Smart manufacturing is targeted as the next generation of manufacturing by many national and international strategic development. The increasingly rich production data, the integration and extensive application of information technology, and the intelligent data processing and system modeling methods have collectively enabled smart manufacturing. Building upon them, manufacturing system modeling, knowledge acquisition, design, and real-time control are the key components [item 1) in the Appendix], [item 2) in the Appendix]. It is still one of the really huge challenges to gather data, transform them into information, and derive knowledge out of this information, especially given the requirement of knowledge that can be trusted as manufacturing systems may harm humans and the environment if they come to the wrong conclusion. Despite the learning and derivation of knowledge, it could be modeled beforehand and taken, for example, as an environmental model for online decisions like in deliberative agent-based systems. Nevertheless, for decisions during operation, real-time requirements, dependability, and security issues are to be guaranteed. Finally, for acceptance and trust, humans need to “understand” the reasons behind automated decisions; therefore, explainability or at least a white box description is an issue of such knowledge-based systems.

中文翻译：

---

智能制造系统基于知识的自动化


智能制造被许多国家和国际战略发展作为下一代制造业的目标。日益丰富的生产数据、信息技术的集成和广泛应用、智能数据处理和系统建模方法，共同使能智能制造。在此基础上，制造系统建模、知识获取、设计和实时控制是关键组成部分[附录中的第 1 项]、[附录中的第 2 项]。收集数据、将其转化为信息并从这些信息中获取知识仍然是真正巨大的挑战之一，特别是考虑到需要可信的知识，因为制造系统如果进入生产环境可能会危害人类和环境。错误的结论。尽管需要学习和推导知识，但它可以预先建模并作为在线决策的环境模型，例如基于协商代理的系统。但对于运行过程中的决策来说，实时性要求、可靠性、安全性问题都是需要保证的。最后，为了获得接受和信任，人类需要“理解”自动决策背后的原因；因此，可解释性或至少是白盒描述是此类基于知识的系统的一个问题。