流程行业大多属于基础行业，在国民经济中具有举足轻重的作用，但因其特殊的生产工艺流程，又使得流程制造行业数字化转型相对离散制造较为缓慢。如何通过数智化提升企业经营韧性及竞争力，化解经济下行面临的挑战；如何解密流程制造数字化转型的数据“黑匣子”，实现真正的数据驱动；如何通过合理的数智化规划及实践，加速企业自身智能转型的进程等等，都是当下流程制造企业数字化转型思考和急需解决的问题。
身处供应链、工艺管理，标准化等工作较为复杂的造纸行业，新加坡金鹰国际CIO王益民在多年的实践中对于流程企业智能工厂的理念、方法论，数智化实践过程中面临的挑战和解法都有丰富的积累。
一、智能工厂赋能流程制造降本增效
“常规的化工行业、冶金行业、信息化工等都称之为流程制造。流程制造有个比较大的特点，即整个投入量到产出是通过一系列流程来支撑的，工艺过程是连续进行的，不能中断。而且工艺过程的加工顺序是固定不变的，生产制造设备按照工艺流程来布置，大多数都安装有自动工控。。流程制造行业的特点使得行业的信息化起步比较早，加之流程行业的工艺设计比较特别，流水线上的工艺决定了产品的成本、质量等。因此，过程控制是流程制造行业面临的重要的挑战和紧迫的需求，而落实过程控制的路径，正当下行业大多数企业正在践行的：智能工厂。
对此，王益民首先引用了工信部对于智能工厂的定义，其中共有7个重要步骤：
实现生产流程可视化和生产工艺优化；
实现物流、能量、数据流，价值流、资产的全过程监控，建立数据采集和监控系统；
采用先进控制系统，工厂自控投用率达到90%以上，关键生产环节实现基于模型的先进控制和在线优化；
建立生产执行系统，生产计划、调度均建立模型；对于存在较高安全与环境风险的项目，实现有毒有害物质排放和危险源的自动检测与监控、安全生产的全方位监控，建立在线应急指挥联动系统；
建立工厂通信网络架构，实现工艺、生产、检验、物流等制造过程各环节之间，以及制造过程与数据采集和监控系统、MES、ERP、SCADA、DCS之间的信息互联互通；
在使用智能化、网络化的前提下，一定要引起对网络安全的高度重视，建有功能安全保护系统。
在产品同质化比较严重的情况下，智能工厂可以显著提高产品质量，降低人员使用成本。一方面需要依靠企业现有的数字化技术，打造前中后的支撑平台。另一方面需要通过云计算、人工智能等先进的技术服务，集成CPS、BI、AI、AR等数字技术，进一步提升企业数字化转型的能力和成果，进而实现企业的智能工厂落地。
二、流程制造行业运营管理需求：生存、可持续发展、风险管控
在运营管理方面，王益民如是说到“流程制造行业智能工厂的运营管理需求，无非是生存、可持续发展、风险管控这三点，其中生存是目前来讲最重要的”。在产品同质化比较严重的情况下，智能工厂可以显著提高产品质量，降低生产成本。
此外，可持续发展也是工厂运营管理的重点。在研发新产品方面，PLM的应用使得通过大数据分析来研发新品成为可能，通过大数据分析、数字智能制造和使用场景的应用实现产品生命周期的管理。风险管控方面也离不开大数据的支持，不同的是，在风险管控方面，企业不仅需要工厂内部的数据，还需要市场外部数据以及第三方数据的支持，基于各种数据来决定企业的经营策略。
数据管理方面，产线实时数据的管理是至关重要。此外，设备运维日志也可以通过大数据分析建立模型来预判设备在未来出现故障的概率，帮助企业合理安排停机检修时间。而质量数据的采集和维护，可以帮助企业在流程制造过程中调整设备的温度、压力、车速等工艺参数，进而保证产品质量的稳定和成本的降低。
运营管理过程中，创新技术在整个生产过程中的应用，例如AI、自动优化供应链，在线设备传感器等可以有效的减少浪费，降低产品成本。
王益民介绍道，“从经营管理的角度来讲，完全可以做到零投资、大回报，如果我们在数字化进程中的某一个阶段缺乏资金，没有办法及时做到数字化的软件硬件设备投资，那重点就应该放在经营管理和精益方面”。
三、流程行业智能工厂建设三阶段
流程行业智能工厂生产管理的重点是生产过程实时透明、作业规范管理、全程可追溯。王益民提到，在了解智能工厂的状态时，可以通过智能工厂的评估模型来得到相关的评估结果。在工艺设计方面，特别是对于已经投产的工厂或者是中小型企业而言，工艺设计比较难改变。这就需要做一些数字化工程，比如可以通过激光扫描3D，把智能工厂做一个数字化的资产模型，关键工业装置做三维可视化模型，通过管理技术实现工厂的制造信息、设备信息、工厂信息等运维信息的集成。
对于信息化技术的关键内容，王益民从五个方面做了介绍。首先是生产作业的规范化，包括生产工艺数据标准化、作业流程规范化、数据驱动有了标准规范的流程协同优化。第二是计划管理的高级排程：承接ERP生产计划，这块的重点在于如何通过EMS的经营、排产，使得废品率减少和产成品的得率得到提高。第三是部分作业执行的精细协同，基于计划管理的高级排程，可以使企业的数据采集应用更加及时高效。第四是底层协同实时化。最后是生产制造的可视化，通过工业大脑，实现整个管理运营的数字化、智能化。
智能工厂的建设可以分为三个发展阶段：第一阶段是数字化和自动控制，包括工业传感器安装、数据采集等。第二阶段是网络化，例如，如何对设备联网，通过软硬一体化建设，使得我们整个设备数据对内对外都能够联系打通。第三阶段是智能化，包括数字孪生技术的应用、云计算大数据和AI建设等。
四、CPS是智能制造的本质
建设智能工厂最重要的一点是“CPS“，CPS是智能制造的本质”。工业互联网与物联网的应用是智能工厂底层必须完成的工作，没有CPS就无法做到智能制造。其次，业务人员的自我需求是推动智能工厂的核心动力。第三是以价值流为导向优化流程。企业在做智能工厂建设时，不能单纯为了满足上级的任务要求，而要以价值为导向，实现投入产出的有效平衡。
另外，要将AI技术服务于业务场景，以减少人工的使用。此外，要注重复合型人才团队培养，这不仅需要我们要有对产线生产设备的控制能力、操作能力，还需要员工有数据分析能力、数据建模能力。同时，在整个实施过程中，高层管理的支持无疑是必需的。
最后是“以点连线、以线带面、小步快跑”。当我们没有特别大的投资来保证数字化工艺实施的情况下，可以在自己的岗位做一些事情。当我们有一些大的数字化投资时，应该考虑是否可以改善数字化供应链，加大CPS投入，做智能排产优化。
王益民总结道：“智能工厂的投资不是立竿见影、马上就能看到效果的，而是需要一定的时间与积累，凡是量变到质变都是需要时间的。先从小的地方做起来，通过飞轮效应来持续改善。”
翻译：
Most of the process industry belongs to the basic industry and plays a pivotal role in the national economy. However, because of its special production process, the digital transformation of the process manufacturing industry is relatively slow compared with discrete manufacturing. How to improve the resilience and competitiveness of enterprises through digital intelligence, to resolve the challenges faced by the economic downturn; How to decrypt the process to make the digital transformation of the data “black box”, realize the real data driven; How to accelerate the process of enterprise’s own intelligent transformation through rational planning and practice of digital intelligence is the current process manufacturing enterprise’s digital transformation thinking and urgent problems.
In the paper industry where supply chain, process management, standardization and other work are more complex, Wang Yimin, CIO of Singapore Golden Eagle International, has accumulated rich knowledge of the concept and methodology of intelligent factory in process enterprises, as well as the challenges and solutions faced in the practice of digital intelligence.
Intelligent factory enabling process manufacturing cost reduction and efficiency increase
“The conventional chemical industry, metallurgy industry, information chemical industry and so on are called process manufacturing. Process manufacturing has a relatively big feature, that is, the whole input to output is supported by a series of processes, the process is continuous, can not be interrupted. And the processing sequence of the process is fixed, the production and manufacturing equipment is arranged according to the process flow, and most of them are equipped with automatic industrial control. The characteristics of the process manufacturing industry make the informatization of the industry start relatively early. In addition, the process design of the process industry is special. The process on the assembly line determines the cost and quality of the product. Therefore, process control is an important challenge and urgent demand for process manufacturing industry, and the path to implement process control is what most enterprises in the industry are practicing: intelligent factory.
In this regard, Wang Yimin first cited the Ministry of Industry and Information Technology’s definition of smart factory, which consists of seven important steps:
Visualization of production flow and optimization of production process;
Realize the whole process monitoring of logistics, energy, data flow, value stream and assets, and establish data acquisition and monitoring system;
The use of advanced control system, factory automatic control rate reached more than 90%, key production links to achieve model-based advanced control and online optimization;
Establish production execution system, production planning, scheduling model; For projects with high safety and environmental risks, automatic detection and monitoring of toxic and harmful substances discharge and hazard sources, comprehensive monitoring of production safety, and the establishment of an online emergency command and linkage system;
Establish the communication network architecture of the factory to realize the information interconnection among all links of the manufacturing process such as process, production, inspection and logistics, as well as between the manufacturing process and the data acquisition and monitoring system, MES, ERP, SCADA and DCS;
Under the premise of using intelligence and network, we must attach great importance to network security and build functional security protection system.
In the case of serious product homogeneity, smart factories can significantly improve product quality and reduce personnel costs. On the one hand, it is necessary to rely on the existing digital technology of the enterprise to build the front, middle and back support platform. On the other hand, advanced technical services such as cloud computing and artificial intelligence are needed to integrate CPS, BI, AI, AR and other digital technologies to further enhance the ability and achievements of digital transformation of enterprises, so as to realize the implementation of intelligent factories of enterprises.
Process Manufacturing industry operation management needs: survival, sustainable development, risk control
In terms of operation management, Wang Yimin said, “The operation management needs of smart factories in process manufacturing industry are nothing more than survival, sustainable development and risk control, among which survival is the most important at present.” In the case of serious product homogeneity, smart factories can significantly improve product quality and reduce production costs.
In addition, sustainability is also the focus of plant operations management. In the aspect of developing new products, the application of PLM makes it possible to develop new products through big data analysis. And realize the management of product life cycle through the application of big data analysis, digital intelligent manufacturing and usage scenarios. Risk management and control cannot be separated from the support of big data. The difference is that, in terms of risk management and control, enterprises not only need the data inside the factory. But also need the support of external market data and third-party data. Based on various data, enterprises can decide their business strategies.
In the aspect of data management, the real-time data management of production line is very important.
In addition, equipment operation and maintenance logs can also build models through big data analysis to predict the probability of equipment failure in the future, helping enterprises to reasonably arrange the downtime for maintenance. The collection and maintenance of quality data can help enterprises adjust equipment temperature, pressure, speed and other process parameters in the process of manufacturing. So as to ensure the stability of product quality and reduce costs.
In operation management, the application of innovative technologies in the entire production process. Such as AI, automatic optimization of the supply chain, online equipment sensors, etc., can effectively reduce waste and reduce product costs.
Wang Yimin introduced, “From the perspective of operation and management. It is completely possible to achieve zero investment and high return. If we lack of funds at a certain stage in the digitalization process, we cannot timely invest in digital software and hardware equipment. Then we should focus on operation and management and lean.”
The construction of intelligent factory in process industry will be carried out in three stages
The focus of intelligent factory production management in process industry is real-time transparent production process, operation standard management, full traceability. Wang Yimin mentioned that when understanding the status of smart factories, relevant evaluation results can be obtained through the evaluation model of smart factories. In terms of process design, especially for factories or small and medium-sized enterprises that have been put into production, process design is difficult to change. This requires some digital projects. For example, the smart factory can be made into a digital asset model through laser scanning 3D. And the key industrial devices can be made into a three-dimensional visualization model. The integration of manufacturing information, equipment information, factory information and other operation and maintenance information can be realized through management technology.
Wang Yimin introduced the key content of information technology from five aspects.
The first is the standardization of production operations. Including production process data standardization, operation process standardization, data-driven process collaborative optimization with standard specifications. The second is the advanced scheduling of plan management: to undertake ERP production plan. Which focuses on how to reduce the reject rate and improve the yield of finished products through EMS operation and production scheduling. The third is the fine coordination of the execution of part of the operations. The advanced scheduling based on the plan management can make the enterprise’s data acquisition application more timely and efficient. The fourth is the underlying collaborative real-time. The last is the visualization of production and manufacturing. Through the industrial brain, the digitalization and intelligence of the entire management and operation are realized.
The construction of smart factories can be divided into three development stages:. The first stage is digital and automatic control, including industrial sensor installation, data acquisition, etc. The second stage is networking, for example, how to network equipment, through the integration of hardware and software construction. So that the whole equipment data can be connected internally and externally. The third stage is intellectualization, including the application of digital twin technology, cloud computing big data and AI construction.
CPS is the essence of intelligent manufacturing
The most important point in building a smart factory is CPS, which is the essence of intelligent manufacturing. The application of industrial Internet and Internet of Things is the work that must be completed at the bottom of intelligent factory. Without CPS, intelligent manufacturing cannot be achieved. Secondly, business personnel’s self needs are the core driving force to promote smart factories. The third is value stream oriented optimization process. In the construction of intelligent factories, enterprises should not simply meet the task requirements of superiors. But should be value-oriented to achieve an effective balance of input-output.
In addition, AI technology should be used in business scenarios to reduce the use of human beings. In addition, we should pay attention to the training of compound talent team. Which not only requires us to have the ability to control and operate the production equipment of the production line. But also requires the staff to have the ability of data analysis and data modeling. At the same time, the support of top management is undoubtedly necessary in the whole implementation process.
Finally, “line with points, line with surface, small step fast.” When we don’t have a particularly large investment to ensure the implementation of digital technology. We can do some things in our own position. When we have some big digital investment, we should consider whether we can improve the digital supply chain, increase CPS input, and optimize intelligent production scheduling.
Wang Yimin concluded: “Investment in smart factories is not immediate or immediate results. But needs a certain amount of time and accumulation, everything from quantitative change to qualitative change takes time. Start small and keep improving through the flywheel effect.”
本文由数字化转型网（www.szhzxw.cn）转载而成，来源于网络；编辑/翻译：数字化转型网宁檬树。
免责声明: 本网站(http://www.szhzxw.cn/)内容主要来自原创、合作媒体供稿和第三方投稿，凡在本网站出现的信息，均仅供参考。本网站将尽力确保所提供信息的准确性及可靠性，但不保证有关资料的准确性及可靠性，读者在使用前请进一步核实，并对任何自主决定的行为负责。本网站对有关资料所引致的错误、不确或遗漏，概不负任何法律责任。
本网站刊载的所有内容(包括但不仅限文字、图片、LOGO、音频、视频、软件、程序等) 版权归原作者所有。任何单位或个人认为本网站中的内容可能涉嫌侵犯其知识产权或存在不实内容时，请及时通知本站，予以删除。
