have a high capability ratio to permit variations in the mean of the process without affecting the quality of the product or operations. In this manner, the practical purpose of the CUSUM method is not really to verify whether the process is out of control or meets the specification limits, but for analyzing the shift in the process mean, stressing the aspects such as performance, productivity, cost, etc. The sensitivity analysis conducted on the CUSUM method establishes that it is not effective in a certain range of shifts in the mean of the process. The analysis on effects of shift in the mean and type-I error on the effectiveness of the method shows that the CUSUM does not predict the existence of shift in the process mean when the real shift is much higher than the shift that is being searched for its existence. A computational program has been developed for calculating the minimum value of α (type-I error) given certain values of the shift in the mean. With these results it is possible to observe that, for a given value of α, the CUSUM method may not detect a shift in the mean greater than the value of the expected shift.

Use of multivariate analysis in controlling a soft drink fabrication process

Multivariate analysis when used to control sets of dependent variables that are present in a soft drink production process provided better insight in detecting out of control states than that of Shewhart control charts.

Group technology concepts as applied to flexible manufacturing systems

Abstract Group technology is a rapidly developing productivity improvement tool that can have a significant impact on the development of totally integrated manufacturing facilities and flexible manufacturing systems. In flexible manufacturing systems (FMS), parts grouping into families can be performed to simplify the long- and short-horizon planning. In this paper, the part families problem applicable to short-horizon planning has been discussed. Some of the traditional parts grouping based on classification and coding systems have been reviewed. New methodologies for overcoming the drawbacks of the classical approach to parts grouping have been presented and discussed. The methodologies are easy to implement since they take advantage of the information already stored in the CAD system. In addition, a heuristic algorithm based on group scheduling concept has been developed and used to determine the optimal group and the optimal job sequence for the batch-type production process.

Standardization of the nutrition and diet division of the university hospital in Santa Maria

This paper presents a proposal of standardization of the Nutrition and Diet Division, (NDD) of a University Hospital based on the quality philosophy in services and standardization of enterprises. The objective is to identify and document the routines of some of the existing categories through a routine manual. The procedures used to standardize the NDD show that more participation of the management in solving divisional problems, and adequate training of the employees are required.

Spectral decomposition in statistical process control

Abstract Spectral decomposition method has been used in obtaining statistical control of time series data analyzed providing quick response in detecting out of control state of the process compared to that of Shewhart control charts.