The dream "meta universe" factory, what is the third-generation digital twin?

I believe many people have heard of the concept of "meta universe" that has recently become popular in different fields such as the Internet, technology, and investment. In various "PPTs", many people have made a very comprehensive and intuitive explanation of the concept of meta universe. In general, the most important point can be summarized in one sentence: the meta universe is a holographic digital world parallel to the traditional physical world.

Does it feel familiar? Especially, if you understand the Industrial Internet or the Industrial Internet of Things, there must be a concept that will immediately appear in your mind at this time: digital twin!

digital twin?

Digital twin is not a new concept. It has long been used in fields such as monitoring, simulation, and simplifying discrete device data. The powerful simulation capabilities of digital twins allow employees and intelligent systems to greatly shorten product design and manufacturing cycles through iterative collaboration, allowing companies to complete more product tests in a simulation environment, and improve them before physical manufacturing. , Thereby saving time and cost, making products closer to customer needs. In this way, companies can explore new product ideas with zero risk and expand the scene infinitely.

Gartner has listed digital twins as one of the top ten new technologies for three consecutive years (2017-2019). According to an IDC report, by the end of 2020, 65% of manufacturing companies will use digital twins to operate products and/or assets , 25% of enterprises hope to use digital twins to reduce the cost of quality defects and service delivery, which shows that digital twins have huge application prospects and space in the digital transformation of enterprises.

With the changes in market demand, digital twins have developed to the third stage. Both the first-generation and second-generation digital twins are tightly coupled with the equipment and cannot bridge the gap between production planning and execution. The third-generation digital twin starts from business needs and builds a common language between the physical world and the digital world, so that artificial intelligence not only completes the closed loop of perception, analysis, decision-making, and execution, but also through the understanding of production goals, when encountering emergencies In the event of an event, real-time feedback of the best route to deliver the order.

Today, we have ushered in the third generation of digital twins, represented by Microsoft's Digital Twin service, leading the third generation of digital twins. At this stage, not only the pursuit of business is traditional for the "people" do provide simulation analysis and decision support for digital twin, it is able to real-time perception of the physical world, as do multi-dimensional artificial intelligence, decision-making close to the implementation of online data analysis support , And return to the physical world to quickly execute the digital twin -through the digital twin to achieve the collection, visualization and contextualization of asset and project data, and then through machine learning to achieve analysis, prediction and empower frontline personnel to make real-time decisions-its core Value has expanded from simulation to decision-making and operation.

What problem does the third-generation digital twin solve?

According to IDC's analysis, 70% of manufacturing companies have begun to list digital transformation projects, including the business side, as their core strategic position; at the same time, 77% of CEOs regard rapid response as the core driving force for companies to gain competitive advantage.

In the past when supply was less than demand, the starting point of Industry 3.0 was to meet the demand for mass customization of a single product. In the past two decades, the market on which we are based has undergone earth-shaking changes. Consumers’ demand for products has become more and more personalized, which has transformed into more and more fragmented requirements on the production side; in recent years, due to the impact of the epidemic and geopolitics , The supply chain fluctuates severely, and the stable, large-scale, standardized production system established by Industry 3.0 is no longer able to cope with it, forcing manufacturing companies to accelerate the use of digital technology, improve the ability of agile response, and forge corporate resilience.

In short, smart manufacturing solves the problem from standardization to agile response, and the specific measure to improve agile response capability is to accelerate the closed loop of manufacturing operations management of "perception-analysis-decision-execution". In order to accelerate this closed-loop management, we need digital twins to solve the problems of "perception" and "execution", and artificial intelligence to solve the problems of "analysis" and "decision making".

application scenarios of 161a6de3309123?

The unit of measurement for "agile" is "response time". The response time starts from "event occurrence" and passes through the time required for "event learning", "root cause analysis", "decision-making" and "implementation". "Events" cover a wide range, such as the order status on the demand side and its changes, the material status on the supply chain side and its late events, the capacity status on the production side, and unplanned shutdowns.

principle, any field that can shorten the response time and improve the agile responsiveness of manufacturing enterprises through digital twins and artificial intelligence means is the application field of intelligent manufacturing.

No matter what plan the manufacturing company makes, it will eventually perform physical execution and deliver a physical product. There are four key elements involved: one is production resources, that is, people, machines, materials, etc.; the second is production content, that is, orders; the third is production technology, that is, manufacturing processes and routes; and the fourth is production management, which is The organizational structure of a workshop, factory, or enterprise.

We express the four elements of the physical world digitally to create a mirror world, or "meta universe" that artificial intelligence can understand, and make real-time analysis, prediction and decision-making through artificial intelligence.

In the context of smart manufacturing, the third-generation digital twin can function at three different levels:

• At the workshop level, it can realize "adaptive production execution" through "material-number fusion";
• At the enterprise level, it can optimize the dimensions of product quality, cost, delivery time, and safety and environment protection;
• At the industrial level, it helps to achieve dynamic synergy between upstream and downstream.

Here, we give an example of a food company to help everyone better understand the role of digital twins.

In the face of global supply chains and production plants, cost reduction and efficiency enhancement and sustainability are a difficult problem for many international production companies. Take a food company as an example. As long as there is a slight parameter deviation in the production equipment, it will cause too many or too little products in the retail package, too much, there will be non-quality costs, and too little will become a quality event. From a cost point of view, this phenomenon will cause a lot of labor consumption and material loss.

In order to solve this problem, Microsoft, Accenture, and Avanade cooperated to build a digital twin platform for smart manufacturing for the company. The platform integrates machine learning technology into the digital twin, uses a large amount of real production data to establish a production line model, and then finds the optimization point of the production line through the combination of the Internet of Things, edge computing and cloud computing, thereby greatly improving production efficiency and reducing labor and materials Waste. The company is currently promoting this digital factory solution globally.

How to plan and implement

Microsoft, together with its leading global partners Accenture and Avanade, is helping many leading companies build an intelligent digital twin that spans all elements of the organization, and is committed to creating factories and supply chains in the real world through the combined application of this technology , Smart twin network with consistent product life cycle.

At the technical level, the integration of big data, digital twins and artificial intelligence is the key to building an intelligent manufacturing technology platform. At the application level, how to promote change and manage risks is the focus of application implementation. We suggest adopting the change management approach of "global planning, small steps and fast running, large-scale promotion".

• overall plan: Industry 3.0 is standardization, and the enabling technology is process informationization and production automation; the core management concept of Industry 4.0 is agile response, and the enabling technology is digital twin and artificial intelligence. This puts forward brand-new requirements for the management thinking, management structure, organizational culture and talent reserve formed by manufacturing enterprises in the past two to three decades. Therefore, before launching specific projects, the company, from high-level to executive front-line personnel, needs to have a unified understanding of the concepts, methods and technologies of intelligent manufacturing.
• "Run in small steps" is the verification iteration . To a large extent, intelligent manufacturing is the process of achieving lean through digital and intelligent means, which not only eliminates the "waste behavior" that does not generate customer value in the manufacturing process, but also conducts pilot verification and verification based on minimum viable products (MVP). Continuous improvement, in just 10-15 weeks, with the shortest time and minimum cost, verify the feasibility of the concept, and make necessary iterative updates to it.
• "Quick promotion" is the large-scale application of . After verifying the feasibility of the concept, what companies need to do is to quickly implement large-scale applications and let the value of innovation radiate a wider range. This can be the replication and promotion between different production lines, or the upgrading of different business units, different factory companies, different upstream and downstream partners, and even domestic and overseas markets.

implementation case?

A global large-scale international fast-moving giant hopes to accelerate the end-to-end digital transformation in the two links of manufacturing and supply chain, so as to realize data-driven operational decision-making. Therefore, they have carried out the construction of digital factories for many years, but they still face the challenges of systemization, platformization and scenarioization.

Microsoft, Accenture, Avanade, and various customer departments have worked together to adopt the implementation method of "global planning, small steps, fast promotion", starting with a category of daily chemical products, and building a production process based on Microsoft Azure. Digital twin model, and the use of machine algorithms to optimize production processes and processes, to achieve two-way control of key assets and equipment. Due to the remarkable results of the pilot, we helped customers to promote the digital twin from a few factories to dozens of factories around the world in just one and a half years. group Culture.

Microsoft Azure Internet of Things Innovation Camp-Manufacturing Digital Transformation Special Session

We specially organized and arranged the Microsoft Azure IoT Innovation Camp. The purpose is to hope that more high-quality customers and partners can understand our products, use product value to solve current business problems, understand more usage scenarios, discover potential needs, and become better acquainted with product services and products through hands-on experiments. use.

On December 1, the first course of Microsoft Azure IoT technology hands-on training camp officially kicked off! We will focus on the application scenarios of the Internet of Things in the manufacturing industry: including factory intelligence, flexible supply chain, data-based predictive maintenance, digital twins, etc. We sincerely invite you to participate!

Time: December 1, 2021 | 09:00-16:00

Venue: Microsoft Artificial Intelligence and Internet of Things Lab (4th Floor, Building 19, Lane 55, Chuanhe Road, Pudong New Area, Shanghai)

(Author of this article: Zhang Simin, Managing Director of Accenture Greater China Industrial X Business and Accenture Greater China Industrial X Team; Microsoft Internet of Things Team; Avanade Smart Manufacturing Team.)