Implementing Digital Twins of Smart Factories with Interval Algebra

Abstract

This paper presents an interval algebra that can be used to create an analytic description of a smart factory. Such a description (recently termed a 'digital twin' of the factory) is used to evaluate alternative manufacturing configurations as part of a search-based optimisation process. Several extensions are proposed to the interval algebra for specifying smart factory production line details. A number of real-life manufacturing scenarios are described, related to Wire-cut Electrical Discharge Machining. The experimental results show the applicability and scalability of the proposed method.

Cloud-Based Integrated Process Planning and Scheduling Optimisation via Asynchronous Islands

“…This section investigates the efficacy of the proposed cloud-based deployment of the GA algorithm based on the asynchronous island management model when applied to the real-world use case described in [4], in which a set of 14 metal parts is ordered to be manufactured in a plant equipped with 12 Wire Electrical Discharge Machining (WEDM) machines of 3 different sizes, operating in 3 various modes each. The considered optimisation problem is an example of a typical manufacturing planning and scheduling problem that can be found in any discrete manufacturing company.…”

“…Such ontology can be then used to build an analytic description of a smart factory (aka digital twin), as proposed in [4], which then can evaluate alternative manufacturing configurations as a part of a search-based optimisation process. As this process is executed in the cloud on-demand only, the applied computing resources can be relatively powerful and large in quantity as long as the computation time is not long-lasting.…”

Process planning and scheduling optimisation with alternative recipes

“…In each iteration, a new generation is formed from the current one in the following manner. Firstly, the objective values (tness) of each individual in the current population are computed using the algebraic model described in [3]. From the current popu-lation, two neighbouring chromosomes are selected randomly with the probability proportional to their tness values and they form a pair of new chromosomes by exchanging their genes in a crossover process (a standard singlepoint crossover operator is applied).…”

Cloud-based dynamic distributed optimisation of integrated process planning and scheduling in smart factories

“…After receiving such a message, OE performs an optimisation process using an objective function customised for a given plant. In the considered cases, the objective function is based on the interval algebra described in [4]. In the proposed solution, this function is generated automatically based on a factory description specified by an XML-based format, derived from the common situation model.…”