Cyber-Physical Production Systems (CPPSs), such as automated car manufacturing plants, execute a configurable sequence of production steps to manufacture products from a product portfolio. In CPPS engineering, domain experts start with manually determining feasible production step sequences and resources based on implicit knowledge. This process is hard to reproduce and highly inefficient. In this paper, we present the Extended Iterative Process Sequence Exploration (eIPSE) approach to derive variability models for products, processes, and resources from a domain-specific description. To automate the integrated exploration and configuration process for a CPPS, we provide a toolchain which automatically reduces the configuration space and allows to generate CPPS artifacts, such as control code for resources. We evaluate the approach with four real-world use cases, including the generation of control code artifacts, and an observational user study to collect feedback from engineers with different backgrounds. The results confirm the usefulness of the eIPSE approach and accompanying prototype to straightforwardly configure a desired CPPS.

翻译：信息物理生产系统（CPPS），如自动化汽车制造工厂，通过执行可配置的生产步骤序列来制造产品组合中的产品。在CPPS工程中，领域专家最初基于隐性知识手动确定可行的生产步骤序列和资源。这一过程难以复现且效率低下。本文提出扩展迭代式流程序列探索（eIPSE）方法，可从领域特定描述中推导产品、流程和资源的可变性模型。为实现CPPS的集成式探索与配置过程自动化，我们提供了一套工具链，可自动缩减配置空间并支持生成CPPS工件（如资源控制代码）。我们通过四个真实用例（包括生成控制代码工件）以及一项面向不同背景工程师的观察性用户研究对该方法进行评估。结果证实了eIPSE方法及配套原型在直接配置所需CPPS方面的实用性。