Ana Esteso1, Andrés Boza1, MME Alemany1 and Pedro Gomez-Gasquet1
Ana Esteso1, Andrés Boza1, MME Alemany1 and Pedro Gomez-Gasquet1
1 Research Centre on Production Management and Engineering (CIGIP), Universitat Politècnica de València, Camino de Vera S/N, 46022 Valencia, Spain
<EMAIL_ADDRESS><EMAIL_ADDRESS><EMAIL_ADDRESS><EMAIL_ADDRESS>Keywords: Optimization, Production Planning, Industry 4.0, MPL, Pyomo.
1. Introduction
The main purpose of digital transformation is to redesign the organizational business through the introduction of digital technologies, achievement benefits such as productivity improvements, cost reductions and innovations [1]. To production enterprises, Industry 4.0 has become one of the most commented industrial business concepts in recent years. This new paradigm in industry promotes the autonomous decision making, interoperability, agility, flexibility, efficiency, and cost reduction among others [2].
According to [3] the four main characteristics of Industry 4.0 include: a) Vertical integration of smart production systems; b) Horizontal integration through global value chain networks; c) Through-engineering across the entire value chain; and d) Acceleration of manufacturing.
Thus, production planning systems must continue to reinforce their vertically integrating role in organizations but making use of this new set of technologies. Mathematical programming, and more specifically linear programming, has been traditionally employed to support the production planning process [4]. Since commercial optimization software usually provides free licenses to academics, this type of software has been used to a greater extent in the academic sphere.
However, the academic nature of these licenses makes it difficult to transfer models developed in the academy to the business sector, where paying for a license may not be profitable. In addition, research projects tend to request both, the developed models, and their code, to be openly published. To meet this requirement, it makes sense for the academy to migrate to free software that allows the free dissemination of mathematical programming models and their codes.
On the other hand, tools based on artificial intelligence have begun to be developed to support production planning. For this, it is important to work with software that allow us not only to solve optimization models but also to connect them with heuristics and artificial intelligence algorithms. This will facilitate the later interoperability with other systems and programming the required agility and flexibility achieving more efficiency, and cost reduction that are in line with Industry 4.0.
In this context, this paper proposes a conceptual framework (CF) to facilitate the migration of mathematical programming models implemented in commercial optimization software to optimization packages developed in high-level programming languages, that allows their integration to the techniques mentioned above. More concretely, this conceptual framework facilitates the migration between the MPL optimiza