Using Nested Simulation for Evaluating Next Period Workload Anticipation Capability of a Control System

Simulation is a powerful tool able to anticipate the behavior of complex systems, often this ability is used for on-line control strategy in several industrial and civil application. Models are currently used for supporting next period workload in gas dispatching networks as well as in electrical power generation management. In the process industry (i.e. steel making plants) simulation is the base technique for ensuring proper appointment among different process (i.e. Ladle Furnaces and Continuous Casting); in maritime logistics several application have been proposed for container yard planning and for mooring optimization. More in the manufacturing sector several application of Scheduling and Simulating systems are well known. In all such applications on-line simulation have been extensively used for supporting operative decision making. In the proposed application a department store is here controlled by on-line simulator devoted to investigate the impact of counter’s management policy. In the common practice, in fact, a fixed number of people is used both for counter’s staffing and for shelves replenishment. Opening a counter, generally speaking, will divert a worker from shelf replenishment in favor of a reduction in the customer’s average time spent in queue resulting a temporary increase in the stock out risk. The effect of such behavior is delayed and a consequent oscillating path will emerge affecting seriously the performances of a real life department store. The paper presents a reflective simulation model able to simulate such system with an embedded simulator in order to investigate the expected performances of such simulation-in-the-loop control model. Methodology and case study are presented and discussed.