Optimization Patterns Enabled by Industry 4.0/5.0 Data

Abstract

Design patterns are widely adopted in software engineering and can be defined as “general design solutions to recurrent problems”. We extend this approach to improve the modeling activities required for addressing optimization problems, specifically those coming from production planning in manufacturing firms. We propose a general framework for optimization patterns, including general, problem and solver specific patterns as well as data connection patterns used to collect and effectively use the large data sets that can be provided by firms that adopt Industry 4.0 paradigm, or also the newer and complementary 5.0 paradigm. We then provide a vertical application of such patterns to address a strategic planning problem inspired by a real application in a manufacturing company operating in Italy. The case study shows all the main activities that begin with the problem description, continues through data definition and optimization patterns selection, finally leading to a prototypical solution to the problem.