Industry 4.0, which describes the transformation of existing production environments toward smart factories, is implemented in ever more manufacturing companies. Smart factories offer diverse advantages such as high flexibility, dynamic scheduling, as well as accurate fault diagnosis and prediction. Hence, manufacturing companies need support for assessing which projects they should implement to transform their production environment. As no such guidance exists in the literature, we propose a multi-dimensional decision model that accounts for interdependencies among production components, for projects with different performance effects, and for digital capabilities constitutive of smart factories (i.e., real-time ability, interoperability, virtualisation and decentralisation). The decision model schedules smart factory projects over multiple planning periods and assesses project roadmaps in line with objectives that comply with established performance measures and the digital capabilities of smart factories. We evaluate and discuss the decision model in interviews with two factory managers and three researchers with great experience in the smart factory domain. Based on a software prototype, we also successfully applied the decision model at a manufacturing company based on real-world data.

工业4.0描述了将现有生产环境向智能工厂的转变，并在越来越多的制造公司中得到实施。智能工厂具有多种优势，例如高灵活性，动态调度以及准确的故障诊断和预测。因此，制造公司需要支持来评估他们应该实施哪些项目以改变其生产环境。由于文献中没有此类指导，因此我们提出了一个多维决策模型，该模型考虑了生产组件之间的相互依赖性，具有不同性能影响的项目以及智能工厂的数字功能构成（即实时能力，互操作性，虚拟化和分散化）。决策模型可在多个计划期间内计划智能工厂项目，并根据符合既定绩效指标和智能工厂数字能力的目标评估项目路线图。我们在与两位在智能工厂领域具有丰富经验的工厂经理和三位研究人员的访谈中评估和讨论了决策模型。基于软件原型，我们还基于实际数据在一家制造公司成功应用了决策模型。