=Paper=
{{Paper
|id=Vol-2900/WS6Paper4
|storemode=property
|title=Evolution of Industry 4.0 Platforms within H2020 Projects and Current Issues
|pdfUrl=https://ceur-ws.org/Vol-2900/WS6Paper4.pdf
|volume=Vol-2900
|authors=Tim Delas,Laura Caroline Ribeiro de Melo
|dblpUrl=https://dblp.org/rec/conf/iesa/DelasM20
}}
==Evolution of Industry 4.0 Platforms within H2020 Projects and Current Issues==
https://ceur-ws.org/Vol-2900/WS6Paper4.pdf
Evolution of Industry 4.0 Platforms within H2020 Projects and
Current Issues
Tim Delasa and Laura Caroline Ribeiro de Meloa
a
Ascora Gmbh, Birkenallee 43, Ganderkesee, 27777, Germany
Abstract
The H2020 framework of the European Union is eager to support research in the Industry 4.0
domain for the European manufacturers to gain a competitive advantage through better
digitization. Originated by extremely costly and aged hardware besides the traditional
approaches for management and knowledge/data transfer, a real need for improvement are
perceived.
Many projects have been conducted in the last 10 years, and within the ZDMP (Zero-Defects-
Manufacturing-Platform) Innovation Action the evolution of the predecessors toward the
current approach are reviewed to identify potential constraints for other projects in the same
research and innovation sector.
Still, EC (European Commission) reviewers stated that it is hard for external parties to
understand the relation between these projects and the concept's evolution to reach the
objective of innovation and digitization of manufacturing processes in the EU. This document
depicts the most important concepts of H2020 platform projects in the manufacturing domain
of Industry 4.0 in the last 10 years with examples focusing on CREMA, vf-OS, and ZDMP, to
demonstrate which issues hindered dissemination of these platforms in the past and how the
following projects are evolving to get a foothold in the market by 2022-2023.
Keywords 1
Zero Defects, Manufacturing Platform, Industry 4.0, H2020, CREMA, vf-OS, Evolution,
Business Concepts, Data Privacy
1. Introduction
Industry 4.0 is revolutionizing the manufacturing industry based on the digitization of assets and
processes as a key innovation driver for manufacturing companies. Factories are complex systems of
other systems, and in many of the manufacturing industry, the digitization still did not happen.
The integration issues faced are manifold, machines often cost millions, therefore, hardware cannot
simply be replaced when interfaces are missing, instead, proprietary interfaces have to be used. Also,
additional sensors may have to be manually added to machines or products, taking time, effort and
additional planning, posing as a business challenge for a platform that should reach the real-world
business sector.
The basic technical challenge was to interface with the machines and planning systems in
manufacturing companies and connect this to a platform to develop solutions on. The author’s first
experience with such a task was 2011 in the FP7(7th framework programme) EU project ADVENTURE
(Adaptive Virtual Enterprise Manufacturing Environment), where a message bus, a global database and
virtual machines composed the platform, and the access to machines was individually programmed (in
a concept called “Gateways”). The technical concepts were lacking, but still, the concept sounded good
enough to rework it for H2020.
Also, technical issues were in all of the projects either conceptually inherent to the proposal or
brought to the projects as a matter of unfit implementation technologies, mismanagement, or lack of
Proceedings of the Workshops of I-ESA 2020, 17-11-2020, Tarbes, France
EMAIL: dellas@ascora.de (T. Delas); lc@ascore.de (L. Ribeiro de Melo)
©️ 2020 Copyright for this paper by its authors.
Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0).
CEUR Workshop Proceedings (CEUR-WS.org)
experience of single partner companies. Research projects are sometimes like Frankenstein monsters,
and when the proficiency with a particular technology is proclaimed by a partner company, but the
conceptualization or implementation of it is lacking in certain areas, some vital parts of the platform
might prove to hinder the whole project. Usually, risk management should come into play, but the
management of large research projects is also a hard task where issues can arise.
Problems can occur in different areas, and the evolution of these different areas are analyzed in the
following sections.
H2020 Research and innovation action CREMA (Cloud-based Rapid Elastic Manufacturing) was a
3-year EU funded project that aimed to create a platform, to facilitate the development of custom
applications to the manufacturing sector. The concept was based on the low-level ADVENTURE
project from the 7th framework programme, where the initial idea of providing a platform preconfigured
for industry use was conceptually validated. Partner companies from the manufacturing sector showed
interest, and therefore the need to drive this field of research was confirmed..
While CREMAs concepts stemmed from ADVENTURE, vf-OS (Virtual Factory Operating System)
was launched a year before CREMA finished, and focused on taking the technical and business concepts
further. Now ZDMP and eFactory are sister projects that started a year before vf-OS came to an end in
December 2019.
2. Technical Concepts Evolution
CREMA did provide working solutions based on a messaging infrastructure, common Cloud storage
for data and many more application-level tools and services. Processes and workflows could be
represented in a BPMN editor called Process Designer, where steps could call services, and the virtual
machines for these could be set up during runtime if necessary, and then the response would be taken
on to the next steps.
This Process Designer was verbalized as “the programming language of CREMA”, to make clear
that technical knowledge is still necessary to build technical solutions, as some partners were
envisioning a programming-free technical solution - something not possible yet for some years. Also,
the Process Designer did not make the design of solutions easier, it just enforced a model which works
with the services and the marketplace of CREMA, which restricts the possibilities of programmers
creating solutions. Also, it had implications on execution sequence and speed, as parallelization is very
basic in BPMN-like structures and also limits programming paradigms. Still, it was possible to
implement many processes and tasks and use any of the services in the Marketplace component.
CREMA lacked technically in many areas, but especially security, federation and data privacy, all
extremely important aspects for industrial companies with confidential business data..
The project vf-OS proposed to improve many of these factors. vf-OS was intended to be an open
operating system for virtual factories composed of a kernel, application programming interface, and
middleware specifically designed for the factory of the future – concepts from computing systems were
transposed on the digital infrastructure of manufacturing plants. Security and federation of components
were planned across the other technical components and concepts to be developed in the research
project.
Also, were CREMA did not endorse 3rd party support, in vf-OS, the integrated development editor
(IDE), innovation hubs, better documentation and even more tools specifically designed to help 3rd
party developers to create solutions on top of vf-OS, coined vApps.
While vf-OS was coming to an end, two more evolutions of the concept were planned: eFactory
intends to focus on a business infrastructure route to combine the different kinds of Industry 4.0 projects
and their solutions and marketplaces, while ZDMP is another more massive refinement of vf-OS,
reinstated as an Innovation Action instead of a Research and Innovation Action using 30 partner
companies (half partners are from the industry) to push the technology in the market. Technically
though, ZDMP puts the focus on industrial-grade solutions (open source as much as possible), instead
of custom-built systems, and adding specific components and tools to implement the Zero Defects
paradigm.
3. Business Concepts Evolution
The concept of CREMA lacked in one major area – the exploitation would only work when the
CREMA solutions would be supported by a company with a valid business model, to keep the servers
running and sell a product to the manufacturing companies. The consortium already realized this from
ADVENTURE but specifying direct exploitation through a legal entity within a proposal was a no-go
in proposals to the European Commission at the time. Therefore, the coordinator decided that the
exploitation deliverables within the project should put together the different factors needed for a
business plan, to at least have the chance of combined exploitation if the plan was promising enough.
The business model, market analysis and dissemination towards a common legal entity was created
within the project, but as the contract with the EU didn’t foresee the creation of a legal entity, partners
commitment was very limited and the experiment of creating a company within CREMA leading to
supporting actors to lose interest, despite the coordinator intention of following the experiment. The
concept was clear, and therefore the creation of a company at the end of the project was now envisioned
in the vf-OS Description of Action (DoA). While the project was mostly focused on technology,
collaboration with ZDMP was planned as the last year overlapped with the start of ZDMP, and the
ZDMP DoA went much further in terms of common exploitation including a legal vehicle, as, by 2018,
the EC adopted the necessity of this.
In ZDMP, a company would not be founded at the end of the project but instead within the first year
of the 4-year-project (which was the last year of the vf-OS project), as at the end of the project funding
period viability of this company did not matter anymore for the participants. Also, the DoA specified
that the company to be founded would be applying as an equal partner and consortium member after its
founding and would get redirected funding from the other involved partners, so that building a real
company was now also in the hands of the ZDMP consortium..
Getting a critical mass of users is also a major issue for a platform. For CREMA, this was practically
impossible. In terms of 3rd party application support, documentation was minimal, technical support
was missing. In vf-OS, the case was much better. As mentioned above, vf-OS incorporated an IDE,
innovation hubs, documentation and tools for 3rd party developers to create vApps for the integrated
marketplace. Still, the motivation of 3rd party developers to test out the platform was nonexistent, as
there were no customers in the vf-OS marketplace. The vf-OS DoA planned hackathons with a bit of
price money, and those were conducted but still not more than a handful of 3rd party developers tried
out the technology and were not excited.
In ZDMP this issue was attacked in manifold ways. The aforementioned business concept is realized
as an internal drive to get a profitable marketplace for 3rd party developers and possibly some products
generating visibility and is now anchored in the DoA. A second tenet is the 3rd party developer support,
which was taken over from vf-OS, combined with the concept of open sub-calls. 3,200,000€ will be
distributed for projects with up to 150,000€ each that use the technology created within ZDMP,
registration for information as soon as it becomes available is possible on the ZDMP website [1]. Also,
the marketplace enables the selling of 3rd party solutions towards customers directly within ZDMP
under the reign of i4FS, which means that i4FS can get a provision of the sales the 3rd party developers
make.
4. Current Challenges
Data privacy in the Cloud is hard. Technically, the main issue is that computing on encrypted data
(i.e. using homomorphic encryption) is not technically viable to use in Cloud databases yet [3].
Therefore, to be able to query a database, data cannot be hidden from the database provider. The
contents can be encrypted for data privacy reasons, but then no useful queries can be done with the
information in the database, which is the main reason to put data into a database in the first place.
Solutions currently are separation of data into secret and non-secret data, and either encrypting the
secret data or store it in a trusted environment.
Storing data on-site vs. in a private or public Cloud so far is the only way to ensure that business
secrets cannot be exploited by rivals, especially as trust in online systems has been eroding since it
became clear that the intelligence services have been corrupting technology infrastructure worldwide,
which indicates that backdoors can be expected everywhere. Hard encryption and on-premise solutions,
therefore, can be the only way forward to secure business secrets currently. Therefore, the only technical
solutions can be seen in a separation of data into private (on-site), semi-private (Private Cloud) and
pubic data (Public Cloud), as the aforementioned project's topic is not data security, neither primarily
nor even secondarily.
Creating a business without a product is difficult. Creating a company in ZDMP is not a problem
but creating a product for companies in the market is difficult. If it was easy, it would have been done
without an H2020 project behind it. Having a promising product though is essential to get people
invested in this company so effort and resources are put into it. Therefore, finding out how to create a
sellable product from such an Innovation Project is one of the main tasks, as this product is not defined
in the DoA. To prevent conflicts of interests by consortium members, a legal body with its employees
needs to exist, so they can define a product (or multiple products or solutions) to sell, which might make
market research and product definition necessary. If all of these steps have happened, a commercial
uptake of the platform is viable.
Currently, the company is coined i4FS (Industry 4.0 Factory Solutions [2]) and at the time of writing
(Jan 2020) consists of seven core companies from ZDMP and vf-OS. The main issue in this early stage
even before the entry as a possible beneficiary of the ZDMP project is finding a business developer who
has the qualifications and the motivation to drive this artificial company, that is even without a concrete
product or business yet. This business developer also should not be directly involved with the
consortium partners, not ask for much money and needs to be convincing the EC, which still has to
approve the amendment inviting i4FS as a beneficiary.
5. Conclusions
The explained opinions demonstrate the issues that H2020 platform projects in the Industry 4.0
manufacturing domain are dealing within the last 10 years with examples focusing on CREMA, vf-OS
and ZDMP, and the current challenges in terms of technical concepts and business concepts.
6. Acknowledgements
The research leading to these results received funding from the European Union H2020 Program
under grant agreement No. 825631 “Zero Defect Manufacturing Platform (ZDMP)”. The content of this
paper does not reflect the official opinion of the European Union. Responsibility for the information
and views expressed in this paper lies entirely with the authors.
7. References
[1] Zero Defects Manufacturing Platform, Register to join the ZDMP Open Call to qualify for your
share of €3.2M funding, 2020. URL: https://www.zdmp.eu/register.
[2] Industry 4.0 Factory Solutions, Industry 4.0 Factory Solutions Web site, 2020. URL:
https://i4fs.com/.
[3] D. J. Wu, Fully Homomorphic Encryption: Cryptography's Holy Grail, 2015. URL:
https://www.cs.virginia.edu/dwu4/papers/XRDSFHE.pdf.