Industrial Engineering

Smart Manufacturing Builds Opportunities for ISEs

https://www.iise.org/iemagazine/2019-04/html/wuest/wuest.html

ISE Magazine May 2019 Volume: 51 Number: 4

By Thorsten Wuest

As an industrial and systems engineer, it is almost impossible to avoid conversations and discussions hovering around smart manufacturing and Industry 4.0. Whether one is active in an industrial or academic context, smart manufacturing seems to be omnipresent at every meeting, conference or trade show.

Here is a concise overview of smart manufacturing, its opportunities and challenges, and why ISEs are uniquely qualified to spear-head the charge to master the fourth industrial revolution.

It is important to understand Industry 4.0 and smart manufacturing. Both stem from the intelligent manufacturing paradigm of the late 1990s. The term “fourth industrial revolution” (aka Industry 4.0 or I4) was coined in 2011 by a German government initiative to prepare its manufacturing industry for the digital future. While the name is certainly controversial – since when are revolutions pro-claimed before they actually happen? – it rapidly caught traction and is now an established terminology across the globe (see accompanying article on page 43).

The Industry 4.0 paradigm builds on the notion that after the first three major industrial revolutions – mechanization of labor, division of labor and computerization of the shop floor – the convergence of the physical and virtual world, in form of so-called cyber-physical systems (CPS) is understood to be the next major interruption of the manufacturing industry and beyond.

The question remains: Was manufacturing not smart, or even “dumb” before? No, not at all. The smartness was however associated mainly with the human operators and process planners – the human experts – and not inherited in the system itself. We all know experienced operators of complex machine tools who seem to just “know” when a machine is about to fail or is experiencing problems. The goal in a smart manufacturing system is to reproduce this gut feeling by collecting data and analyzing it to draw conclusions, ultimately providing valuable insights for decision support.

The term “smart manufacturing” originated in the U.S. and is defined as “a data intensive application of information technology at the shop-floor level and above to enable intelligent, efficient and responsive operations” (“Panel on Enabling Smart Manufacturing,” Evan Wallace and Frank Riddick, Advances in Production Management Systems conference, 2013). While there are several more comprehensive definitions available, they all emphasize the use of information and communication technology data and advanced data analytics to improve manufacturing operations at all levels of the digital supply network.

An important aspect that differentiates smart manufacturing from many other initiatives is the specific emphasis on human ingenuity within the framework. Humans are not to be simply replaced by artificial intelligence and cognitive automation on the shop floor. Instead, their capabilities are to be enhanced by smart, customized solutions for the specific area. The importance of product and process information and data, enabling technologies and human or machine inherent knowledge is commonly accepted.

Smart manufacturing is often used in conjuncture with advanced manufacturing. At times, the terms are used as synonyms. However, this is not accurate and we need to be clear about the meaning of each term to avoid confusion. Smart and advanced manufacturing describe distinct areas of the new manufacturing realities and can be seen as two sides of the same Industry 4.0 medallion (as seen in Figure 2). Smart manufacturing at its core focuses on connectivity, virtualization and data utilization, while advanced manufacturing focuses on manufacturing process technologies such as automation, robotics and additive manufacturing. Nevertheless, there is no sharp dividing line between the two concepts, and in order to be successful in the future, companies need to embrace both.

Smart manufacturing technologies, characteristics, enabling factors

An important aspect of smart manufacturing for both research and industrial application revolves around associated technologies. For instance, it is important for companies to understand what technologies are relevant and might be worthwhile for assessment or investment. Meanwhile, researchers seek to identify areas where additional research is required to further develop technologies that will address current needs and pre-pare students for manufacturing careers.

The following is an introduction of the technologies, characteristics and enabling factors are associated with smart manufacturing. This is a summary of our recent journal publication regarding this topic (“Smart Manufacturing: Characteristics, Technologies and Enabling Factors,” Sameer Mittal, Muztoba Ahmad Khan, David Romero, Thorsten Wuest, 2017). If you are interested in a more in-depth analysis, visit https://link.iise.org/SmartManufacturing. In a comprehensive study, we identified 38 different technologies, 27 characteristics and seven enabling factors associated with smart manufacturing. The technologies ranged from machine learning to augmented reality; the characteristics from agility to decentralized control; and enabling factors from STEP standards to MTConnect. We then clustered the list based on their se-mantic similarity, as illustrated in Figure 3.

It has to be noted that additive manufacturing is included in this list despite the association with advanced manufacturing based on the methodology used. It showcases that some authors use smart and advanced manufacturing interchangeably.

Research opportunities for industrial and systems engineers

This section, discusses a selection of pressing research issues that requires the attention of the ISE community to successfully sup-port the manufacturing industry in its digital transformation. This is not a comprehensive list, as there are several more issues that deserve attention. However, the following challenges pose interesting and worthwhile problems for industrial and systems engineers to tackle.

Given the width of the field, the research issues are presented in three categories: technical research issues, methodological research issues and business case research issues (following “Industrie 4.0 and Smart Manufacturing – A Review of Research Issues and Application Examples,” Klaus-Dieter Thoben, Stefan Wiesner and Wuest, 2017). This again reflects the interdisciplinary nature and complexity of the field. Technical research issues include:

  • Standards and interfaces. To harvest the promise of smart manufacturing on the shop floor and beyond, a strong foundation that allows us to connect and communicate is key. Well-developed and widely accepted standards and interfaces are crucial to achieve this vision.
  • Sensors and actuators. With the dawn of the internet of things, sensors and actuators are everywhere, more powerful and cheaper than ever. However, in an industrial setting, the requirements for sensors and actuators are more rigorous than in our homes. Therefore, we need to continue to develop this field to provide the nodes for our growing network of manufacturing things.
  • Data quality. Data are the lifeblood of all smart manufacturing initiatives. We need new ways to assess and ideally guarantee a high quality of our data as an input for our advanced algorithms and decision-making.
  • Data handling. Developing new and expanding existing database systems and platforms is required. These systems must be capable of storing, retrieving and manipulating vast amount of data on premises, in the cloud and in-between (fog/edge) as well as variating combinations of these three approaches (hybrid systems), to facilitate collaboration and truly enable cross-domain learning.
  • Data analytics and machine learning. Deriving valuable insights from a vast amount of data (“big data”) re-quires continuous efforts in developing new and adapting current algorithms to optimize predictions, computing efficiency and ease-of-use, to name a few.
  • Data security/cybersecurity. Manufacturing facilities are among the most attacked entities, according to the U.S.

government. With increasing connectivity, the threat and potential damage of cyberattacks increases exponentially. It’s therefore necessary to invest efforts in developing new safeguards for industrial smart manufacturing systems that improve security while minimizing the negative effect on intended data exchange, sharing and flow.

Methodological research issues include:

  • Reference models. Smart manufacturing systems are inherently complex. This represents a significant entry barrier for many companies. Reference models providing a structure and guidelines to manage this complexity are needed, as well as new adaptations and extensions of existing ones for specific industries and special-use cases, such as small- and medium-sized enterprises.
  • Deriving insights from large amounts of data are only one side of the medallion. Communicating these insights in an appropriate, efficient and effective manner is equally important to create value and impact. For example, the C-level executive requires a very different visualization of the same data than the operator of a certain machine tool or the maintenance team. Visualization is strongly related to certain technologies such as digital twins, dashboards and virtual and augmented reality applications.
  • Services and applications marketplaces. Given the complexity of a smart manufacturing system, one common approach is to address the different functionalities through composable (micro-)services. Orchestrating those and creating efficient marketplaces to bring the various stakeholders together is a challenge that has yet to be fully addressed.
  • Requirements engineering. This remains a continuous issue for all engineering projects. With the dawn of smart manufacturing and Industry 4.0, the possibility to collect (through IoT) and analyze large amounts of usage data (big data) opens up new opportunities to derive insights in the real users’ needs and requirements directly from how they interact with the products and systems. New methods and ways to automate the translation of data and insights into design requirements need to be developed.
  • Operator 4.0. While certain tasks in the manufacturing environment will be increasingly automated, both physical and cognitive in nature, we believe the human operator will remain a key part of a smart manufacturing system. New ways to provide additional capabilities to the human operator are referred to as the tech-augmented “Operator 4.0.” Case studies and innovative solutions to extend the Operator 4.0 are in high demand.

Business case research issues include:

  • Smart manufacturing revolves around data col-lection, sharing and analysis. This introduces new challenges in the data privacy area, which is different from the data security aspect. The ethics behind sharing and analyzing user data, for example, need to be critically assessed as well as new rules and standards are needed.
  • Similar to most new developments that re-quire new technologies, redesign of processes and training, entering the smart manufacturing journey will require a significant investment. Especially for small or medium enterprises, this initial investment might pose a barrier as they are more likely to have limited resources, monetary and in expertise. Identifying ways to reduce this initial investment through, for example, new open source and modular solutions, will impact adoption in those cases. Collecting lessons learned and best practices from recent implementations, as well as case studies, will further lower the bar to engage in modernizing manufacturing operations.
  • Servitization and servitized business models. Servitization as a business strategy is a disruptive form of value (co-)creation. The availability of connectivity and real-time access to machine tool data enables the adoption of new business models based on pay-per-use or pay-per-outcome principles. For example, offering a complex machine tool as a product service system (PSS) provides multiple benefits to both the manufacturer of the machine tool as well as to the user. The manufacturer has access to the usage data as input for next generation designs, a continuous revenue stream and a closer relationship with its customers. At the same time, users benefit from reduced initial investments, reduced maintenance efforts and regular upgrades. While these theoretical benefits are very attractive, there are several issues to be figured out regarding these new business models, such as revenue sharing, data ownership, etc.

After covering the background of smart manufacturing and Industry 4.0, discussing associated technologies and enabling factors and identifying opportunities to advance the field, the question remains: Why are ISEs uniquely qualified to address the challenges put forth by the digital transformation of industry?

The answer is simple: This brave new world requires inter-disciplinary experts who are trained to think in systems, actually in systems of systems, and to deal with complexity both efficiently and effectively. Who is better at that than industrial and systems engineers?