Smart workload input-output control of production systems: A proof of concept

Abstract

Workload control (WLC) relies on order release (input control) and capacity adjustments (output control) to regulate manufacturing throughput times. Research concerning integrated input-output control is very scarce. A novel approach is proposed: a closed-loop integrated model, where automatic feedback control is added and where order release and capacity decisions are a continuous function of the shop floor current state. In existing models, capacity increments are triggered by a threshold utilisation and have a fixed value, defined by means of experiments. In the proposed approach, these values are dynamically and flexibly defined (by the controller), addressing a gap regarding input-output parameterisation. The model was applied to a shop with unidirectional flow and three control rules were tested in scenarios of unbalanced work in process (WIP) and demand fluctuations. The results showed that the simultaneous input-output control is effective, leading the system to restore WIP balance, reach stability and absorb demand fluctuations, especially when global information is provided to the controller. The results of continuous closed-loop model simulation were integrated into a discrete-event simulation (DES) model, to compare the proposed capacity policies with the policies presented in the literature. The comparison showed that the proposed policies are cost-effective for a reasonable throughput time reduction. Moreover, the closed-loop model provided good parameters for a step-increment capacity policy that minimizes throughput times. Closed-loop control models are responsive and prescriptive, in contrast to DES models used in WLC research, which are descriptive. With this approach, WLC research can benefit from the strengths of both methods.