pdj_1-2024_english

Editorial
News in brief
New members
prostep ivip Symposium 2024
Interview with Ulrich Ahle
Interview with Frank Möring
Interview with Jan-Steffen Kuhlmann
FAICE – An innovative approach to the collaborative use of AI in engineering
Making Artificial Intelligence Available for Use in PLM
Making effective use of digital twins in the product lifecycle
Integration of VDA 231-300 in DIN SPEC 91383
Multilateral Platform-Based Collaboration
Decentralized Data Spaces with Gaia-X
Events
Demonstrate presence in the important PLM-Journal!
Imprint

Content 3 prostep ivip product data journal 2024-1 Content 4 Editorial 6 News in brief New Members 13 Dentsu Soken Inc. 14 Material one 15 Microsoft Deutschland 16 Secondmind Symposium 18 prostep ivip Symposium 2024 28 Interview with Ulrich Ahle 31 Interview with Prof. Dr. Rainer Stark, Dr. Maurice Preidel, Alexander Palmisano 34 Interview with Frank Möring 38 Interview with Jan-Steffen Kuhlmann Projects 40 Revision of DIN SPEC 91383 — JT Industrial Application Package (JTIAP) 44 FAICE – An innovative approach to the collaborative use of AI in engineering 48 KI-Marketplace: How artificial intelligence is redefining engineering Experience 54 Making Artificial Intelligence Available for Use in PLM 60 Making effective use of digital twins in the product lifecycle 68 Integration of VDA 231-300 in DIN SPEC 91383 Science & Research 74 Multilateral Platform-Based Collaboration 80 Building the Future: Decentralized Data Spaces with Gaia-X 88 Events 89 Web Seminars 90 Demonstrate presence in the important PLM-Journal 94 Imprint Die DIN SPEC 91383 JT Industrielles Anwendungs paket wurde 2018 initial veröffentlicht und 2021 erstmalig überarbeitet. In dieser DIN SPEC wird der Nutzen von JT für die Verwendung und den Einsatz in der Anwendungs industrie dargestellt. Sie baut auf der ISO 14306 auf. Products and product/ service systems are increa singly being developed in value networks. Platforms, and in particular dataspa ces, offer promising approa ches to this multilateral collaboration. However, there are still technical, legal and other challenges to be overcome before these platforms are ready for largescale industrial use.

Editorial Editorial prostep ivip product data journal 2024-1 4 Embracing the Future of Product Data Management Dear Members and Colleagues, As we embark on another chapter of innovation and progress, it is my distinct pleasure to address you through the pages of the Product Data Journal. The prostep ivip association has always been at the forefront of product data management, driving forward-thinking solutions and setting industry standards. As Chairman, I am proud of the strides we have made and excited about the opportunities that lie ahead. The landscape of product data management is evolving rapidly, influenced by advancements in digital transformation, Industry 4.0, and the increasing integration of AI and machine learning into our processes. These technolo- gies are not just trends; they represent a fundamental shift in how we approach the lifecycle of products from conception through to decommis- sioning. Our association’s role in this transformation is pivotal, as we guide our members through the complexities and challenges of adopting these new technologies.

Editorial 5 prostep ivip product data journal 2024-1 This year, we have several key initiatives that I believe will significantly impact our community and the broader industry. Firstly, our focus on enhancing interoperability remains unwavering. The ability to seamlessly exchange data across different platforms and stakeholders is crucial for the efficiency and effectiveness of product development processes. Our work on stand- ardizing protocols and frameworks continues to pave the way for a more connected and cohesive industry. Secondly, we are committed to deepening our engagement with emerging technologies. Artificial Intelligence and Machine Learning offer unparalleled opportunities to optimize product data management. From predictive analytics to automated decision-making, these tools can transform how we handle data, improving accuracy, efficiency, and innovation. Our upcoming projects and workshops will delve into practical applications of these tech- nologies, providing our members with the knowledge and skills to leverage them effectively. Moreover, sustainability has become an integral part of our mission. The environmental impact of our industry cannot be overlooked, and it is our responsibility to integrate sustainable practices into our data management strategies. This includes everything from reducing the carbon footprint of our IT infrastructure to promoting the use of sustainable materials and processes in product development. By prioritizing sustainability, we are not only protecting our planet but also ensuring the long-term viability and success of our industry. Collaboration is the cornerstone of our success. the prostep ivip association thrives on the active participation and contributions of its members. Your expertise, feedback, and engagement are what drive our initiatives forward. I encourage all of you to take an active role in our projects, attend our events, and share your insights. Together, we can overcome challenges and seize opportunities that will shape the future of product data management. In conclusion, I am immensely optimistic about the future of our industry. The advancements we are making in interoperability, technology adoption, and sustainability are setting new benchmarks for excellence. As we continue to innovate and collaborate, I am confident that we will not only meet but exceed the expectations of our stakeholders and set new standards for product data management. Thank you for your continued support and dedication. Let us embrace the challenges and opportunities ahead with enthusiasm and a commitment to excellence. Sincerely, Henrik Weimer Dr. Henrik Weimer | AIRBUS Speaker of the board prostep ivip Association

News in brief 6 prostep ivip product data journal 2024-1 German Chapter of INCOSE (GfSE) started CASCaDE — joined standardization process with VDA and Prostep iViP @ OMG On 18.03.24, a small delegation of the German Chapter of INCOSE (GfSE) participated in the Object Management Group’s (OMG) quarterly technical meeting (TC), the Q1 Member Meeting, in Reston, Virginia near Washington D.C. These were the chairman of the GfSE Walter Koch, the representative of the GfSE in the OMG and member of the board of directors of OMG, Uwe Kaufmann, the representative of corporate members and association coop- eration, Christian von Holst, and our valued member and SysML v2 develop- er, Tim Weilkiens. A meeting of the “Manufacturing Technology & Industrial Systems Domain Task Force (ManTIS DTF)” was held on Monday morning. In this meeting, the joint project of the GfSE with the German Association of Automotive Industry (VDA) and the Prostep iViP for standardizing interlinked data, the current working name CASCaDE (Collaborative Advanced Specification, Content and Data Exchange), was presented. In this project, the two formats SpecIF of the GfSE and “Digital Data Packet (DDP)” of the prostep iViP are harmo- nized and merged into a common standard. This is to be developed and published by OMG (similar to ReqIF, SysML and others) and then also become ISO standard via the ISO PAS process. The goal is to have an inter- national standard available that allows the exchange of linked data between development partners across departments and organizations, regardless of the tool. The keywords here are “model-based collaboration” (e. g. MoSSEC) and “exchange of digital twins” (see “Collaborative Digital Twin”). This standard will also support other current activities relaying on connected data, like “long term archiving” (e. g. LOTAR) or supporting the “Code of PLM Open- ness (CPO)”. This initiative was presented at the Assembly of Members of the GfSE last November and is now entering the working phase. This is a novel approach of the GfSE to form strategic alliances with other important associations and also focus more on standardization activities and processes. The next steps here are the preparation of a so-called “Request for Proposal (RfP)” and its presentation at the OMG in the next meeting in Las Vegas in June. Later this year, a first submission of a harmonized standard is planned in the Q3 or at least the Q4 meeting of OMG. Overall, the project is planned to last three years in order to achieve a harmonized and published standard.

News in brief 7 prostep ivip product data journal 2024-1 Harmonization of SpecIF and DDP in new standard proposal CASCaDE Wiring System Data with Semantics / VEC now Published as an Ontology / Announce ment VECDay — “Bordnetze Digital” The Vehicle Electrical System Workflow Forum (VES-WF) and the ECAD Implementor Forum (ECAD-IF) groups address current issues and questions related to the digitization of the physical vehicle electrical system process chain. The automotive companies and suppliers involved in the VES-WF define the “Vehicle Electric Container” (VEC) as a complete digital product model and develop methods for cross-company, model-based develop- ment. ECAD-IF is the platform for system vendors. It supports vendors and users in implementing and testing solutions based on the above standards through implementation guidelines and joint test rounds. The activities of the VES-WF and the ECAD-IF are closely interlinked and in constant exchange with each other. This year, the VES-WF succeeded in adding a new facet to the applicability of the VEC model. In addition to the established XML schema, the VEC model will from now on also be published as an OWL 2 ontology with a corresponding SHACL schema. Publishing the VEC as an ontology creates new possibilities and offers several decisive advantages:

News in brief 8 prostep ivip product data journal 2024-1 1. Semantic interoperability: unlike the XML Schema representation, which defines syntax only, an ontology defines both syntax and semantics. This ensures that different systems can interpret the data in a consistent and meaningful way, leading to better data integration and understanding. The VEC model and its documentation contain a wealth of semantic rules and definitions that cannot be formally expressed in a purely syntactic definition such as the XML schema. By using ontologies, we aim to signif- icantly reduce the likelihood of dialects arising. 2. Data integration and linkage: ontologies and RDF facilitate the integration of data from different heterogeneous sources by providing a common framework for linking and merging distributed data. We believe that these are key factors in solving today’s challenges, such as the exchange of data in case of incremental changes, the support of distributed devel- opment systems or the creation of cross-domain digital threads. 3. Support for data rooms: industry trends and initiatives point to the cloud and data rooms. Ontologies are a key technology in this area, enabling more efficient sharing, integration and analysis of data or distributed data. The publication of the VEC as an ontology is our contribution to making data from the physical vehicle electrical system even more useful in the future. It should be noted that the OWL 2 ontology is being released in addition to the existing XML representation and is not a replacement. Both displays remain side-by-side, allowing users to choose the best option for their needs. Please do not hesitate to contact us with any questions, comments or feed- back you may have. This is the only way we can continue to refine and im- prove the VEC. Further insights into the digitization of the process chain of physical vehicle electrical systems, as well as opportunities for exchange and discussion, will be offered this year at “Bordnetze Digital”, the new home of the established VEC Day, which will be held for the first time at ARENA2036 in Stuttgart. You can expect an exciting and varied program covering all aspects of digitiza- tion in the wiring systems industry. We look forward to welcoming you! Registrations at: https://arena2036.de/de/bordnetze-digital-2024 Are you interested in, or have you become inspired by the idea of further digitizing the process chain of physical automotive electrics? Then contact us and become part of the project group! Contact: Philipp Kremer philipp.kremer@prostep.com

News in brief 9 prostep ivip product data journal 2024-1 prostep ivip Software Day — How to manage current embedded software development challenges Join us and be part of a new initiative to enhance software topics within the prostep ivip Association. We will explore the scope of key issues such as safety, security, MBSE, and virtual homologation to facilitate collaboration between OEMs and suppliers. Together, we aim to form a new expert group within prostep ivip to address the most relevant challenges in the software community such as n Software regulation challenges (e.g. UN ECE, ISO 27001, TISAX, ISO 26262, CVE, CVSS) n Innovative software architectures (CI/CD), n Collaboration and exchange (OEM, supplier) n Long Term Archiving (LTA) of software environments Register now to get in touch about the topics that interest you in this envi- ronment! Save the date on 8 October 2024 in Darmstadt and register now for participation. https://www.prostep.org/en/events/events Contact: Nora Tazir nora.tazir@prostep.com prostep ivip publishes JTIAP V3 In the JT Workflow Forum (JT-WF), both the possible areas of application and the resulting requirements for the JT format are further developed. The JT Industrial Application Package (JTIAP) consists of the specification of the JT format and the collection of use cases for JT. The publication of the new version of these documents reflects the latest developments in these areas. On the one hand, the use cases were reviewed and revised by the JT-WF, and on the other hand, the JT specification was adapted to JT version 10.5 Rev B. For both parts, the requirements arising from the VDA guideline VDA 231-300 for the exchange of material data were considered. During the revision, there was close cooperation with DIN, so that the changes made are consistent with the also revised DIN SPEC 91383. Contact: Raphael Becker raphael.becker@em.ag

News in brief 10 prostep ivip product data journal 2024-1 prostep ivip annual report 2023 available as an epaper and for download Even though prostep ivip events — and the annual symposium in particular — are often in the spotlight of the association’s activities, the actual back- bone is generally formed by the project groups and their work results. If you want to get an overview of the results, activities and highlights of 2023, take a close look at the prostep ivip annual report. It is available in German and English as an e-paper and can also be downloaded as a pdf. https://prostep.epaperpro.org/annualreport2023_english/ Further prostep ivip events: JT Day, STEP AP242 Day, Tokyo Systems Engineering Summit Further training, dialogue, networking — we offer all this and more at our numerous prostep ivip events. In addition to the new events already men- tioned, such as Software Day (8 October 2024 in Darmstadt) and Bordnetze Digital (26 September 2024 in Stuttgart, a joint event with ARENA2036), the following events will also take place in autumn n JT Day (23 October 2024 at the Fraunhofer Forum) hosted by Fraunhofer IPK, n STEP AP242 Day (13 November 2024 at ZAL Hamburg, hosted by AIR- BUS) and the 3rd Tokyo Systems Engineering Summit (27 November 2024). n Tokyo Systems Engineering Summit (27 November 2024, right — in Tokyo — with the kind support of NTT Data). Participation in Software Day, JT Day, STEP AP242 Day and Tokyo Systems Engineering Summit is free of charge — and open to both members and non-members of prostep ivip. Information, programme notes and of course the possibility to register can be found on our homepage. https://www.prostep.org/en/events/events

News in brief 11 prostep ivip product data journal 2024-1 Collaboration with and between Standards Standards are the foundation for successful collaboration within and across companies, but collaboration between standards is just as important to cover the entire digital thread. To achieve that several project groups met jointly in Darmstadt from 11-14 March. The LOTAR project used the face-to-face meetings to make progress on several new and updated documents of the EN/NAS 9300 series of stand- ard which will be submitted to the industry associations ASD and AIA for publication later this year. The LOTAR MBSE team used the opportunity for direct interaction with prostep ivip’s Digital Data Package (DDP) project, since an MBSE archive package and a DDP follow similar concepts. The CAx Interoperability Forum (CAx-IF) concluded its 53rd round of test- ing, which for the first time included broader testing of Persistent IDs. Such identifiers enable the tracking of changes in the model in use cases such as downstream communication as well as design collaboration. The highlight of the week was a joint workshop of all project groups work- ing on STEP AP242 XML. In addition to LOTAR, CAx, PDM, and EWIS (Elec- trical), this also included the JT projects. In total, 50 users and implementors from Japan, Europe, and the US discussed common scenarios for using STEP AP42 XML in combination with JT and STEP geometry to enable capa- bilities such as Kinematics, Assembly-level PMI, and Electrical Harness In- stallation. Previous roadblocks were resolved, and numerous pilot activities across the different project groups will follow later this year. All in all, the four-day meeting with up to five parallel sessions and more than 40 attendees on site, plus just as many joining remotely, was a great success and rendered many results that will drive project work for the next six months. The LOTAR and CAx-IF teams will meet again 16-20 September in Phoenix, AZ, USA. Contact: Jochen Boy, prostep ivip jochen.boy@prostep.com

New members 13 prostep ivip product data journal 2024-1 DENTSU SOKEN INC. Under its corporate vision of „HUMANOLOGY for the future: bringing people and technology together to shape the future,“ DENTSU SOKEN will coordinate the three functions of System Integration, Consulting and Think- tank to deal in earnest with society as a whole, including companies, government agencies, municipalities and consumers. In doing so, we aim to create a cycle that runs from the proposal of issues to their technology-driven solutions, supporting and realizing an evolution into a better society. DENTSU SOKEN will promote X Innovation beyond the boundaries of technologies, industries, companies and regions, harnessing the power of people and technology to pioneer the future and continually create new value. Kazuya Motoki General Manager Engineering Technology Development motoki.kazuya@dentsusoken.com Contact

New members 14 prostep ivip product data journal 2024-1 material.one material.one AG is part of the adesso Group, one of the leading IT service providers in German-speaking countries. Core sectors also include the automotive and manufacturing industries. adesso (FTE) currently employs more than 8,500 people. material.one is a supply chain collaboration platform for product and process approvals. The digital network helps global manufacturers and their suppliers collaborate in areas such as quality management, sustainability and product certifica- tion. material.one dissolves existing process and company boundaries and transfers company processes seamlessly into the supply network. With material.one, manufacturers, suppliers and companies from the TIC sector (testing, inspection, certification) sector simplify their collaboration. The consistent digitization of product requirements gives all participants in the supply network access to digital verification plans and proofs. Bernd Löhle CEO bernd.loehle@material.one www.material.one Contact

New members 15 prostep ivip product data journal 2024-1 Microsoft Deutschland Microsoft Deutschland GmbH was founded in 1983 as a subsidiary of Micro- soft Corporation (Redmond, U.S.A.). In Germany, we employ over 3,000 people at our seven locations in Berlin, Frankfurt, Hamburg, Cologne, Mu- nich, Stuttgart and Walldorf. Together with our 30,000 partners in Germany, we support companies with innovative solutions for the intelligent cloud and the intelligent edge so that they are successfully positioned for digital transformation. Furthermore, Microsoft is a leading global provider in many other areas such as productive software solutions, IT security, innovative hardware and development platforms, which are also based on open source technology. We connect more than 750 million people worldwide through our LinkedIn career network. With Xbox and Game Pass, we enable a cross-platform gaming experience. Moreover, Microsoft is involved in a va- riety of initiatives and projects together with politics, business and science so that everyone can participate in the progress of the digital society. Robert Lokner Director Automotive/EMEA Manufacturing & Mobility Industry robert.lokner@microsoft.com Contact

New members 16 prostep ivip product data journal 2024-1 Secondmind Secondmind exists to help automotive engineers design better cars faster, and achieve greater sustainability through machine learning. Secondmind is the optimization engine for the software-defined vehicle. From reducing design simulation time and calibration overhead, to contin- uously improving energy efficiency and performance throughout the vehicle lifecycle, Secondmind cloud-native optimization solutions give engineers the tools and confidence to re-imagine the design, development and expe- rience of driving, and accelerate the transition to carbon-neutral mobility. Secondmind was founded in 2016 and is backed by leading venture funds including Amadeus Capital, Atlantic Bridge, and Cambridge Innovation Capital, among others. Secondmind is the recent winner of the Autotech AI Breakthrough of the Year Award 2023. Gary Brotman CEO at Secondmind gary.brotman@secondmind.ai www.secondmind.ai Contact

prostep ivip Symposium 2024 19 prostep ivip product data journal 2024-1 This strategy is bearing fruit as this year’s symposium has shown. The event placed a firm focus on new topics such as artificial intelligence (AI), software-defined products, collaborative digital twins and dataspaces. It attracted some 700 participants from Germany, Europe, USA, Japan and Korea, approximately ten percent up on the previous year’s attendance. Not only was the public more international; there were also a refreshingly large number of new, younger faces alongside the “old hands” from the PLM family. The symposium was held under the motto “Digital Engineering – Ready for Seamless Collaboration”. With four keynote speeches, 98 presentations and seven workshops, the agenda of the two-day event at the MOC Event Center in Munich was even more packed than ever before. Those attending had to make a difficult choice between up to five parallel series of talks. There were four Expert Corners where they could meet recognized specialists and discuss topics such as the challenge posed by software-defined product development, the use of AI in engineering or the successful implementation of standards. Guided tours through the exhibition One new feature of this year’s event was the guided tours through the exhi- bition. These gave visitors an opportunity to get a feeling for the range of solutions available in various subject areas. The tours were well received by participants and exhibitors alike. The exhibition was fully booked and featured 33 software vendors and service providers. The exhibitors included newcomers such as Arorian, a company specializing in IT consulting and software development in the fields of PLM, ALM, IoT and SLM, as well as familiar names such as Accenture, Aras or CoreTechnologie that had already been present at the exhibition in the past.

prostep ivip Symposium 2024 20 prostep ivip product data journal 2024-1 “We hadn’t exhibited for the previous two years and had noticed that our market profile was slipping and that we were becoming less visible for our partners and service providers,” explains Andreas Grave, Director Regional Sales Central Europe at the PLM developer, Aras. “We revived many contacts here and the visit was definitely worthwhile for us.” A similar opinion was expressed by Daniel Spiess, Managing Director Engineering at Accenture, one of the world’s leading providers in the field of R&D transfor- mations. “For Accenture, taking part in the symposium is an unmissable opportunity to benefit from high-quality networking with customers and partners and to help actively shape the future of digital product develop- ment.” To make attendance at the symposium more attractive for younger people, the Association offered its second Start-Up Award for new enterprises. The candidates explained to the audience what they were all about in short three- minute presentations. There was Threedy GmbH with its instant3Dhub, which allows enterprises to use complex 3D data without the need for any extensive infrastructure for different 3D applications; KI-Marktplatz GmbH, which provides businesses with a platform where they can make available or search for AI applications, and the company Secondmind Ltd., which aims to speed up automotive development and simulation using machine learn- ing. Voting on the symposium’s app, the participants chose KI-Marktplatz as the clear winner of the award, which brings with it a prize of 2,500 euros. “Our platform is ready at the technical level and will be launched very soon,” explained Rik Rasor, one of the Managing Directors of the company, which has its origins in a research project. BMW and PTC as premium sponsors The premium sponsors of this year’s symposium were the carmaker BMW and PLM vendor PTC. In their keynote address on the art of digital engineering, Dr. Julien Hohenstein, VP Processes, Digitalization Governance Idea to Order at BMW, and Jan-Steffen Kuhlmann, VP Idea to Offer, Quality, offered some fascinating insights into the digitalization strategy that the automotive manufacturer is pursuing in order to speed up vehicle development, and particularly that of electric vehicles, in the coming years. According to Hohenstein, the aim is to use best-in-class practices to boost the speed of development without negative repercussions at the cost level. This means that there is considerable pressure on R&D.

prostep ivip Symposium 2024 21 prostep ivip product data journal 2024-1 As Kuhlmann explained, in the field of R&D, BMW is not only bringing IT closer to the specialist departments but is also modernizing its system land- scape. (See also the interview on page 38) In addition to the modernizing requirements management with Codebeamer, the list of digitalization measures is headed by the construction of a groupwide engineering platform, the introduction of a modern CAD system and the creation of an end-to-end data model for product structures and processes. More specifi- cally, the PRISMA PDM system is to be replaced by the Dassault Systèmes 3DExperience platform and the old, mainframe-based TAIS BOM manage- ment solution will give way to SAP IPPE. All of this will then run in an integrated engineering workplace user interface. Agile hardware and software development In addition to harmonizing the IT landscape, the use of AI plays an important role in boosting the efficiency of R&D processes, as Hohenstein pointed out. Potentials have been identified, for example in the fields of requirements checking, engineering searches, simulations, or software development and testing. Frank Möhring, VP Solutions Consulting CER at PTC, took up the baton and explained PTC’s AI strategy as well as a number of major use cases in the PLM environment, e.g. the capability to assess the impact of changes more accurately. At the same time, he reminded the audience that AI has to be used responsibly. Not everything that is technically possible is socially acceptable. In his keynote speech, Möhring took a look at the challenges of software- defined or, more accurately, software-dominated product development. He argued that although carmakers still have to take account of mechanical aspects, software has become the most important driver of innovation. According to Möhring, now is the time to take a fresh look at the develop- ment process and start developing both software and hardware in agile cycles. To achieve this, PTC’s next-generation automotive solution contains new modules, for example in the form of software from pure-systems that is intended to make the management of software variants more efficient (See also the interview on page 34). Focus on artificial intelligence AI was one of the topics that received the most attention at this year’s symposium. In the Expert Corner “AI in Engineering Data Hub”, Uwe Kloss (RLE), Professor Rainer Stark (Berlin Technical University) and Dr. Martin Strietzel (PROSTEP) discussed the obstacles standing in the way of AI use in industry, starting with the fact that most enterprises do not possess suffi- cient in-house data to support its implementation. The experts all agreed that AI is a collaborative issue and encouraged businesses to cooperate more with one another. The problem is that many of them have misgivings about sharing their data. It remains to be seen whether the availability of trustworthy dataspaces will change this attitude. The question of what happens when solutions are created with the help of generative AI was also discussed. According to Stark, what is needed is someone who is able to assess these solutions and choose between them.

prostep ivip Symposium 2024 22 prostep ivip product data journal 2024-1 What generative AI is already capable of became clear at the workshop entitled “From Requirements & Code to Road With Generative AI”. In his presentation at the start of the event, Chief Futurist Lars Thomson (future matters) told the audience that ever since it has been able to learn from us, the power of AI should not be underestimated. However, he preferred to speak not of “artificial” but rather “augmented” intelligence. In his opinion, we are still a few years away from having an “artificial assistant” with which we can communicate in natural language. Thomson predicted that AI will bring about a tenfold increase in productivity over the next ten years. In one workshop, representatives of a variety of companies, including Accenture, AVL, IAV, Microsoft, PTC and TÜV Süd, demonstrated the advan- tages of using AI, based on the example of an end-to-end process chain extending from the acquisition of the requirements with Codebeamer, their implementation in software code with GitHub and on to validation and homologation by TÜV Süd. Participants were impressed by the automated implementation of requirements in code form using GitHub Copilot. According to Steffen Linz (IAV), enlisting the help of AI makes it possible to achieve time savings of 50 percent in the requirements engineering cycle, including code generation and testing and on through to error analysis. To speed up the homologation process, TÜV Süd has worked together with Accenture and Microsoft to develop a GenAI-based assistant that promises to bring about further time savings. Early specification of training data For us to be able to use AI to develop and operate smart services, we need to have the right data as early as possible in the process. As part of his doctoral thesis, Dr. Maurice Preidel has developed a semantic approach that offers a more systematic way of specifying relevant data than is possible with conventional brainstorming. This ensures a high of quality of the training data and thus the AI models, avoids unnecessary iterations during the development of smart services, and reduces costs. It also simplifies collabo- ration between data scientists and subject specialists. To acknowledge his research, the Association awarded Preidel this year’s Scientific Award for the best doctoral thesis (See also the interview on page 31).

prostep ivip Symposium 2024 23 prostep ivip product data journal 2024-1 One of the first companies to report on its practical experience of using AI was MAZDA Motor Corp. Tomohiko Adachi (MAZDA) and Gary Brotman (Secondmind) explained how they have used AI to cut down the simulation work involved in engine development. To do this, MAZDA developed a model-based method for calibrating simulation data, which was then fine- tuned using Secondmind’s AI technology. To optimize the filling of the com- bustion chamber, it is necessary to compare and harmonize simulation and test data. As Adachi explained, with Secondmind’s help, it was possible to reduce the time required for this task by nearly 60 percent. MAZDA has a financial stake in the company that is being started up to optimize vehicle development. The benefits of trustworthy dataspaces The speakers in the Expert Corners as well as at the various presentations looked closely at the use and value of trustworthy dataspaces. According to Kuhlmann, BMW intends to move over from the need-to-know to the need- to-protect principle and provide more open access to its data, for example in Catena-X, in order to be able to interact with its suppliers. Nevertheless, Ulrich Ahle, CEO of Gaia-X, pointed out that there is as yet no readiness to accept such approaches in the core area of product development and called for a change of mindset. “Dataspaces make it possible to both share and protect data.” (see also the interview on page 28). In the opinion of Henrik Weimer, Director Digital Engineering at Airbus, who was present in the Expert Corner on the topic of software-defined products, the Gaia-X concept is particularly interesting in the context of model-based systems engineering (MBSE). However, he said that the various partners are not yet fully integrated in the process. And he went on to say that the chal- lenge when exchanging MBSE artifacts lies in protecting the know-how they embody. In the opinion of Dr. Juan Carlos Mendo, who gave a presentation on Boeing’s strategy for using MBSE standards, it is not sufficient just to exchange SysML diagrams if we want to improve interoperability in the BSE process. The American aircraft manufacturer is adopting a pragmatic approach that combines various standards for the exchange of the different types of model depending on the supplier level and use case. One of these is the Digital Data Package specified by the prostep ivip Association, which was presented in a workshop by Michael Kisch (em:engineering Methods) and Torsten Schmied.

Symposium 24 prostep ivip product data journal 2024-1 New forms of data exchange Trustworthy dataspaces permit new forms of data exchange and new use cases, as Dr. Marvin Hubl (SSC Services) and Yübo Wang (Mercedes-Benz) explained in another workshop. More precisely, the aim is no longer to exchange any documents but instead to synchronize data using the Eclipse Dataspace Connector (EDC). The new paradigm allows providers to retain sovereignty over their data and to control its use or impose a time limit on this. Sven Forte (CONTACT) presented two interesting industrial use cases in his contribution on the use of the Asset Administration Shell (AAS) for the management of engineering data. On the one hand, the mechanical engi- neering company and manufacturer of motor spindles, GMN, uses AAS to monitor IoT-capable spindles manufactured in-house and, on the other, it is able to offer the end users of its spindles services such as predictive mainte- nance based on AAS and the data acquired during operation, irrespective of the machines in which the spindles are installed. To this end, the company intends to collate the operating data from its spindles in a Gaia-X-compliant dataspace. Replacing PDM system at KnorrBremse This year’s symposium saw a whole range of presentations on the digitaliza- tion initiatives of companies from outside of the automotive and aviation industries. Franziska Freudenberger (Knorr-Bremse) and Markus Kelch (UNITY) presented a very large-scale project at Knorr-Bremse that has been running for more than six years and involves over 200 project participants. The aim of the PLM4Rail project is to replace the old PDM system, which is installed at 36 sites and accessed by more than 7,300 users, with the Wind- chill PLM solution. To do this, it is necessary to migrate 2.5 million parts, 6.1 million documents and 150 highly customer-specific integration adapta- tions. The greatest challenge lies in operating the two systems in parallel and ensuring their 1:1 synchronization.

Symposium 25 prostep ivip product data journal 2024-1 The dominant topic at last year’s symposium was sustainability. However, this was a less pervasive topic at this year’s event. Nevertheless, it is an issue to which BMW continues to attach considerable importance, as Hohenstein emphasized. He reported that the company met its climate aims for last year in terms of CO2 emissions per car and year. Mercedes-Benz is currently considering ways in which its IT organization can contribute to reducing the company’s CO2 emissions. Accounting for four percent of global emissions, the IT, communications and data industries are very much part of the problem, as Dr. Laura Casella (TWT) and Anke Wiemann (Mercedes-Benz) explained in their presentation on green innovation. The factors behind successful standardization Standardization is one of the prostep ivip Association’s key areas of compe- tence and a topic that is high up on the agenda of the Standardization Strategy Board (SSB). The specialists at one of the Expert Corners discussed what it is that contributes to the successful implementation of standards and the ways in which users and software vendors alike can benefit from this. Although the community needs the support of the vendors, it must also agree on the right standards and move more quickly, explained Dr. Frank Ramsak (BMW). At the same time, Dr. Sebastian Handschuh (Mercedes-Benz) refuted the idea that standardization takes too long and pointed out that the ISO has speeded up its processes. According to a survey conducted by the SSB, the greatest obstacles to standardization include the lack of supporting commercial IT tools as well as the lack of partners to put them to practical use. However, Professor Rainer Stark (Berlin Technical University) defended the vendors against the accusation that they would block any such move. He explained that they were faced with a dilemma arising from the fact that users are often unable to agree on uniform semantics. To conclude, Markus Krastel (em:engineering methods) stressed that engineering and IT standards are becoming ever more impor- tant for the synchronous collaboration of the future. He explained that there are enough standards. Crucial for their successful implementation in soft- ware solutions is the ability to link them and set clear priorities.

Symposium 26 prostep ivip product data journal 2024-1 The CDT project group starts its work In the presentation given by the Association’s new Collaborative Digital Twin (CDT) project group, Handschuh took this idea one step further and said that asynchronous data exchange is an obstacle to collaboration and that a more digitally-based, synchronous approach is needed. In the long term, it does not make any sense to send the entire telephone directory around to everyone every time a number changes. Consequently, the data must be linked at cross-enterprise level via the CDT. Peter Gerber (Schaeffler Technologies) and Dr. Sebastian Schweigert- Recksiek (em:engineering methods) presented the various work packages being addressed by the project group, which is working with various imple- mentation partners to develop a demonstrator for the CDT. To do this, it will cooperate closely with other prostep ivip project groups and external partners such as Catena-X, ISO and others. Its work will be channeled via various use cases, the first three of which are almost complete thanks to the preliminary work performed by earlier project groups. One of the most important of these is collaborative requirements engineering on the basis of directly linked requirements. Next prostep ivip Symposium in the BBC Berlin The next prostep ivip Symposium will be held 13 and 14 May 2025 in the BBC Berlin – with Dassault Systèmes as one of the premium sponsors. In his concluding keynote address, Dominic Kurtaz, Managing Director Eurocentral, gave the audience a small taste of what is awaiting them: The vision of a virtual twin that permits a glimpse into the past in order to get to grips with a “Future in Motion”. According to Kurtaz, the aim is to create a networked and personalized mobility experience. To achieve this, it will be necessary to master the complexities involved. This cannot be done using AI alone. To exploit the power of AI, it will be necessary to adopt a virtual-twin strategy.

Symposium 27 prostep ivip product data journal 2024-1 All photos by one click

Interview with Ulrich Ahle 29 prostep ivip product data journal 2024-1 Question: Mr. Ahle, you have been with Gaia-X for six months now. How is the initiative developing? Ahle: Things have developed very positively. Gaia-X has become a reality and the first data rooms are in productive operation. One of them is Agdatahub for the agricultural sector in France. Catena-X has also been productive since last October. We are increasingly generating new light- house projects using the Gaia-X trust protocol. We have also better aligned the organizations and projects involved in the creation of trusted data spaces in Europe, so that our activities complement each other to form an overall architecture. Question: Catena-X is very much focused on the supply chain. Where do you see potential for the creation of shared data spaces at the prostep ivip symposium? Ahle: In the end, product development is widely distributed along the value chain. Especially in the automotive industry, suppliers are responsible for a significant part of the development. When you create a digital mockup, you need parts from everyone, and you need to do calculations with your development partners. And we all know from experience over the last few decades how time-consuming it is to connect PLM systems and enable data exchange across company boundaries. This is where data rooms take us to the next level of productivity, making it much easier to collaborate across companies and connect with new partners. That‘s why it‘s very important to me that product development is taken into account when creating data rooms. This is currently not happening enough. Question: Are companies still reluctant to share their data in these data rooms? Ahle: Data rooms don‘t work without data, but in this area, product creation is no different from logistics or other areas. The key to data room acceptance is data security on the one hand and what we call data sovereignty on the other: that the producer of the data can determine and technically enforce what someone who has access to that data can and may do with it. This is a prerequisite for different companies and, in case of doubt, competitors to be willing to share their data in a common data room. This functionality exists. Therefore, this should no longer be an obstacle.

Symposium 30 prostep ivip product data journal 2024-1 Question: My understanding of Gaia-X is that you don‘t provide the infrastructure for these data rooms, just the rules for building them. Could this strategy change? Ahle: No, we‘re not going to change that. Some people said at the beginning that Gaia-X would be the next cloud hyperscaler. We are not and we don‘t want to be. Together with the community, we define the rules and standards for trusted data exchange, map them in a so-called trust frame- work and implement the whole thing in the Gaia-X Digital Clearing House. However, these clearinghouses are not operated by Gaia-X, but by partners from the community. Question: You mentioned Agdatahub and Catena-X as examples of func- tioning data rooms. What other data rooms are currently under construction? Ahle: In total, there are currently 180 other projects, 17 of which have managed to qualify as Lighthouse projects. In manufacturing, there is EuProGigant, a joint project between partners from Germany and Austria to create a production ecosystem, which is already quite advanced. A brand new project is Data4Industry-X, funded by the French government and involving companies such as Schneider Electric and automotive supplier Valeo. Question: What about Manufacturing-X? Ahle: Manufacturing-X is only the framework for various projects such as Factory-X or Aerospace-X. These projects are now underway. What makes this special is that countries from Europe and other continents have worked together from the beginning to enable global interoperability. Mr. Ahle, thank you very much for the interview.

Interview with Prof. Dr. Rainer Stark, Dr. Maurice Preidel, Alexander Palmisano 32 prostep ivip product data journal 2024-1 At the evening event held at this year‘s symposium, Prof. Dr. Rainer Stark, as chairman of the member group Academia, presented the Scientific Award for the best doctoral thesis, which is endowed with 4,000 euros, to Dr. Maurice Preidel from Berlin Technical University. The 1,000-euro award for the best master‘s thesis went to Alexander Palmisano from Graz Technical University. Together with the award winners, Stark discusses the importance of the awards in this interview. Question: Prof. Dr. Stark, how important are the Scientific Awards for the association? Stark: The key impetus they provide is twofold: First of all, new topics being addressed by the research community are made available to us in such a way that we can make them viable for use in industry in the form of solutions created by vendors. Secondly, we want to pave the way into the community for bright young talent who are familiar with these topics so that they can network and find potential employers. Question: Dr. Preidel, Mr. Palmisano, have you already found an employer? Preidel: I‘m currently in the process of setting up my own online education business and I am, of course, finding potential customers here. That is why I‘m delighted to have won the Scientific Award. Palmisano: I‘ve already found an employer. I‘ve ended up at Siemens in Graz. Question: How many of these Scientific Awards has the association presented so far? Stark: We introduced the Scientific Awards in 2012 and have presented them every year except one, so that must make it 24 or 25 awards. We usually have two or three contenders for each award but sometimes as many as four or five. Question: How did the two of you first find out about the association‘s Scientific Awards? Preidel: In my case, it was an obvious choice because Prof. Dr. Stark is my supervisor and I already have a long relationship with the prostep ivip Symposium. Polisano: And I found out about the awards from AVL, where I wrote my master‘s thesis. Question: Can you give us a brief summary of the subject matter that you researched in your master‘s thesis and doctoral thesis? Preidel: For my doctoral thesis, I developed a methodology that allows me to specify, at an early stage of the product development process, the data that I need to train AI for my smart services. I normally don‘t generate this data until the product has been built. I therefore have to think systematically about what I need early on in order to avoid expensive iterations later. Palmisano: The primary aim of my master‘s thesis was determining the health of battery management units for electric vehicles and combining two worlds in the process: the cell data, which is measured before the battery pack is developed, because you need to know its characteristics, and the data set generated by a fleet of vehicles. The aim was to improve predictions regarding the service life of the batteries based on cell and fleet data.

Interview with Prof. Dr. Rainer Stark, Dr. Maurice Preidel, Alexander Palmisano 33 prostep ivip product data journal 2024-1 Question: What is the main benefit of your work, Mr. Preidel? Preidel: The main benefit is that you can systematically specify the data that today tends to be generated in brainstorming workshops, where you are dependent on the people attending the workshops. The methodology allows you to achieve a higher level of process reliability and determine at an earlier stage whether or not the AI training process will be successful. Question: Your answer to the same question, Mr. Palmisano? Palmisano: The more precisely you can determine how healthy the battery is, the easier it is to reuse it in a second lifecycle. At the moment, battery management units are often based on estimations. In the future, there will be a move in the direction of data-driven machine learning. It will then be easier to adjust certain operating parameters with the aim of extending their service life. Question: Prof. Stark, what is the association doing to publicize the Scientific Awards to an even greater extent? Stark: We use a variety of different channels, from the association‘s website to LinkedIn. I also regularly remind my colleagues in the member group, which comprises universities and research institutions, to provide graduates with information early on because we want them to be the primary triggers. We are also thinking about whether we should publicize the awards directly via the universities, but the universities, of course, have numerous awards. Gentlemen, thank you very much for this interview. Mehr Informationen zum Scientific Award und Bewerbungen für 2025 unter https://www.prostep.org/ueber-uns/scientific-award

Interview with Frank Möring 35 prostep ivip product data journal 2024-1 Question: The title of your keynote was “Software-Defined Everything”. How well does PTC support software development? Möring: Product lifecycle management has actually always included managing software artifacts, but 20 or 30 years ago, software played an important but nevertheless subordinate role in the product lifecycle. Back then, people started incorporating compiled software in BOMs in order to establish a link between the hardware and the software. The acquisition of MKS Integrity meant that PTC’s portfolio included ALM software very early on, the aim being to expand PLM functionality to encompass requirements, test and source code management. Question: But the Integrity solution wasn’t integrated in Windchill until many years later. Does it actually still exist? Möring: Windchill RV&S, as it is called today, still exists and is an extremely robust tool available on the market, precisely because of the very close cou- pling with Windchill. We also offer a modeling tool that uses SysML to link

Interview with Frank Möring 36 prostep ivip product data journal 2024-1 software and hardware artifacts, thus mapping the V-model. The software can keep track of requirements management and test management across all domains. However, the fact that the amount of software in products is growing and requires agile development and online collaboration, we have further enhanced our position in the field of ALM by acquiring two more products, namely Codebeamer and pure::variants. Question: What are the particular strengths of these two tools? Möring: Codebeamer is a web-based technology for agile software devel- opment that enjoys a high level of acceptance among both developers and administrators. With pure::variants, we are expanding our range of solutions to include software variant management, i.e. making it possible to map soft- ware configurations to ensure that I, for example, can identify which vehicle in which model series has which software version on which control unit in which month of manufacture. This is a challenge that many carmakers have tried to tackle using software developed in-house. Question: How long will it take until the solutions are integrated to such an extent that they are able, for example, to exchange data between Code- beamer and pure::variants? Möring: There’s news to report on that front: Good interfaces between pure::variants and Codebeamer and between pure::variants and Windchill RV&S existed prior to the acquisition. We’re now working on improving integration of the tools, but we are also open to tools from other vendors. Question: What role does artificial intelligence, which you addressed in your keynote speech, play in the context of software development in the vehicle? Möring: Thanks to the progress made in recent years, artificial intelligence has the potential to offer significant improvements in software development in the very near term. Let me give you an example. When it comes to requirements management, complex tasks have to be mastered using primarily text, image and video data. This data is particularly well suited to achieving a higher level of process quality in less time using GenAI technology. Keep in mind that hundreds of thousands of requirements exist for a vehicle. These requirements not only have to meet all formal quality standards, they also have to be checked for conformity, for example. AI can already do this today with the help of large language models. The second point I want to make is that AI makes it easier to understand these large amounts of data. The user can interact with the data and, for example, ask a specific question like how many requirements exist for a certain seat shell.

Interview with Frank Möring 37 prostep ivip product data journal 2024-1 Question: Do you see possible applications for generating the actual soft- ware? Möring: Yes, of course. The symposium included a workshop on precisely this topic, during which participants were able to watch a complete cycle through the entire V-model, supported at every step by AI. What’s involved here is also using defined requirements to automatically generate code with Microsoft GitHub, for example. In other words, we’re also in the process of automating the code generation process. Question: Are AI tools already being used productively by your customers? Möring: Yes, they are. As part of our Industry 4.0 initiatives, we started implementing machine learning algorithms, which make it possible to pro- cess mass production data more precisely at a very early stage. Another area that we’ve been working on for a while now is the option of using generative AI to optimize the geometry of a component while giving due consideration to load or production-related aspects. This is a functionality that PTC’s CAD tool Creo offers, and one that is already being used by numerous customers. In the field of augmented reality, AI technologies are being used to recognize physical components, for example. This creates a link between the physical product on the one hand and its digital description on the other and makes it possible to map digital twins. Mr. Möring, thank you very much for this interview

Symposium 38 Interview with Jan-Steffen Kuhlmann, Vice President Idea to Offer, Quality at BMW Group IT AI is support for the engineer „The Art of Digital Engineering“ was the title of the keynote speech held by Jan-Steffen Kuhlmann, Vice President Idea to Offer, Quality at BMW Group IT, together with Dr. Julien Hohenstein, Vice President Processes, Digitalization, Governance, at the prostep ivip symposium. In this interview, he explains the role played by artificial intelligence. Question: Mr. Kuhlmann, is this your first time at the prostep ivip Symposium? Kuhlmann: Yes, it has to do with my new role. In July 2023, I assumed overall responsibility for IT in the R&D department. Prior to that, I spent five years working on the transformation of the BMW Group‘s IT infrastructure. Question: What is the idea behind the new digitization units at the BMW Group? Kuhlmann: All in all, this is a corporate initiative that has its origins in production and logistics. They have a digitization unit for modern technology that is directly geared to the respective process. Using this as a model, we have created a unit in each business division that is responsible for the use of new technologies in the respective process. We have also assigned process ownership governance to this unit because process flows will change massively as a result of digitization.

Interview with Jan-Steffen Kuhlmann 39 prostep ivip product data journal 2024-1 Question: How important is the software in the product for the BMW Group today and how is this reflected in the organization? Kuhlmann: Software development is fundamental to both in-vehicle digiti- zation and business processes. This is reflected in the fact that Group IT, together with our development division, has set up several IT & software hubs abroad to meet the growing demand. In Portugal, for example, we have a joint venture, CTW (Critical TechWorks), which now employs more than 2,500 software developers. We are also establishing a new joint venture in India this year and have set up another IT hub in China for software development in vehicles. Question: You talked a lot about AI in your presentation. Are these real applications? Kuhlmann: Yes, we already have many productive AI-based applications throughout our value chain, for example in the purchasing process or in engineering, where the internal image generation tool IMAGINE supports our designers. We have also implemented the first generative AI use cases, such as document validation and data queries based on natural language input, and our goal is to roll these out broadly this year. We are also working on AI support for specific engineering disciplines, such as AI-based prediction of simulation results. Question: Will we eventually get to the point where the engineer just enters the features and AI designs the entire car? Kuhlmann: AI has great potential and is already supporting our developers in their daily work. AI can help us digitize processes even further, and it can also guide our employees through processes, suggesting alternative solu- tions and helping them make decisions. However, we retain decision- making authority for ourselves and our employees, and we are committed to empowering all of our employees to use AI and apply it profitably. Question: What benefits do you expect from AI? Kuhlmann: We conducted an analysis of our processes earlier this year and identified significant potential that we plan to leverage over the next two years. For example, by introducing a co-pilot in software development, which will further shorten the vehicle development cycle. Together with our international IT hubs, we analyzed how GitHub Co-Pilot, for example, could support us. We have found that this already works very well for the common programming languages JAVA and Python. We can use the co-pilot to generate code, but we can also use it to generate tests and thus achieve higher code quality. Again, the next step is to deploy and scale across our departments. Mr. Kuhlmann, thank you very much for the interview.

Revision of DIN SPEC 91383 41 prostep ivip product data journal 2024-1 Motivation and relevance The scope of JT and the technical capabilities it offers have grown consider- ably in recent years in response to the application requirements of industry. Originally developed for visualization, and in the mean time for interopera- bility and long-term data storage, JT is increasingly being used as a neutral 3D data format for new applications. These are continuously taken into con- sideration by stakeholders from industry and international standardization organizations in order to protect investments made by industrial users, ven- dors and service providers. The capabilities, performance and openness of JT have made a remarkable contribution to its acceptance and dissemination in the IT landscapes in in- dustry. JT is actively used by millions of users worldwide in the automotive, aerospace and consumer goods industries alone, and this number is rising. The intensive use of JT for over a decade constantly leads to new ideas and requirements with regard to the format specification. Strategies, methods and applications have continued to make great strides in order to ensure interoperability, performance and consistency when processing JT data. The catalyst for updating the DIN SPEC prior to expiration of the 3-year deadline was new requirements from the VDA’s “Digital Materials Data Man- agement in the Prototyping Process” project group relating to the JT data format in the context of handling material and surface information. The pro- step ivip Association, acting as a stakeholder for industrial users and ven- dors, is addressing this need and has applied to DIN for an update of the DIN SPEC. The basic structure of the DIN SPEC has been retained. It is shown in the following section. Structure and content of the DIN SPEC With DIN SPEC 91383, representatives of the industrial users have devel- oped a comprehensive JT specification, referred to as JT Industrial Applica- tion Package, under the lead of the prostep ivip Association and the JT Workflow Forum. It contains both an adapted JT File Format Specification in document PSI 14-1, which is based on the JT Open File Format Description 10.5 (Rev. B), and a description of 38 industrial use cases in document PSI 14-2, the Comprehensive Collection of Industrial Use Cases. Unlike ISO 14306, which only covers 16 of these use cases, and primarily in the context of visualization, the DIN SPEC covers all use cases involving the use of JT as a process format that are relevant to industry. For use cases that require precise geometry, for example, it specifies the use of XT B-Rep for precise geometry as normative.

Revision of DIN SPEC 91383 42 prostep ivip product data journal 2024-1 Figure/ Abbildungen 1 Deutsch ISO Foundation (JT & STEP AP242) Scope DIN SPEC 91383 is an open, vendor-independent, state-of-the-art, license- free specification based on ISO 14306 (JT file format specification for 3D visualization) and ISO 10303-242 (STEP AP242 Domain Model XML), see Figure 1. ISO 14306 ISO 10303-242 Domain Model XML Extended scope Use cases published by prostep ivip/VDA JT Workflow Forum (PSI 14-2) Precise Geometry XT B-Rep mandatory for precise geometry (normative) Functionality Enhancement Performance JT Version Requirements to fulfill JT use cases Compatibility to JT version 10.5 More powerful publicly available LZMA Compression compression Figure 1: The modular concept for DIN SPEC 91383 — JT Industrial Application Package It provides clarity and investment protection for vendors in the context of making appropriate functionalities available to industrial users, which is reflected in the extensive range of solutions and tools available on the market. Revision of the DIN SPEC Figure 2: The members of the DIN SPEC 91383 update consortium After the prostep ivip Association officially requested an update of the DIN SPEC in September 2023, project groups from the prostep ivip Association and the VDA joined forces to create the consortium required for this purpose. The constituent meeting (kickoff) took place in January 2024. At the meeting, the representatives of Mercedes were elected to head up the consortium. This was followed by two work meetings held in February and March, at which the content to be updated was agreed on. Special focus was placed on updating the use cases. Obsolete use cases were removed

Figure/ Abbildungen 3 Englisch Revision of DIN SPEC 91383 prostep ivip product data journal 2024-1 43 Request submitted 1. Work meeting 2. Work meeting Signing group kickoff DIN SPEC release Preparation DIN approval processes Update content Release process July 23 Aug. 23 Sept. 23 Oct. 23 Nov. 23 Dec. 23 Jan. 24 Feb. 24 March 24 April 24 May 24 June 24 Figure 3: Timeline of the DIN SPEC update from the DIN SPEC. A new use case, “JT for Hybrid Design in Context for In- house Usage” was added. The “JT for Material Specification” use case has been revised in accordance with VDA 321-300, and the newly defined at- tributes for material and surface specifications have been added to the JT File Format Specification PSI 14-1. The revision of the content of the DIN SPEC was officially approved at the end of April with a unanimous vote by all the companies represented in the consortium. The consortium has thus completed its work. DIN will perform the editorial review, revision and approval of the new edition of the DIN SPEC as stipulated in the PAS procedure in May and June. The new edition will be published via DIN Media and will be available for use worldwide free of charge. Summary and outlook Collaboration within the consortium was characterized by the goal-oriented, constructive participation of all the parties involved with strong focus placed on results. Collaboration with DIN was also outstanding. We had a very com- mitted project management team that not only provided the consortium with reliable support in every phase of the PAS procedure but also in the event that someone became ill or was on vacation. This made it possible to update a document of key importance to industry, such as the DIN SPEC, within as little as nine months from application through to publication. In a world that is spinning ever faster, this is an enormous advantage compared to the time-consuming ISO processes. (By way of comparison: An update of ISO 14306 was launched in February 2022. The process is still ongoing and will not be brought to a close in 2024). DIN SPEC 91383 is the benchmark for the use of JT as a 3D process format in industry. It is the only document worldwide that covers all relevant JT- specific application scenarios for industry and vendors: visualization, inter- operability, long-term archiving and data exchange. Because of its crucial importance for industry, our next step is to turn the DIN SPEC into a DIN standard. Contact Bernd Watzal Mercedes-Benz AG bernd.watzal@mercedes-benz.com Contact Lars-Christian Bütow :em engineering methods AG lars-christian.buetow@em.ag

44 FAICE – An innovative approach to the collaborative use of AI in engineering Ruslan Bernijazov, Rik Rasor, AI Marketplace GmbH Rachel Bauer, prostep ivip Association Inspired by lively discussions at the symposiums held in 2023 and 2024 and the results of a survey conducted by the prostep ivip Association at the end of 2023 (see also the PDJ article entitled „Multilateral Collaboration on Platforms“), the „FAICE – Framework for AI-Enabled Collaborative Engineering“ project was officially launched on 7 May 2024 with a virtual kick-off meeting. With over 50 international participants from a wide range of companies and industries, the meeting indi- cated the community‘s great interest in the topic artifi- cial intelligence for collaborative engineering.

FAICE 45 prostep ivip product data journal 2024-1 The motivation for launching the project was the realization that the indus- trial application of AI in engineering is currently still subject to several chal- lenges that need to be addressed together. On the one hand, this concerns the large amount of high-quality, well-balanced data that is needed for training, fine-tuning and the efficient application of AI solutions and which cannot be generated by individual companies – even if they are large com- panies. On the other hand, AI solutions are currently being developed and used almost exclusively in-house, meaning that neither is the potential offered by shared data for improving results being fully exploited, nor can these AI solutions normally be used across different companies in collaboration scenarios. This means an increase in cost, time and effort for the individual companies and reduces the quality and scope of AI. Another key problem is how highly confidential data is handled in the product development pro- cess. Although contracts guarantee the protection of this data, how exactly it is used in AI applications is often unclear. Traceability and clear rules are therefore needed, particularly in the field of product development. If it is clear which data is used when, by whom, for what purpose and with which model, the efficient use of AI can be optimized. This transparency increases trust and makes the more efficient and targeted use of AI technologies in engineering possible. The FAICE project therefore focuses on creating a framework and standards for the secure collaborative use of AI solutions. Working together with the partner involved, existing technical solutions are adapted to the meet the needs of collaborative use cases, and solution modules are developed that benefit all association members. Figure 1: Use case for the first demonstrator

FAICE 46 prostep ivip product data journal 2024-1 The kick-off was preceded by the development of a first demonstrator based on an example use case at the end of 2023, see Figure 1. This meant examining a case involving interaction between OEM and supplier in which the fictitious OEM “OhmElectra” grants a specific supplier (“DriveControl”) targeted access to individual factory standards. A shared AI solution for checking specifications now checks the conformity of the specification against a variety of different standards and indicates any violations. Individual users are granted different permissions: some standards are publicly available (e.g. as part of laws), some are licensed documents, some are factory standards that can only be made available within the framework of a specific collaboration. A framework was developed that controls and deterministically specifies individually for each user what access to the corresponding standards a chatbot solution is granted for the purpose of comparison. The, to some extent, sensitive data (see factory standards) can therefore only be used by users with authorized access for the purpose of the automatic comparison of specifications with standards. This does not involve the data flowing into the AI model itself. It is only used for individual queries via a separate backend. This prevents unauthorized access via the AI model itself, even when methods like prompt injection are used. Figure 2: Concept for the framework for the collaborative use of AI in engineering

FAICE 47 prostep ivip product data journal 2024-1 In addition to presenting this first demonstrator and the associated example use case, additional possible use cases that would be worked on during the course of the project were also discussed at the project kick-off. Different examples of the collaborative use of AI in engineering were discussed and enriched with further content. For example: n Improving AI-based requirements management using shared data n Automatic generation of SysML models from natural language based on a common model repository n AI-based “translation aids” for generating semantic adapters that make it easier to integrate, for example, specifications and models across com- pany boundaries n Support when selecting relevant content for specific collaborations n Development of a chatbot for standards that makes targeted, relevant content available to individual users from an extensive pool of standards. It is planned that the exciting discussion on the proposed use cases be continued, the vision and mission of the project be defined, and the project goals and approach be specified in greater detail at an all-day, on-site work- shop to be held on 6 June 2024 in Darmstadt. All association members are cordially invited to participate, further develop the use cases or contribute new use cases, and work together to develop the framework for the secure, collaborative use of AI in engineering. If you are interested in participating or would like further information, please feel free to contact us. Contact Ruslan Bernijazov Co-Founder AI Marketplace GmbH ruslan.bernijazov@ki-marktplatz.com Contact Rik Rasor Co-Founder AI Marketplace GmbH rik.rasor@ki-marktplatz.com Contact Rachel Bauer Research & Innovation Manager prostep ivip e.V. rachel.bauer@prostep.org

How artificial intelligence is redefining engineering 49 prostep ivip product data journal 2024-1 The role of AI in product development Product development is a complex process that involves numerous steps from the initial idea to series production. Since the introduction of the German government’s national AI strategy in 2018 and the subsequent AI innovation competition, the role of AI in this area has changed dramatically. AI is now an essential part of modern development processes, significantly increasing efficiency and creativity in product development. From the idea to implementation: AI as a game changer AI has already established itself in many sectors and processes in industry, whether in the financial sector, transportation or logistics. However, data in engineering is different: it is often proprietary, available in specific tools, subject to frequent changes and has complex data formats. It would be wrong to assume that AI is primarily used to design and develop completely new products. Rather, the strengths of AI lie in the numerous facets and recurring tasks that require large, easily processable data sets. Very specific applications, such as the simulation of complex systems, are particularly exciting. In recent years, we have seen greater penetration and maturity among both providers and demand, not least because engineering IT is becoming increasingly data- and model-based with concepts such as PLM (Product Lifecycle Management) and MBSE (Model-Based Systems Engi- neering). Application examples along the Vmodel AI can be used in every phase of engineering to optimize processes and increase efficiency. Here are some application examples along the V-model: 1. Product planning: Market analysis is a decisive factor in product plan- ning. AI solutions can use natural language processing (NLP) and machine learning to help identify trends and competitive activities. One example is the use of NLP to analyze patents and scientific publications in order to identify technological trends at an early stage and derive strategic decisions from them. This leads to better positioning in the market and can secure long-term competitive advantages. 2. Requirements elicitation: Requirements elicitation is often a time-con- suming process that can be considerably simplified by AI. One specific use case is requirements specification analysis. Here, requirements are automat- ically extracted from extensive documents, classified and evaluated. This not only saves time, but also increases the precision and completeness of the requirements, which in turn forms the basis for a successful develop- ment process.

How artificial intelligence is redefining engineering 50 prostep ivip product data journal 2024-1 3. System design: The creation of MBSE system models (Model-Based Systems Engineering) and multi-domain models plays an important role in system design. These models help to understand the system behavior on a holistic level with different variants. SysML models, which as graphical mod- eling languages are also based on a representation in XML, are additionally extended by SysML v2 with a textual notation. Similar to software develop- ment, where large language models are already successfully generating code today, architecture models can also be created automatically in the future. Nearby applications are the derivation of models from requirements, checking models for semantic and syntactic consistency, the textual expla- nation of diagrams and models for people without modeling expertise, e.g. by using a system model as a formal knowledge representation for chat assistants. 4. Discipline-specific design: AI also offers numerous advantages for specific design in the areas of software, mechanics and electrics/electronics (E/E). One example is PCB (Printed Circuit Board) design, where AI algo- rithms can take over the routing and optimization of circuit diagrams. The interpretation of circuit diagrams and co-piloting in software development can also be significantly improved by AI. In mechanical design, generative design can be used to create and evaluate a large number of design alter- natives within a very short time. Another example in mechanics is intelligent part management. Such solutions make it possible to search through millions of components in seconds, significantly improving search efficiency. The rapid identification of already used or geometrically very similar parts can significantly reduce manufacturing, development and storage costs. In addition, the reuse of components in design allows resources to be used more efficiently and development times to be shortened. Figure 1: Examples of AI applications along the V-model

How artificial intelligence is redefining engineering 51 prostep ivip product data journal 2024-1 5. System integration and validation: In the critical phases of integration and validation of mechatronic systems, AI can make a significant contribution to increasing efficiency. AI assistance systems support the automation of error detection and test execution. By analyzing test reports and automati- cally detecting deviations, AI significantly improves quality assurance. A concrete example of the potential of AI in the field of software is the ability to automatically extract and categorize those results from a large amount of test data that indicate specific sources of error. This reduces manual effort and increases the overall efficiency of the test process. 6. Production planning: In production planning, the optimization of sequence planning plays a decisive role. AI can provide significant support here through the targeted analysis of current and historical production data as well as the consideration of framework conditions and multi-criteria objectives. An example from the automotive supply industry shows that the use of AI-based software has reduced production times by more than 25 percent. The solutions use ERP data, machine data and real-time data from production to suggest an optimized sequence for the production of orders. By using AI technology, conventional analog planning boards and spread- sheets can be successfully replaced, resulting in a drastic reduction in plan- ning time and a significant increase in production efficiency. Generative AI: more than just hype Current foundation models of generative AI already have a variety of use cases, from the automation of simple tasks to the generation of complex content. A good example of the application of generative AI in software development is copilots, which help programmers to write code or create documentation. This illustrates the potential of generative AI to provide support in highly specialized and technical areas. These models use large amounts of data from all over the internet, including texts, videos or git repositories. Generative AI also has the potential to create development artifacts such as CAD models and circuit diagrams in the future. In the field of engineering, however, CAD data, for example, is not public available, which makes the development of specific models challenging. Nevertheless, it is conceivable that such models will also be developed in the future. At present, the solutions are still in their infancy and artifacts are reminiscent of toy building blocks, but development is progressing and could open up new possibilities in the engineering sector in the near future (see Fig. 2).

How artificial intelligence is redefining engineering 52 prostep ivip product data journal 2024-1 Figure 2: Development of generative AI for development artefacts such as CAD – today and in the future The role of the developer One of the frequent questions in this context is whether AI will replace human developers. In fact, AI serves more as a tool to support developers (see Figure 3). Routine tasks such as searching for data and creating docu- mentation can be taken over by AI, giving developers more time for creative and complex tasks. This symbiosis makes it possible to combine the strengths of human intelligence with the efficiency of AI. This collaboration results in innovative solutions and products that would not be possible with- out the use of AI. The question is therefore not so much what kind of jobs in development will be replaced by AI, but rather what part or activity of each job can be supported by AI in the future. Demographic change and the associated shortage of skilled workers also require innovative solutions to preserve and pass on the knowledge of experienced employees. AI can help overcome these challenges by acting as a knowledge repository and mediator. The continuous development of the technology will make it possible to overcome these barriers and further improve the integration of AI into product development. Benefits of AI integration, challenges and future prospects The advantages of integrating AI into product development are manifold. One of the most important aspects is the reduction in time-to-market. Faster and more precise development processes mean that products can be brought to market more quickly. This is a decisive advantage, especially in highly competitive industries. In addition, the optimization of material costs through AI enables significant savings. Better planning and management of resources enables companies to work more efficiently and reduce their costs, resulting in higher profitability.

How artificial intelligence is redefining engineering 53 prostep ivip product data journal 2024-1 Figure 3: Symbolic image – AI assistants as a tool to support developers *inside Despite the many advantages, there are also challenges. The protection of company data and the need to train specific models for specific applica- tions are just some of the obstacles that need to be overcome. After some initial hype, we are now in the phase where productive solutions are being developed and implemented. In the coming years, we will see how these solutions prove themselves in practice and what real benefits they bring. Conclusion The integration of artificial intelligence into product development is more than just a passing trend. It offers companies the opportunity to optimize their processes, shorten time-to-market and boost the creativity of their developers. Despite the existing challenges, the benefits outweigh the risks. The future of product development is closely linked to the advancement and integration of AI technologies, and the companies that embrace these changes will be ahead of the competition. The key to success lies in the ability to use AI as a supporting tool that enhances human creativity and efficiency in equal measure. Contact Rik Rasor AI Marketplace GmbH rik.rasor@ki-marktplatz.com +49 5251 414 9960

Making Artificial Intelligence Available for Use in PLM 55 prostep ivip product data journal 2024-1 CONTACT intends to use AI to make its Enterprise Search even more powerful. (All images: CONTACT Software) Research into artificial intelligence (AI) is an exciting and rewarding field. CONTACT’s journey began with the development of its IoT portfolio for the smart factory and participation in joint projects on a variety of different use cases in the context of the product lifecycle. AI is a main area of research within the company ever since CONTACT Research was founded in 2022. The AI team started with initial work on Natural Language Processing (NLP), then came the hype surrounding ChatGPT and Co. This led to a plan to use a Large Language Model (LLM) in the next project in order to gain in-depth experience with this technology based on a specific use case. The aim was to code a prototype AI solution which, assuming the results were good, could then be integrated into CONTACT’s standard software. The idea A suitable AI topic was quickly found: semantic searches. With Enterprise Search, the CONTACT Elements platform offers full-text searching for all specialist applications that build on it, which delivers the desired infor- mation on products, projects and processes within seconds. This traditional keyword search works exceptionally well but is stretched to its limits when similar terms or whole sentences are involved. CONTACT Research intends to circumvent these restrictions with an LLM and expand Enterprise Search in such a way that customers are also able to use imprecise terms, synonyms and questions formulated as sentences when performing searches in future.

Making Artificial Intelligence Available for Use in PLM 56 prostep ivip product data journal 2024-1 Prior to its initial experiments with an LLM, the AI team first evaluated which data sources met the requirements in terms of quantity, quality, availability and security. In general, the following criteria need to be clarified for every AI project: n The data set must be large enough to obtain meaningful results – but not excessively large. This would require the acquisition of computing resources and lead to longer processing times at the start of the project, hindering rapid progress. n The data set must offer a sufficiently high input quality in order to be able to realistically evaluate the advantages and disadvantages of the tested methods. n It is a good idea to choose a data source for the initial approach that is either already available or easily accessible in order to speed up the solu- tion-finding process. n The protection of internal company know-how and personal data must also be taken into account. Depending on the tools and methods used to process the data, this cannot always be guaranteed, for example if data is processed on external servers. Either the data set has to be selected accordingly, or anonymization or encryption methods need to be used to exclude sensitive information. Last but not least, a limited project scope from the very start facilitates a step-by-step approach that delivers results as early as after the first attempts. These results are helpful on the road to the later AI application. CONTACT’s documentation portal serves as a data source for the LLM project.

Making Artificial Intelligence Available for Use in PLM 57 prostep ivip product data journal 2024-1 Taking all these criteria into account, a decision was made to use the official documentation for CONTACT’s software products as the data source. It is easily accessible, well maintained and only contains information that is avail- able to all customers and partners, which makes it non-sensitive in terms of confidentiality. The goal was clear from the beginning: everyone should be able to intui- tively access the knowledge stored in different formats, such as manuals, tutorials, specifications and examples – by simply asking for it regardless of the format. Implementation The semantic search complements the traditional keyword search by also finding results that are related to the search terms with regard to content. It includes a variety of methods such as the use of ontologies and thesauri, query expansions and semantic indexing. CONTACT Research first imple- mented semantic indexing. The AI team broke down the software documentation into short text excerpts called “chunks”, embedded them using an LLM and stored them in a vector database. Incoming search queries are embedded using the same LLM, and then the answers are found by performing a similarity search in the vector database. The results are output as a sorted list based on the similarity value. Each entry contains the text snippet and a link to the documentation so that the original source can be found easily. “The speed and quality with which our AI team achieved the initial results came as a pleasant surprise to us all,” says Dr. Udo Göbel, who is responsible for CONTACT’s AI strategy.

Making Artificial Intelligence Available for Use in PLM 58 prostep ivip product data journal 2024-1 The infrastructure On the way to a prototype solution, the team came across several aspects that it had not considered beforehand. The following are a few practical tips relating to the required infrastructure, which comprises an AI pipeline, vector database, computer capacities and database structure: n You can either code the processing pipeline for your LLM project yourself or make use of frameworks – depending on which option seems best suited for rapidly developing the respective AI application. n There are a large number of vector databases available. We chose a database that was easy to set up and use, and we initially used it from the cloud and later as a local installation. Most providers offer you both options. The test environment for the semantic search will be made available as a web service.

Making Artificial Intelligence Available for Use in PLM 59 prostep ivip product data journal 2024-1 n Normal laptops and desktop computers are not suitable for calculating embeddings for large amounts of text. You need a server with good com- puting power to carry out embeddings and storage space for hosting the database. We performed the calculations in-house and stored the embedded data with a cloud provider. This variant has the advantage that you maintain control over the data content, as the vectors are stored in the cloud without plain text. n We also planned the structure of the database ourselves and set up the tables with all the necessary fields. However, we only started doing this later on in the project and had to keep making adjustments. It is therefore a good idea to carefully consider the architecture in advance in order to avoid such issues. The results The results from our LLM project are so good that we were able to make a test environment available to selected users during the research phase. Meanwhile, planning has begun on how to best integrate the new AI application into the CONTACT Elements platform. Further information on this and other research projects can be found on the CON:RS blog at https://medium.com/contact-research/. Contact Leif Sabellek Senior Research Scientist CONTACT Software leif.sabellek@contact-software.com

Making effective use of digital twins in the product lifecycle 61 prostep ivip product data journal 2024-1 Currently, the various disciplines are all engaged in developing their own digital twins in order not to lose ground in the competitive field of digitali- zation. However, one thing that is often forgotten is the fact that all these different digital twins nevertheless still represent a product. This often leads to work being duplicated either because the various stakeholders are not aware of the twins that already exist or because these existing twins are sometimes too specialized to be of use for other applications. The aim of this article is to provide an overview of which digital twins can be found where in the product lifecycle and how these can be linked across disciplines. It ends by pointing out the identified benefits in order to pro- vide support to individuals and teams responsible for developing strategies in connection with digital twins. Making effective use of digital twins in the product lifecycle 1. The term “digital twin” 2. The general product lifecycle 3. Use of the digital twin 4. Benefits resulting from digital twin management The term “digital twin” Even though the term “digital twin” has been in constant use for years, its actual meaning depends greatly on the context. One generally accepted definition describes the digital twin as a virtual mapping of a product or machine. “Fraunhofer IOSB understands the term “digital twin” to refer to a concept in which products, as well as machines and their components, are modeled using digital tools, including all the geometric, kinematic and logic data. A digital twin is a map of the physical assets in a real factory and can be used to simulate, control and improve them. Industry 4.0 working groups dis- cuss digital twins in connection with the so-called asset administration shell and Industry 4.0 components.”

Making effective use of digital twins in the product lifecycle 62 prostep ivip product data journal 2024-1 A distinction is made between different types of digital twin depending on the application: n Component twins: map a component in a product/digital twin n Asset twins: describe a product or machine n System or unit twins: map a system or unit consisting of different assets n Process twins: describe the interaction between different twins over a de- fined timeframe The general product lifecycle The product lifecycle describes all the phases that a product passes through from the initial idea, to development through to disposal/market phase-out. The literature contains many different diagrams describing the product lifecycle. Intrinsic product lifecycle: Requirements, problems, corporate objectives → product planning → development, design → production, assembly, testing → marketing, con- sulting, sales → use, wear, repair → recycling, thermal recovery, landfill At the same time, the product is traced and monitored. Economic product lifecycle Consideration of a product’s market presence (Pahl, Beitz, Feldhausen, & Grote, 2007) t c u d o r P t n e m p o e v e D l r o f n o i t a r a p e r P Phase h Introduction c n u a L t e k r a M Growth Phase Maturity Stage Saturation Phase Deterioration Phase Sales volume Profit Sales volume Profit i l g n n n a P t c u d o r P Costs Losses

Making effective use of digital twins in the product lifecycle 63 prostep ivip product data journal 2024-1 Use cases Use Case Digital Twin Digital Thread Digital Shadow Lifecycle Phase Assessing the impact of changes Optimization of logical processes Commissioning Virtual demonstrations Quality control Predictive maintenance Usage monitoring Documentation of asset changes Global Service Building knowledge Virtual product representation Representation of the environment Asset monitoring Virtual training Virtual advertising Design optimization Prediction of service life Product as a service Close the quality loop Waste management Software updates OTA Usage simulations Optimizations in production Packaging optimizations Recycling simulations Digital maintenance 3D model Simulation model of changes Behavioral simulation model Product videos Sensor data VR experience Process model Logic model Asset shell Plant model AR experience Environment model Material Data AR / VR Data MBSE Data AR / VR Data MES Data CAx Data ERP Data CAx Data IoT Data ERP Data Asset Data Usage Data IoT Data Asset Data Usage Data IoT Data AR / VR Data AR / VR Data AR / VR Data Asset Data AR / VR Data Recycling Data CAM Data ERP Data Product development Manufacturing Sales Usage Repair / Maintenance Disposal / Recycling Source: ZWF Zeitschrift für wirtschaftlichen Fabrikbetrieb, downloaded from www.hanser-elibrary.com, Hanser-Library Use of the digital twin A digital twin can be used to perform a variety of different tasks. These can be simple tasks such as providing a digital description of an object or more demanding ones such as creating a logical map of a complex system. This breadth of requirements means that there will generally be different digital twins for one and the same object. This can be illustrated by taking a look at the entire product lifecycle. During product planning, digital twins play a less important role because they usually simply act as placeholders for potential products. The ways these products are described can vary dramatically from company to company. It is important to make sure that all the information that is collected during this early phase is also available for later use. The product development phase sees the emergence of many different digital twins that support a wide range of different use cases:

Making effective use of digital twins in the product lifecycle 64 prostep ivip product data journal 2024-1 The most widespread application for digital twins can be found in their use for different simulations. These can range from simple component strength calculations through to complex system simulations. They may give rise to different types of information: loads, behavior under certain boundary con- straints, optimization and many others. In addition to product development, simulations are also used in the production field in order to optimize product manufacture even before the first prototype is created. This allows the decisions needed to ensure cost-effective manufacturing to be made at the start of the product development process. For many companies, it is not yet standard practice to use digital twins during sales and operation. In these areas, AR is creating opportunities for customers and users to acquire experience before a product is manufac- tured This virtual acquisition of experience makes it easier for customers to decide which is the best product for them. It leads to increased customer satisfaction and consequently helps enhance customer loyalty. What is more, during the period when the product is in active use, digital twins can help minimize downtimes and speed up maintenance work. This is possible thanks to the simulation of, for example, deterioration parameters based on the real behavior of the product in use. In addition, upcoming maintenance operations can be planned and simu- lated in detail based on a digital twin, allowing service personnel to equip themselves with the necessary tools and spare parts in advance and prepare as best possible for the maintenance work. Here again, the work carried out is digitally documented and complex tasks are supported by AR. When a product reaches the end of its lifecycle, potential alternative uses or optimum disposal can be planned and simulated based on a digital twin.

Making effective use of digital twins in the product lifecycle 65 prostep ivip product data journal 2024-1 Benefits resulting from digital twin management Given the various uses for digital twins described above, it can be seen that it is not possible to cover all requirements using just a single twin. That is why companies must think long and hard about how they will create and manage their digital twins very early on in the product development pro- cess. During the development phase, it is obvious that the digital twins will be linked with information from PLM systems. It is important that the links are assigned to specifically identified versions to ensure that design changes are not linked to the wrong twin. Currently, some companies employ a Digital Twin Solutions Architect, whose task it is to create the various digital twins. However, this role usually has a reactive mandate and the twins are not created until a concrete request is received. Although this ensures that the best digital twin is always created, it also causes avoidable delays in the development process. Some compa- nies, however, specify in their development manuals which digital twins are to be created at what time and consequently make sure that the digital twins are always available. This approach can be supported by workflows in PLM systems. If other digital twins are needed, then they can be created by the relevant experts upon request. This is only possible if all the necessary information is accessible and easy to find. In such cases, a PLM system based on the use of parts numbers has proved to be an excellent tool. Here, users can find all the available infor mation based on the parts numbers and use this to create the required digital twin. It should be noted here that ALM and software development are considered to be components of the PLM system. This makes it possible to retrieve all the necessary information from the ALM and software development fields based on a material or parts number and then make effective use of this.

Making effective use of digital twins in the product lifecycle 66 prostep ivip product data journal 2024-1 Production is the point at which various digital twins come together. On the one hand, these may be created by the company itself, e.g. for products or equipment and tools developed in-house, and, on the other, they may come from external suppliers and partners (e.g. a processing center). If companies want to manage their production development and production process definitions in a PLM system, then the procedure is the same as during product development. Otherwise, they will need to choose a system that makes it possible to link the digital twins to the manufacturing data and the products. This makes sure that these can be used if the product is modified and that any required adaptations to the twins are identified at an early stage. In the past, the question of managing digital twins relating to already supplied products has received only inadequate attention and the corresponding twins have usually been stored as standalone objects without links. Thanks to the improving quality of digital twins and their area of application, the concept known as AAS (asset administrative shell) has been developed. Organizations such as the IDTA have taken up the question and are devel- oping guidelines and standards regarding the ways in which the informa- tion defining digital twins can be stored in standardized form. As a result, it is now also possible to link digital twins to one another and thus develop new business models. This means that not only the digital twin of a deliv- ered machine but also the digital twins of its components can easily be called up to check whether or not a spare part from another supplier can be used. This information is also important to companies that are responsible for disposing of products. Because the twins cover all lifecycle phases, detailed information can be accessed quickly and makes efficient, environ- mentally responsible disposal possible. Nevertheless, one very important task that remains to be resolved in this area is that of protecting company know-how. It is important to create digital twins with different details so that, for example, an end user only receives information that is necessary for operation and maintenance, whereas a service partner receives more detailed information allowing it to adapt its services optimally to customers’ requirements. The manufacturer has the complete digital twin and can use this to continuously improve its products and harmonize its portfolio with the specific needs of its user base.

Making effective use of digital twins in the product lifecycle 67 prostep ivip product data journal 2024-1 It should be noted that digital twins have significant benefits during every product lifecycle phase, both for the companies that use them and for their customers. As a result, companies that do not make extensive use of digital twins will be faced with ever more economic challenges in the future. At the same time, however, there is a danger that companies will lose their way among the multitude of digital twins and invest so much time in their creation and maintenance that they will no longer be able to operate economically. That is why it is important for them to develop a strategy for using digital twins from which they can then derive the necessary management processes. In this way, it will be possible to make cost-effective use of digital twins by improving existing products and exploring new business models. One of the first steps on this path is to ensure data continuity across system boundaries. This can be achieved using TTPSC’s connectedPLM concept. In this, the data from a wide variety of different systems is read and processed in real-time. User-specific applications not only make it possible to link and retrieve relevant data but also help simplify complex workflows and create certain digital twins on the basis of standards and guidelines. Wolfgang Rüdell Principal Business Analyst at Transition Technologies PSC wolfgang.rudell@ttpsc.com Contact

68 Digitalization of the product engineering process: Integration of VDA 231300 in DIN SPEC 91383 Jonas Rothe, Director Business Development, material.one AG Ongoing digitalization is having an impact on all areas of industrial production and, in particular, on the product engineering process. An important step in this context is the integration of VDA Recommendation 231-300 in DIN SPEC 91383 – a step towards a higher level of standardization and the digitalized future of the product engineering process. This makes an important contribution to increasing efficiency and shortening the time-to-market in the development/innovation process (not only) in the automotive industry.

Integration of VDA 231-300 in DIN SPEC 91383 69 prostep ivip product data journal 2024-1 Futureoriented standards and digitalization in the product engineering process (PEP) VDA Recommendation 231-300, which was published by the German Auto- motive Industry Association (VDA), provides the foundation for standard- ized data exchange in the context of material sampling. Its integration in DIN SPEC 91383 puts this standard on a broader and more binding footing. The CAD-neutral JT data format, which is specified in accordance with ISO 14306, plays a key role here. This format makes it possible to clearly describe and embed material and surface requirements in 3D datasets, resulting in improved collaboration along the entire supply chain. Importance of endtoend digitalization for entry onto the market End-to-end digitalization of the PEP is vital when it comes to increasing speed and innovative power for market entry. The precise and efficient exchange of data is essential in a global market in which development processes are often distributed across multiple companies and ever-more requirements need to be met. Increasing the level of digitalization also serves to improve transparency and traceability in the supply chain – also because of new environmental requirements for materials (e.g. proportion of secondary materials, recyclate, etc.). The use of digital tools and stand- ards means that companies can ensure that everyone involved has access to the same, up-to-date information. This reduces errors and avoids misunder- standings, thus contributing to quality assurance. Figure 1: Supply chain collaboration networks standardize data exchange and support networking between all participants in the supply network

Integration of VDA 231-300 in DIN SPEC 91383 70 prostep ivip product data journal 2024-1 Secure data exchange via industry cloud networks The exchange of standardized data in the PEP ideally takes place via industry cloud networks that guarantee data security and integrity. These cloud soft- ware solutions provide a secure infrastructure for accessing and using material and product data. They make it possible those involved to commu- nicate and exchange data productively and securely, which in turn increases the efficiency and quality of collaboration along the entire supply chain. Information from documents that are available in different file formats (e.g. pdf or xlsx), for example, is systematically digitalized to ensure that the data can be used securely via automated processes in other data exchange operations. Practical application: material and surface information in the JT format A concrete example of the application of VDA 231-300 is the coding of material and surface information in the JT format. This information is attached to the respective bodies in the 3D dataset as attributes and is thus available to all downstream processes. This makes both seamless inte- gration and use of the data along the entire supply chain possible. A clear and concise use case is the exchange of material information and surface qualities as attributes in the JT format, which enables standardized and more efficient collaboration. Figure 2: Example of a property field in the JT ToGo Viewer Source: VDA

Integration of VDA 231-300 in DIN SPEC 91383 71 prostep ivip product data journal 2024-1 Standardizing material and surface information in JT files means that every- one involved in the supply chain has access to consistent and accurate data. This increases both effectiveness and accuracy in the development process and also ensures that the correct materials are used. Use of material databases during development Design engineers can access a material database containing standardized construction materials during development. This data is integrated in the JT file. The most important information however is the ID, which allows the dataset in the database to be identified. This ensures that everyone along in the supply chain has access to consistent and accurate information. This method promotes efficiency and accuracy in the development process. It also makes certain that the correct materials are used. Figure 3: Illustration of the process based on a supply chain collaboration platform Source: material.one AG; ©infografik.pro

Integration of VDA 231-300 in DIN SPEC 91383 72 prostep ivip product data journal 2024-1 Support for subcontractors and automated inspection plans Subcontractors also benefit from this standardization, as they receive the information they need for planning and manufacturing the components. Access to the material database allows them to view the data required for the sampling and approval process. Ideally, the material database is connected to a requirements database for industry standards and the factory standard of the OEM. Cloud-based software platforms like material.one support this process by automatically creating Design Verification Plan and Reports (DVP&R) with detailed inspection plans. This concept works along the entire supply chain and also includes Testing, Inspection and Certifi- cation (TIC) service providers to ensure the integrity of the information. The integration of VDA 231-300 in DIN SPEC 91383 represents significant progress in the digitalization of the PEP. Standardization and digital data exchange not only increase efficiency and speed but also improve the quality and reliability of products along the entire supply chain. Figure 4: Screenshot of the product release process in accordance with VDA 231-300: Components in tree structure with different inspection requirements Source: material.one AG

Integration of VDA 231-300 in DIN SPEC 91383 73 prostep ivip product data journal 2024-1 Conclusion The integration of VDA 231-300 in DIN SPEC 91383 marks an important step forward in the digitalization and standardization of the PEP. The CAD-neutral JT data format enables the precise exchange of material and surface information, which significantly improves efficiency and quality along the supply chain. Secure data exchange via industry cloud networks and the use of material databases enhance the consistency and accuracy of the information. Sub- contractors benefit from standardization as it provides them with access to the information they need and automated inspection plans with the support of platforms like material.one. Integration increases overall efficiency, quality and sustainability in the PEP and paves the way for an innovative and competitive future in the auto- motive industry. Jonas Rothe Director Sales & Business Development material.one AG jonas.rothe@material.one +49 177 7954595 Contact

74 Multilateral PlatformBased Collaboration Rachel Bauer, prostep ivip Association Prof. Dr.-Ing. Stephan Husung, Technische Universität Ilmenau Products and product/service systems are increasingly being developed in value creation networks. Plat- forms, and in particular data spaces, offer promising approaches to this multilateral collaboration. However, there are still technical, legal and other challenges to be overcome before these platforms are ready for large-scale industrial use. More research is needed in order to identify the challenges and opportunities of multilateral, platform-based collaboration in greater detail. A preliminary introduction is provided by a survey conducted by the prostep ivip Association and the Product and Systems Engineering Group at the Technische Universität Ilmenau.

Multilateral Platform-Based Collaboration 75 prostep ivip product data journal 2024-1 Motivation Technical products and the associated services are increasingly being developed, produced and made available in value creation networks that bring together different development partners. On average, suppliers account for 50% of value added, in the automotive industry even at 78% (see “Lieferantenentwicklung: Stand der empirischen Forschung” [Supplier development: Status of empirical research] by Durst and Sucky in: Bogaschewsky et al, Supply Management Research, Gabler 2010). At the same time, products are becoming ever more complex. In addition to the development of mechatronic systems towards cybertronic systems and products that are integrated into the environment as part of a ‘system of systems’ and rely on targeted interaction with the surrounding systems to fulfil stakeholder needs, an increasing proportion of product-related services can also be observed. Products are thus becoming product-service systems (PSS, see also “Innovation in Product-Service System Engineering Based on Early Customer Integration and Prototyping” by Exner, Damerau and Stark, published in Procedia CIRP Volume 47, 2016). The development of complex PSS in networks involving multiple develop- ment partners demands intensive, trust-based collaboration between the participating partners. Among other things, value creation networks such as these are obliged to make use of extensive, heterogeneous, distributed data and the participants must be extremely dynamic in the way they approach projects in order, for example, to increase product maturity and allow for the targeted integration of development partners. In the early phases of product development, it is essential to ensure smooth collaboration both between different internal departments and between development partners active across different enterprises. The decisions taken during this phase will significantly determine the quality and the environ- mental and economic characteristics of the future PSS. The necessity and challenges of setting the strategic course at an early stage and, above all, the effort required for any necessary corrections in later phases of life are often illustrated by the concept of the “Rule of Ten” (see, for example, Ehrlenspiel et al, “Kostengünstig Entwickeln und Konstruieren” [Cost-effec- tive development and design engineering], Springer Verlag 2020). Development decisions must also be made based on valid information. In the context of value creation networks, it is necessary to make this informa- tion available and use it securely via a consistent, shared database. This shared database must not only allow the exchange of information between different development partners but also between different phases in the product lifecycle (see also Figure 1).

Multilateral Platform-Based Collaboration 76 prostep ivip product data journal 2024-1 Platform-based collaboration approaches are becoming increasingly important for the industry in order to handle this complexity (see also “Collaborative Model-Based Systems Engineering Using Dataspaces and SysML v2” by Li, Faheem and Husung, published in Systems 12, 2024). Data- space-based initiatives such as Gaia-X, Catena-X and Manufacturing-X in particular offer promising approaches. However, there are currently still numerous uncertainties and technical barriers that complicate or prevent a comprehensive transition of value creation processes to corresponding collaboration platforms, particularly in the area of product development. In order to analyse the challenges and future potential of platform-based product development in more detail, the prostep ivip association, together with the Product and Systems Engineering group at the TU Ilmenau, devel- oped a survey that was distributed to the members of the prostep ivip asso- ciation. The aim was to identify the challenges, opportunities and risks of platform-based collaborative product development from an industrial perspective. Figure 1: Need for collaboration between the different phases of the product lifecycle and between different partners in the value creation network

Multilateral Platform-Based Collaboration 77 prostep ivip product data journal 2024-1 Survey results and need for research The survey, which was entitled “Multilateral development of complex systems” was conducted among the members of the prostep ivip Association in December 2023 (the results of the survey can also be found at https:// www.prostep.org/mediathek). In addition to identifying the challenges and opportunities of platform-based multilateral collaboration, the aim was also to provide the impetus to use the results for follow-up projects and activities (e.g. theme-specific research projects). Results were obtained from 29 participants, approximately half of which were from manufacturing industry, a quarter from IT and the remainder from research or other areas, and these give an initial insight into industry’s views concerning multilateral collabora- tion. In terms of subject matter, the participants were focused primarily on the fields of simulation and the early phases of product development (the “left branch” of the V-model). The issue of “IP protection” was the one most frequently stated as a core challenge for platform-based collaboration. This involves a number of different aspects: An important aspect is that it is essential to create a relationship of trust between the partners in order to make effective collaboration possible. This requires both the presence of a demonstrably trustworthy technical infrastructure that permits the secure exchange of information and data and certainty regarding legal and contractual issues. Other key challenges to collaboration identified by the participants were the absence of standardized interfaces, the lack of interoperable systems, tools and processes, as well as the extremely heterogeneous nature and constantly expanding scope of the information to be exchanged. The need to create and/or identify a “single source of truth” was identified as a further challenge in light of the disparate, cross-enterprise sources of information. Last but not least, “soft” factors such as human resistance to change and a lack of understanding of the technology were considered to be significant obstacles to successful platform-based collaboration. Accordingly, the participants listed the following prerequisites for the indus- trial use of data spaces. On the one hand, the most important prerequisites for the active sharing of data are: clarity regarding usage rights, reliable IP protection and clearly defined intellectual property rights. On the other hand, certainty regarding liability for data correctness, certainty concerning usage rights, and the availability and recency of the data are vital prerequi- sites for the productive use of shared data.

Multilateral Platform-Based Collaboration 78 prostep ivip product data journal 2024-1 In addition to exploring the current challenges and prerequisites for multi- lateral platform-based collaboration, the survey also provided an opportu- nity to develop hypothetical future scenarios. The objectives expressed by the participants for future collaboration included non-redundant data linkage, full interoperability with no discontinuities, synchronous real-time collaboration and simple, efficient, trust-based communication that protects the interests and rights of the collaboration partners. Security risks and concerns regarding IP security (in particular when working together with competitors), technical problems (difficulties at the levels of system integration and data interoperability, performance and cost difficulties), as well as the difficulty of establishing joint processes across enterprise boundaries were all identified as obstacles to the emergence of this form of collaboration. In the light of the various challenges mentioned – reliable IP protection, interoperability of systems and processes, certainty regarding the usage and IP rights to shared data, certainty regarding liability, as well as the quality, recency and availability of the data – multilateral, platform-based collabora- tion in particular in the engineering field still seems to face a large number of difficulties. This is all made more complicated by the heterogeneity, complexity, sensitivity and dynamic nature of engineering data, as well as by the large volumes involved. As the survey showed, there are currently still great uncertainties surrounding the industrial use of platforms for multi- lateral collaboration in the engineering field. These are making it difficult to fully exploit the potential offered by this form of collaboration and develop new business models. In addition to the value added that multilateral, platform-based collaboration brings to the development partners, such approaches can also lead to the development of completely new business models on the part of suppliers and service providers who, among other things, will be in a position to monetize their services and the supplied data. Further activities are focusing on the design and evaluation of possible new business models in connection with suitably adapted cooperative models in value creation networks. Outlook Some initial initiatives are already addressing some of the above-mentioned challenges. For example, the FAICE project (Framework for AI-enabled Collaborative Engineering) was launched by the prostep ivip Association in early 2024 (see also PDJ article “FAICE – an innovative approach to the collaborative use of AI in engineering”. The aim of this project is to tailor AI models for use in collaborative engineering scenarios by building on existing technologies in the field of artificial intelligence as well as in that of secure data exchange between peers. It seems likely that data space tech- nologies will play a leading role, in particular with regard to this latter aspect.

Multilateral Platform-Based Collaboration 79 prostep ivip product data journal 2024-1 Part of the prostep ivip project “Collaborative Digital Twins (CDT)” is – besides the definition and content elaboration of digital twins in general – the development and establishment of the potential of data spaces for engineering. The focus of the project CDT in this context is on the identifica- tion and prototypical implementation of suitable user stories from engi- neering, using existing data space technologies and, if necessary, adapting them. The examples presented here illustrate some initial attempts to make the advantages of multilateral, platform-based collaboration available to product development. However, these initiatives only address some individual aspects (e.g. the collaborative use of AI). Further research work is needed in order to take full advantage of the potentials for engineering and industry. In particular, it will be necessary to develop possible future scenarios for platform-based collaboration and possibly also test prototype solutions in order to come to a realistic appraisal of the opportunities and risks as well as of potential new business models. Contact Rachel Bauer Research & Innovation Manager prostep ivip Association rachel.bauer@prostep.org Contact Prof. Dr.-Ing. Stephan Husung Department of Product and System Development Technische Universität Ilmenau stephan.husung@tu-ilmenau.de

80 Building the Future: Decentralized Data Spaces with GaiaX Sonika Gogineni, Cansu Tanrikulu, Dr. Kai Lindow Fraunhofer Institute for Production Systems and Design Technology Maximilian Stäbler German Aerospace Center Institute for AI Safety and Security As digital transformation accelerates, data has become a cornerstone of business success (Otto et al. 2022). Initially, data was just a byproduct of processes; now, it’s a valuable asset and even a product itself (Fraun- hofer-Gesellschaft 2016). The modern era demands data to be available and shared securely among multi- ple parties, ensuring privacy and interoperability (Zheng and Cai 2020). This comes with a set of safety and secu- rity requirements. One solution for sharing data is to use Data Spaces. Data Spaces enable secure and reliable data exchange without needing physical data integra- tion or a common database schema (Otto et al. 2022). They support data sovereignty and privacy, key in our interconnected world (Schleimer and Duparc 2023).

Decentralized Data Spaces with Gaia-X 81 prostep ivip product data journal 2024-1 Enter GaiaX Gaia-X is an ambitious initiative designed to create a federated, open, and decentralized data infrastructure that aligns with European values (Federal Ministry for Economic Affairs and Energy 2019). It aims to facilitate data sharing while ensuring privacy and security. European Data Spaces call for the establishment of domain-specific Data Spaces. Data Spaces act like virtual bridges, linking diverse data points with- out needing them to be under one roof. They are about connecting, not collecting, ensuring data flows securely and seamlessly across borders while adhering to strict privacy and sovereignty guidelines (Otto et al. 2022; Cappiello et al. 2020). Gaia-X builds trust and ensures compliance, facilitating economic growth and driving innovations across sectors while focusing on European stand- ards. (Federal Ministry for Economic Affairs and Energy 2019; Schleimer and Duparc 2023). There are two ways of creating value defined by Tardieu in the Data Spaces 1. Volume Sharing: Partners share similar types of data, increasing the volume for better insights. 2. Diverse Data Integration: Partners share different but complementary types of data to create unique services (Tardieu 2022). Key Services in GaiaX Gaia-X’s latest architecture introduces “Federation Services” under ecosystem governance (GAIA-X European Association for Data and Cloud 2023). These services Identity, Compliance, Data/Services Exchange, and Catalogue services. These form the backbone of a decentralized data ecosystem. include

Decentralized Data Spaces with Gaia-X 82 prostep ivip product data journal 2024-1 Building the Data Space: The Basic Component Model (BCM) The BCM provides a comprehensive overview of the essential services needed to create a functional Data Space, bridging the gap between tech- nical and conceptual models. Let’s break it down: At its core, the BCM identifies two levels with a total of four key services for a Data Space: n Governance Level: - Identity Service: Verifies and manages digital identities of participants. - Compliance Service: Ensures all participants adhere to data govern- ance policies. n Federator Level: - Data/Services Exchange Service: Facilitates the secure exchange of data and services. - Catalogue Service: Manages and organizes metadata, making data discoverable and usable. Figure 1: Basic Components Model

Decentralized Data Spaces with Gaia-X 83 prostep ivip product data journal 2024-1 Data, along with Services and Applications, are positioned outside the core model (Figure 1) but are interconnected with these essential services. This setup highlights the interdependence of various components without suggesting a specific direction of data flow. The onboarding process to the Data Space involves a few critical steps: n Self-Description (SD): Participants provide a self-description, stored in the SD Register within the catalogue. n SSI Verification: The self-description is verified according to the Govern- ance Framework and stored in the participant’s Identity Wallet. n Metadata Entry: Data enters the Data Space through metadata encapsu- lated in a Semantic Model. The Semantic Model is crucial as it integrates data sets from various sources and links them to business applications. This model is stored in a Semantic Hub within the catalogue, discoverable via a Marketplace accessed through an API Wrapper and a User Interface. The data flow and its compliance in a Data Space is managed through: n Data Exchange: The Data Provider sends data to the Data Consumer through a connector, utilizing the semantic model for description. n Logging Service: Tracks data usage, ensuring compliance with Data Usage Policies and agreements. n Contracting and Payment: The Logging Service connects to Contracting, which then links to the Payment Infrastructure, enabling seamless data trading.

Decentralized Data Spaces with Gaia-X 84 prostep ivip product data journal 2024-1 Approach to Identifying Services for Use Cases in GaiaX In the Gaia-X 4 Future Mobility projects, various use cases are being integrated into the project Data Space. These use cases were initially developed by multiple stakeholders in an unstructured manner, making it challenging to analyze and implement them effectively. To address this, a structured template was developed. This template (Table 1) provides a clear and concise overview of the implementation plan, covering essential components such as service descriptions, use case details, participant roles, and resource requirements. The template, developed in accordance with the Gaia-X architecture model and enriched with insights from the Industrial Data Space Whitepaper by Fraunhofer, helps standardize the process. It ensures that all service offer- ings align with Gaia-X principles such as data sovereignty, interoperability, and transparency. Here’s how the template helps: n Service Offerings: Defines the service name and provides a short descrip- tion. n Participants: Identifies the trust anchors, providers, and consumers involved. n Resources: Specifies the virtual and physical resources required. n Policies: Details the provider and consumer policies, including usage and pricing models. Figure 2: Selected Use Case : Drone selection service

Decentralized Data Spaces with Gaia-X 85 prostep ivip product data journal 2024-1 Categories Subcategories Details of use case Name Drone selection service Service Offering Short description Participants Resources Policies Trust Anchor Provider Consumer Virtual Resource Physical Resource Provider policy / usage policy Consumer policy / search policy Pricing A service that collects drone informa- tion from different providers, gathers consumer needs, and selects a drone accordingly Trust Anchor in the Gaia-X 4 AMS Data Space Drone Selection Service Provider Fire department Drone data models, drone operation data Infrastructure, drone hardware information Limited On demand Pay per use Table 1: Service offering template filled for drone selection service Practical Application: Drone Selection Service The drone selection service from the Gaia-X 4 AMS project highlighted in Tanrikulu et al. was chosen as a prime example to illustrate the application of the template. It effectively demonstrates the interaction between providers and consumers and shows a practical use case in rescue mobility. The service aligns the resources of multiple drone providers with specific consumer requirements, such as specific sensor or camera specifications, making it an ideal choice to illustrate the service template. To effectively map the extracted information from the various use cases, understanding the requirements of stakeholders, for example in this appli- cation such as the fire department is the first step. One key requirement identified is the necessity for a trusted provider for drone services. This implies that the drone selector service should possess the capability to access a catalogue and offer a preselection of drones tailored to the specific needs of the consumer. Drone selector services require a catalogue, to be able to search and find all possible drone providers, logging mechanisms and data usage policies should be determined and included in the contracting. The mapping of the example service with BCM components is detailed in table 1.

Decentralized Data Spaces with Gaia-X 86 prostep ivip product data journal 2024-1 A semantic model should be defined outlining the parameters and require- ments specific to fire department missions. This includes factors such as mission location, type (e.g., search and rescue, firefighting), payload capac- ity, sensor requirements, and regulatory compliance. The drone providers describe their offerings within the catalogue using standardized formats compatible with the drone selector service’s semantic model. To do so, a comprehensive catalogue is needed. Each listing of an offering should include detailed information about the drone’s capabilities, including sensor types, camera specifications, payload capacity, and any specialized features relevant to fire department operations. The catalogue should have semantic search and filtering capabilities imple- mented, enabling the fire department to specify mission requirements using natural language queries or predefined parameters. This ensures precise matching of drones to mission needs based on sensor types, camera capabilities, and other specified criteria. The identity of drone providers and fire departments should be authenticated through Self-Sovereign Identity (SSI) verification mechanisms to ensure trustworthiness. SSI verification enhances security and reliability in drone selection, only verified providers are considered within the search results, ensuring trustworthiness and reliability. Once a suitable drone provider is identified, the service facilitates the contracting process by generating necessary documentation and facilitating communication between the fire department and the provider. Contracts outline terms and conditions related to drone services, including pricing, delivery timelines, sensor configura- tions, and service level agreements. The Data Space’s data usage policies should ensure that all data collected during missions adheres to privacy regulations and ownership rights. This includes specifying protocols for data sharing and access control, particu- larly regarding sensitive information collected during drone missions. Payment requirements should be defined from drone providers and con- sumers to streamline transactions between the fire department and drone providers. The drone selector service’s operations should be monitored in regards of compliance and regulations, continuously optimizing the service to meet the consumers’ needs. Logging mechanisms should be implemented to record the search and selection process, as well as contracting and policies. This includes logging search criteria, filtered results, and ultimately selected drones. If a connector is used for data exchange, data usage may also be logged to ensure the negotiated usage agreements. Detailed logs facilitate transparency and accountability in the selection process. Founded by the European Union NextGenerationEU The project GAIA-X 4 AMS is funded by the German Federal Ministry of Economic Affairs and Climate Action (BMWK) based on a decision of the German Bundestag (FKZ: 19S21004I).

Decentralized Data Spaces with Gaia-X 87 prostep ivip product data journal 2024-1 The BCM services, such as the semantic hub and identity wallet, are not mandatory and can be considered optional in the Data Space. They may be excluded if not required for other use cases. Conclusion The future of decentralized Data Spaces is set to transform the way data is shared, managed, and utilized. By enhancing interoperability, security, and governance, Data Spaces will unlock new opportunities for innovation and collaboration across various sectors. As technology and regulations evolve, Data Spaces will become a cornerstone of the digital economy, driving growth and efficiency in a data-driven world. The BCM model, along with the service offering templates provides an initial approach to identifying service components for decentralized Data Spaces, meeting the specific needs of various use cases. By continuously mapping diverse use cases though the templates, the model evolves to address emerging challenges and improve service provision. References Federal Ministry for Economic Affairs and Energy. 2019. “Project GAIA-X: A Federated Data Infrastructure as the Cradle of a Vibrant European Eco- system,” Fraunhofer-Gesellschaft. 2016. “Industrial Data Space: Digital sovereignity over data,” GAIA-X European Association for Data and Cloud. 2023. “Gaia-X Architecture Document - 23.1O Release,” available at https://docs.gaia-x.eu/technical-committee/ architecture-document/23.10/. Otto, B., Hompel, M. ten, and Wrobel, S. 2022. Designing Data Spaces, Cham: Springer International Publishing. Schleimer, A. M., and Duparc, E. 2023. “Designing Digital Infrastructures for Industrial Data Ecosystems – A Literature Review,” Wirtschaftsinformatik 2023 Proceedings. 32. Tanrikulu, C., Gogineni, S., Kondak, K., and Lindow, K. “Conception and Require- ments Identification of Gaia-X-Based Service Offerings,” Tardieu, H. 2022. “Role of Gaia-X in the European Data Space Ecosystem,” in Designing Data Spaces, B. Otto, M. ten Hompel and S. Wrobel (eds.), Cham: Springer International Publishing, pp. 41-59 (doi: 10.1007/978-3-030-93975- 5_4). Zheng, X., and Cai, Z. (eds.). 2020. Privacy- Preserved Data Sharing Towards Multiple Parties in Industrial IoTs. Contact Contact Contact Cansu Tanrikulu Fraunhofer-Institut für Produktionsanlagen und Konstruktionstechnik IPK +49 30 39006-415 cansu.tanrikulu@ipk.fraunhofer.de Sonika Gogineni Fraunhofer-Institut für Produktionsanlagen und Konstruktionstechnik IPK +49 30 39006-175 Sonika.gogineni@ipk.fraunhofer.de Dr. Kai Lindow Fraunhofer-Institut für Produktionsanlagen und Konstruktionstechnik IPK +49 172 1068963 kai.lindow@ipk.fraunhofer.de

Events 88 prostep ivip product data journal 2024-1 Events September 26.09.2024 Bordnetze Digital Home | ARENA2036 Stuttgart October 08.10.2024 Software Day Darmstadt 23.10.2024 JT Day (Gastgeber Fraunhofer IPK) Berlin November 13.11.2024 STEP AP242 Day (Gastgeber: AIRBUS) Hamburg 27.11.2024 Tokyo Systems Engineering Summit (Premium Partner: NTT Data) Tokyo December 03.12.2024 AI Day (in Kooperation mit it’s OWL) Online May 2025 13/14.05.2025 prostep ivip Symposium, Berlin SYMPOSIUM 2025 Newsletter Already subscribed? The prostep ivip newsletter keeps you informed about all activities around the asso- ciation – be it about the work and results of the project groups, events, the KI marketplace and much more. Click here to subscribe! Did you know? In each newsletter we publish in addition to prostep ivip events also event-tips by our members (right of selection reserved). Social Media Well informed with LinkedIn! This is the easiest way to receive our news – via our LinkedIn channel prostep ivip Association! It‘s best to connect right away and always stay up to date.

Web seminars 89 prostep ivip product data journal 2024-1 prostep ivip web seminars Another service that the prostep ivip Association provides to its members is participation in a monthly web seminar free of charge. Some of the web seminars are open to non-members. This series of events usually takes place on the last Friday of a given month and comprises a half-hour presentation followed by a discussion among the international group of participants. You can find out more about the web seminars and how to register by contacting sarah.giese@prostep.com or online at https://www.prostep.org/en/events/webseminars. The dates for 2024: 13.09.2024 Easing the crosscompany collaboration in simula tionbased Systems Engineering (SmartSE) Hans-Martin Heinkel (BOSCH) Dr. Melanie Kluge (BHC) members only 27.09.2024 Modelbased engineering collaboration in Collabora tive Digital Twins (CDT) members only 18.10.2024 Modelbased, crosscompany collaboration in the process chain „physical wiring system“ (VEC) Johannes Becker (4Soft) Nebenredner: Philipp Kremer (PROSTEP AG) open to non-members 25.10.2024 Integrated process chains in the virtual vehicle develop ment based on FDX (Functional Data Exchange) Guido Schneider (Peak Solutions) Martin Hagedorn (Mercedes-Benz) open to non-members 29.11.2024 Linking product data in different formats for data exchange (Digital Data Package) Torsten Schmied (PROSTEP AG) Michael Kirsch (:em engineering methods AG) members only

Media 90 prostep ivip product data journal 2024-1 Magazine for members prostep ivip Association ProductDataJournal 2024-1 Demonstrate presence in the important PLMJournal! Established and leading the field since 1994: the ProductData Journal from prostep ivip! Where experts from industry and science present twice a year the latest developments and projects. prostep ivip – Your Digitalization Partner: 40 % Users, 40 % System Administrators and 20 % Research & Development This is the way to reach 100 % of TopManagement, IT and Business of our members! Knowledge transfer is the highest ambition of the think tank prostep ivip. Take advantage of our network and reach your target audience through advertising: the decision makers of the top companies in the automotive, aerospace and manu- facturing industries, as well as research! The ProductData Journal is published with 2.200 copies as a bilingual summer and winter edition. In addition to high-quality scientific content the ProductData Journal will offer you the possibility to promote your company or products in a more targeted and personalized way. Here you will find all the important information for a potential advertisement in the next ProductData Journal!

Media 91 prostep ivip product data journal 2024-1 Advertising formats and prices valid from 01.01.2019 1/1 page vertical format (portrait) Euro 2,500 1/2 page vertical format (portrait) Euro 1,500 1/2 page horizontal format (landscape) Euro 1,500 1/4 page portrait Euro 900 1/1 page full bleed DIN A4 or in type area 180 x 256 mm Placement on the cover pages U2, U3 or U4, max. 3 ads per issue, 4-color. Euro 2,500.00 1/2 page portrait full bleed 101,5 x 297mm or in type area 180x256 mm or 1/2 page landscape full bleed 210 x 146 mm or in type area 180 x 225 mm. Placement in editorial section, max. 1 ad per issue, 4-color. Euro 1,500.00 1/4 page in type area 86,5 x 125 mm Placement in editorial section. Max. 2 ads per issue, 4-color. Euro 900.00 Phone +49 151 20528679 fabienne.kreusch@prostep.org prostep ivip Verein Dolivostraße 11 64293 Darmstadt www.prostep.org Contact

Media 92 prostep ivip product data journal 2024-1 General Information for Authors To begin with... Please submit an abstract of the article beforehand Text/Article n Sections New Products, New Members 1/2 page, ca. 1.200 characters (incl. blanks) + 1 graph n Reference value per page 3.500 characters (incl. blanks) + 1 graph n Project Contributions 14.000 characters (incl. blanks) + 4 graphics n Science & Research 17.500 characters (incl. blanks) + 5 graphics n Avoid product names in article titles and headlines n Text formatting should be limited to title, introduction, subtitles, and enumerations n Please include name of author, company/organization, title, phone and/ or email address as contact information, gladly with a photo n Please be sure to include an English version of the text right away n Layout: proof-reading and editing possible for German as well as English version

Media 93 prostep ivip product data journal 2024-1 Photos/Graphics n Save text and photos/graphics separately n Please state image assignments clearly in the text, clear image captions n Generate graphics in German and English! n Graphic formats: TIFF, JPG, Powerpoint for graphics, EPS n Resolution: at least 300 dpi Furthermore… n Place your emphasis on a specific project, not on a product. For editorial reasons we reserve the right to publish your article in a later edition/issue. Please understand that the editorial office also reserves the right to edit and abridge for layout and/or placement reasons. Please feel free to contact us prostep ivip Association Phone +49 151 20528679 fabienne.kreusch@prostep.org

Symposium 94 prostep ivip product data journal 2024-1 Imprint ProductData Journal, a Journal covering Applica- tions, Products and Standards in Product Data Technology Publisher prostep ivip Association Editorial prostep ivip Association Fabienne Kreusch Dolivostraße 11 64293 Darmstadt Phone:+49 151 20528679 fabienne.kreusch@prostep.org Graphics, Layout and Production prostep ivip Association in cooperation with Müller-Stoiber & Reuss Poststraße 9 64293 Darmstadt Germany Frequancy of Publication Biannual Subscription Rate Free of charge for members of the prostep ivip Association Copyright Articles, including extracts thereof – may only be re- printed with the written permission of the publisher, provided that the author and source are acknowl- edged. Named articles represent the opinion of the author, not necessarily that of the editors. All docu- mentation submitted for inclusion, particularly dia- grams, drawings, brochures, must be free of third party rights. On submitting an article, the author/ company automatically grants approval for further printing without remuneration in all publications pro- duced by the publisher, who retains the copyright for published manuscripts. Copyright by prostep ivip Association ISSN 1436-0403

Editorial

News in brief

New members

prostep ivip Symposium 2024

Interview with Ulrich Ahle

Interview with Frank Möring

Interview with Jan-Steffen Kuhlmann

FAICE – An innovative approach to the collaborative use of AI in engineering

Making Artificial Intelligence Available for Use in PLM

Making effective use of digital twins in the product lifecycle

Integration of VDA 231-300 in DIN SPEC 91383

Multilateral Platform-Based Collaboration

Decentralized Data Spaces with Gaia-X

Events

Demonstrate presence in the important PLM-Journal!

Imprint