Integration of AI into Manufacturing Execution System (IMES)

加拿大的制造业必须变得更有竞争力，并在全球市场上占有更大的份额，目前约为1%。向加拿大制造商引入具有智能和适应性生产技术的工业4.0可以为实现这一目标提供显着优势。这些技术寻求将自动化技术与数据分析相结合，以做出智能制造决策，这些决策对扰动具有鲁棒性，对生产中的变化具有灵活性。该提案描述了UBC实验室和NTWIST公司之间的研究伙伴关系，NTWIST公司与德国的研究和工业合作者合作，旨在设计和实施智能制造执行系统（IMES），用于制造业的广泛应用。拟议的IMES是一种通用的、适应性强的多智能体生产协调和自动化系统，可用于不同行业，以促进其生产过程的数字化，并更快地向工业4.0迈进。该研究响应了将人工智能与机器学习和数据分析功能无缝快速集成到工厂中的需求。目前，即使是非常先进的MES系统也没有这种能力。UBC团队的应用人工智能专业知识和NTWIST在广泛行业的领域专业知识将被结合起来，以加速加拿大制造业的数字化。具体而言，该项目有助于i)研究和开发一种新的数据驱动的IMES决策范式；ii)通过使用开发的IMES在OPC UA驱动的制造系统中训练其代理，快速集成用于控制子系统的机器学习系统；以及iii)在按比例缩小的网络物理生产测试平台中实施开发的IMES。拟议的IMES最终将在加拿大和德国的工厂进行测试，这些工厂将分别由NTWIST和我们的德国合作者带入该项目。

Canada's manufacturing sector has to become more competitive and claim a larger share of the global market, currently at about 1%. The introduction of Industry 4.0 with intelligent and adaptable production techniques to Canadian manufacturers can offer significant advantages for realizing this objective. These techniques seek to integrate automation technologies with data analytics to make intelligent manufacturing decisions that are robust to perturbations and flexible to changes in production. This proposal describes the research partnership between UBC laboratories and NTWIST Inc., which in collaboration with research and industry collaborators in Germany, aim to design and implement an Intelligent Manufacturing Execution System (IMES) for broad applications in the manufacturing sector. The proposed IMES is a generic and adaptable multiagent production coordination and automation system that can be utilized by different industries to facilitate the digitalization of their production processes and faster progression towards Industry 4.0. The research responds to the need for seamless and rapid integration of artificial intelligence with machine learning and data analytics capabilities into a factory. Such capabilities are currently absent in even highly advanced MES systems. The applied artificial intelligence expertise of the UBC team and the domain expertise of NTWIST in broad industries will be incorporated to accelerate digitalization of manufacturing in Canada. Specifically, the project contributes to i) the research and development of a novel data-driven decision-making paradigm for IMES; ii) rapid integration of a machine learning system for controlling subsystems through training of their agents in an OPC UA driven manufacturing system using the developed IMES; and iii) implementation of the developed IMES in a scaled-down cyberphysical production testbed. The proposed IMES will ultimately be put to test in both Canadian and German factories which will be brought into the project by NTWIST and our German collaborators, respectively.