Adaptive simulation of a keyboard assembly cell

Abstract

Describes the development of an adaptive simulation model for a keyboard assembly cell for real‐time decision support. Discusses the architecture of the modelling and control system, including the movement of entities and conveyors, describing how up to four different keyboard types may be modelled, with a PC cell controller continually monitoring the state changes of the assembly line, passing the data captured to the simulation model created in ARENA.

Automated manufacturing system discovery and digital twin generation

“…The optimization function (12) represents the maximization of the model score expressed by equation (6). Constraints (13) and ( 14) limit the number of nodes and arcs included in the model, respectively.…”

“…The goal of this section is to determine which configuration of weights contributes the most to the objective function described in equation (12). The function is a weighted sum of the scores defined in section 4.1.…”

“…A simulation model of the flow line has been developed in Rockwell Arena and used to generate five independent event logs through simulation experiments representing the production of 1000 parts. For the calibration experiments, we have exploited a space-filling design with a 256 points over the values of the weights of objective function (12). Each point has been replicated five times.…”

“…Traditional model generation techniques use the available data to estimate model parameters, such as inter arrival times or rework ratios. These approaches are based on an existing model structure, which is typically derived from previous knowledge or from interviews with company experts [12]. More recent approaches introduced the exploitation of process mining techniques [13], which enable to retrieve the system topology from the manufacturing system data [14].…”

Some aspects of real‐time production control in distributed flexible assembly systems

“…Based on the earlier works of Anger (1989), Chandi and Mistra (1981), Lamport (1981), Ranky (1982Ranky ( , 1985Ranky ( , 1986aRanky ( , 1990Ranky ( , 1997a, Ranky and Francis (1986) and Tanik and Chan (1995) as well as related publications that underline the importance of research in this direction Bartholomew (1997), Bowman (1997), Chen (1997), Estrada et al (1997), Gong (1997), Hase (1997), Rooks (1997), Spedding (1997), Walton and Whicker (1996), Westkampfer (1997) and Wiendahl (1997), let us discuss a new, extended algorithm to illustrate the operation and messaging processes in our distributed model of dynamically scheduled lean and flexible design and manufacturing systems operating on a global basis. In order to simplify the understanding of our model, let us assume that all processes that are subject to our analysis actually occur within the same time frame.…”