Raw material release rates to ensure desired production lead time in Bernoulli serial lines

Production systems are often managed to maximise their production rate (). In some cases, this leads to excessive work-in-process and, as a result, long production lead time (). This issue is of particular importance in systems with unlimited' buffer space, where may also become unlimited. The question addressed in this paper is: What should be the release rate of raw materials so that the desired is obtained, while is maximised? We answer this question for serial lines with infinite buffers and machines obeying the Bernoulli reliability model. Specifically, given a serial line, we quantify the set of attainable s (feasible set), and for each point in the feasible set provide the release rate (as a function of machine parameters) that guarantees the desired lead time, while maximising the production rate. In addition, we offer a feedback control law that enforces the desired system behaviour. The development is based on a recursive aggregation procedure that leads to analytical estimates of . The accuracy of these estimates is evaluated by simulations and is shown to be acceptable for practical applications (well within 5%). The results obtained offer small- and mid-size enterprises the possibility of managing their production lead time, when neither limited buffers nor kanban/CONWIP systems are available, by releasing raw materials on, say, an hourly basis.