Modeling and controlling work‐in‐progress in discrete manufacturing systems

Abstract

Purpose

The purpose of this paper is to apply the proportional integral (PI) control algorithm in discrete manufacturing enterprises to maintain lower and steadier work in progress so as to improve on‐time delivery.

Design/methodology/approach

A sensitivity constrained optimization model is designed on the frequency domain, whose optimum algebraic solutions are then obtained easily. Two controllers, a backlog controller and an input‐rate controller, are devised, which correspond to the integral control and the proportional control of PI controllers, respectively. Interacting with each other, these controllers have made the engineering implementation of PI controllers a reality.

Findings

Simulation is carried out in certain motorcycle production lines. Results confirm that PI controllers also possess good control effects in the discrete manufacturing industry, as well as in the process industry.

Research limitations/implications

A continued departure from the nominal may happen repeatedly if the root causes of changing are not detected and identified. Moreover, PI controllers can mask process defects, failures, and drifts, and this may lead to eventual catastrophic failures. So, statistical process control should be utilized in PI controlled processes to detect significant changes for long‐term process improvement.

Practical implications

PI controllers possess potential in discrete enterprises.

Originality/value

PI controllers are tried for process improvement in discrete manufacturing enterprises.

Citation

Cai‐fen, G. and Ran‐zhe, J. (2011), "Modeling and controlling work‐in‐progress in discrete manufacturing systems", Kybernetes, Vol. 40 No. 5/6, pp. 842-847. https://doi.org/10.1108/03684921111142386