Thorsten Wuest

Machine learning in manufacturing: advantages, challenges, and applications

The nature of manufacturing systems faces ever more complex, dynamic and at times even chaotic behaviors. In order to being able to satisfy the demand for high-quality products in an efficient manner, it is essential to utilize all means available. One area, which saw fast pace developments in terms of not only promising results but also usability, is machine learning. Promising an answer to many of the old and new challenges of manufacturing, machine learning is widely discussed by researchers and practitioners alike. However, the field is very broad and even confusing which presents a challenge and a barrier hindering wide application. Here, this paper contributes in presenting an overview of available machine learning techniques and structuring this rather complicated area. A special focus is laid on the potential benefit, and examples of successful applications in a manufacturing environment.

Cloud-Based Automated Design and Additive Manufacturing: A Usage Data-Enabled Paradigm Shift

Integration of sensors into various kinds of products and machines provides access to in-depth usage information as basis for product optimization. Presently, this large potential for more user-friendly and efficient products is not being realized because (a) sensor integration and thus usage information is not available on a large scale and (b) product optimization requires considerable efforts in terms of manpower and adaptation of production equipment. However, with the advent of cloud-based services and highly flexible additive manufacturing techniques, these obstacles are currently crumbling away at rapid pace. The present study explores the state of the art in gathering and evaluating product usage and life cycle data, additive manufacturing and sensor integration, automated design and cloud-based services in manufacturing. By joining and extrapolating development trends in these areas, it delimits the foundations of a manufacturing concept that will allow continuous and economically viable product optimization on a general, user group or individual user level. This projection is checked against three different application scenarios, each of which stresses different aspects of the underlying holistic concept. The following discussion identifies critical issues and research needs by adopting the relevant stakeholder perspectives.

Smart manufacturing: Characteristics, technologies and enabling factors:

The purpose of this article is to collect and structure the various characteristics, technologies and enabling factors available in the current body of knowledge that are associated with smart manufacturing. Eventually, it is expected that this selection of characteristics, technologies and enabling factors will help compare and distinguish other initiatives such as Industry 4.0, cyber-physical production systems, smart factory, intelligent manufacturing and advanced manufacturing, which are frequently used synonymously with smart manufacturing. The result of this article is a comprehensive list of such characteristics, technologies and enabling factors that are regularly associated with smart manufacturing. This article also considers principles of “semantic similarity” to establish the basis for a future smart manufacturing ontology, since it was found that many of the listed items show varying overlaps; therefore, certain characteristics and technologies are merged and/or clustered. This results in a set of five defining characteristics, 11 technologies and three enabling factors that are considered relevant for the smart manufacturing scope. This article then evaluates the derived structure by matching the characteristics and technology clusters of smart manufacturing with the design principles of Industry 4.0 and cyber-physical systems. The authors aim to provide a solid basis to start a broad and interdisciplinary discussion within the research and industrial community about the defining characteristics, technologies and enabling factors of smart manufacturing.

Can a Product Have a Facebook? A New Perspective on Product Avatars in Product Lifecycle Management

In today’s business environment, customer expectation towards product lifecycle information accessibility and quality is rising. Concepts such as the Internet of Things (IoT) respond to these demands enabling products becoming “intelligent” and capable of interaction. Simultaneously, society is changing with people spending more and more time online. Social networks allow them to interact richly with both their personal and professional contacts. Users of social network share information about their lives with a wide network of people who can respond directly. Being the most widely used social network Facebook’s functionality and usability have continuously evolved, culminating in the introduction of the timeline. The timeline is a representation of the users’ entire life, de facto managing the user’s “lifecycle” information. Considering the above developments, the question arises, whether it is feasible for a product to have a Facebook which acts as its product avatar, and whether that would contribute towards fulfilling the increasing customer demands towards product lifecycle information accessibility and quality?

Accessing servitisation potential of PLM data by applying the product avatar concept

Manufacturers of complex, high-value consumer products are increasingly forced to think of ways to satisfy their customers’ needs, stand out from competition and access new revenue streams. One way to accomplish that is to utilise the potential of product service bundles, which allow customers to enjoy a more holistic experience of products. The objective of this study was to investigate how the development of a novel concept, the product avatar and its application to existing products can contribute to servitisation. After establishing a solid foundation through a literature review, the theory behind the product avatar concept is introduced. In the following, the support of the product avatar concept for the main drivers of servitisation is discussed. The theoretical findings are then evaluated through a case study describing the development and application of the product avatar to leisure boat product lifecycle management data in the European boat industry. The results indicate that the product avatar concept does support servitisation by e.g. supporting industrial stakeholders to create new services around their core product which, in turn, may create new revenue opportunities.

Digital Representations of Intelligent Products: Product Avatar 2.0

Customer expectations towards products are constantly increasing. They are not limited to product quality alone but also include the accompanying services and information provided. Intelligent Products allow the retrieval and communication of large amounts of information from all stages of the product lifecycle. Customers have become used to user-centric presentation and customizable information presentation from their experience with the Web 2.0 and Social Networks. Implementing Product Avatars as parts of Social Networking Services (SNS) as digital representations of physical products allows the presentation of individually customized information in a familiar environment for different stakeholders. This can raise the acceptance and the lower the adoption threshold for Product Avatars by increasing their availability and usability on both stationary and mobile devices.

Towards Product Avatars Representing Middle-of-Life Information for Improving Design, Development and Manufacturing Processes

In today’s globalized world, customers increasingly expect physical products and related information of the highest quality. New developments bring the entire product lifecycle into focus. Accordingly, an emphasis must be placed upon the need to actively manage and share product lifecycle information. The so-called Product Avatar represents an interesting approach to administrate the communication between intelligent products and their stakeholders along the product lifecycle. After its initial introduction as a technical concept, the product avatar now revolves around the idea individualized digital counterparts as targeted digital representations of products enabling stakeholders to benefit from value-added services built on product lifecycle information generated and shared by Intelligent Products. In this paper, first the concept of using a Product Avatar representation of product lifecycle information to improve the first phases, namely design, development and manufacturing will be elaborated on. This will be followed by a real life example of a leisure boat manufacturer incorporating these principles to make the theoretical concept more feasible.

Comparing mining and manufacturing supply chain processes: challenges and requirements

The mineral raw materials industry is essential for manufacturing by providing the basic materials for their value adding processes. In the last decade, the integration of operations of the industry within global manufacturing supply chains has progressed greatly. The processes of the different stakeholders have been extensively analysed and modelled according to standardised frameworks such as the supply chain operations reference (SCOR) model. However, as of today, not all stakeholders are integrated to the same extent. Especially, in the early part of the supply chain, deep integration of the mineral raw materials industry is still an exception. This industry and its processes differ greatly from the average manufacturing company’s processes. Not being directly comparable results in the absence of applications of standardised modelling tools for supply chains like the above-mentioned SCOR model. These circumstances hinder the integration, understanding and exchange between industries that rely significantly on each other. In a first attempt to create a basis for further research, this study analyses, elaborates and compares the challenges and requirements of supply chain processes, with a special focus on sourcing processes, in manufacturing and mining. Based on these findings, an adaption of the SCOR model towards applicability in mining and mining supply processes is presented, followed by a critical discussion of the results and implications, later concluded by a short outlook on further research.

EXPERIMENTAL RESEARCH DATA QUALITY IN MATERIALS SCIENCE

In materials sciences, a large amount of research data is generated through a broad spectrum of different experiments. As of today, experimental research data including meta-data in materials science is often stored decentralized by the researcher(s) conducting the experiments without generally accepted standards on what and how to store data. The conducted research and experiments often involve a considerable investment from public funding agencies that desire the results to be made available in order to increase their impact. In order to achieve the goal of citable and (openly) accessible materials science experimental research data in the future, not only an adequate infrastructure needs to be established but the question of how to measure the quality of the experimental research data also to be addressed. In this publication, the authors identify requirements and challenges towards a systematic methodology to measure experimental research data quality prior to publication and derive different approaches on that basis. These methods are critically discussed and assessed by their contribution and limitations towards the set goals. Concluding, a combination of selected methods is presented as a systematic, functional and practical quality measurement and assurance approach for experimental research data in materials science with the goal of supporting the accessibility and dissemination of existing data sets.

Social Factory Architecture: Social Networking Services and Production Scenarios Through the Social Internet of Things, Services and People for the Social Operator 4.0

The prevailing industrial digitalisation flagship initiative, Industrie 4.0, gathers a substantial part of its functionality from the human in the system. This will drive a need for focus on both human and social dimensions of technology. The paper explores the roles of the Social Operator 4.0 in smart and social factory environments, where humans, machines and software systems will cooperate (socialise) in real-time to support manufacturing and services operations. A Social Factory Architecture based on adaptive, collaborative and intelligent multi-agent system is proposed for enabling such cooperation. Further, production scenarios are proposed, to show how social operators, social machines, and social software systems will communicate and cooperate via enterprise social networking services to accomplish production goals in the Social Internet of Things, Services and People.