Stochastic model for manufacturing cost estimating

  • Authors:
  • C. T. Abraham;R. D. Prasad

  • Affiliations:
  • IBM Corporate Headquarters, Armonk, New York;IBM Corporate Headquarters, Armonk, New York

  • Venue:
  • IBM Journal of Research and Development
  • Year:
  • 1969

Quantified Score

Hi-index 0.00

Visualization

Abstract

The unit manufacturing cost (i.e., its estimator) for a given manufacturing program with stochastic demand and operation yield is assumed to be a random variable. For a simple series production line the probability distribution of the unit manufacturing cost has been derived by either the transform method, which uses Mellin and Laplace transforms, ort he method of moments, which uses either the Gram-Charlier series approximation or the Pearson system of frequency curves. The estimates and 90%-confidence intervals for the base manufacturing cost are computed for two device-component products. The model cost estimates are very close to the actual values and the confidence intervals are sufficiently narrow to be useful in applying contingencies to the predictions.