(These comments refer only to the paragraphs within the remit of industriAll European trade union, i.e. Executive Summary, §1,2,3,6 and 7)

General observations on the document

The document is visibly a first draft, collecting the feedback from several separate discussions. No time seems to have yet been available to place the messages in coherence, nor to avoid repetitions.

Contradiction between 3 incompatible goals

One important contradiction appears in the document between (1) the strategic intention of achieving inter-operability, (2) the legitimate political aims of avoiding private monopolies and of distributing power and value added fairly among all participants in the supply chain, and (3) the reliance on spontaneous market evolutions to achieve this, with no public intervention. Very little evidence[1] is provided that these three goals can be achieved simultaneously. Historical experience tends to show that only two of these goals are reached simultaneously:

• inter-operability following a spontaneous market evolution towards private monopolies (goals 1 and 3, sacrificing goal 2) – the most probable outcome if nothing is done
• chaos and inefficiency of competing parallel technical solutions in an un-regulated market (goals 2 and 3, sacrificing goal 1)
• inter-operability following public intervention towards a technical and legal solution providing a fair distribution of power and value added along the supply chain (goals 1 and 2, sacrificing goal 3) – the most desirable outcome for trade unions, but also for the cohesiveness of European societies, and for the viability of European manufacturing industries.

The “penguin effect” is in fact a coordination problem – which only public intervention can solve

The so-called “penguin effect” is mentioned several times in the report (§7.1, p.46; §7.4.2, p.53), to illustrate situations in which none of the players risks taking the first dive by committing to a given technical solution or standard – thereby blocking the evolution towards technological progress. This phenomenon is also known as the “coordination problem” in industrial economics, and is well documented as one type of market failure – for which only public intervention, by giving a neutral, coordinating signal to all players in a market (which are also competitors), enables the transition to take place.

The number of Digital Industrial Platforms should be one per Integrated System (see definition and examples below)

The report rightfully condemns the quest for a “holy grail”, one platform meeting all needs, and applicable to all usages. However, I would contest the idea that the number of integrated Digital Industrial Platforms is so high that the current mess (360 registered Internet of Things initiatives!), or a milder form thereof, be considered as the norm for the future.

I would argue that efficiency requires us to have one Digital Industrial Platform per Integrated System to be served. An “Integrated System” in this approach would be defined as a system within which digital, automated communication and interaction enhance its overall performance (examples will be given below). Indeed, if there are several, incompatible and competing Digital Industrial Platforms to serve a given Integrated System, then the benefits of the automated communication and interaction are lost, because of gaps, overlaps, and redundancies.

From our discussions, I would consider that there is a limited number of such Integrated Systems:

• design, manufacturing, transport, maintenance, repair and recycling of non-food products, which I could designate as “Smart non-food industry”, with a functional scope broader than the “Smart factory” alone (which tends to restrict itself to manufacturing)
• Smart electric power grids
• Connected and autonomous driving
• growing & harvesting, raising & slaughtering, processing and transport of food products, which I could designate as “Smart food industry”, with a functional scope broader than the “Smart agriculture” alone (which tends to restrict itself to the initial stages of the supply chain)
• Smart healthcare
• Smart cities
• Smart homes.

A given item can belong to one or several of these Integrated Systems: a patient only belongs to the “Smart healthcare” Integrated System; a meat processing machine belongs to the “Smart non-food industry” Integrated System (by its nature) and to the “Smart food industry” Integrated System (by its function); an ambulance is an item in the “Smart non-food industry” supply chain, a component of the “Connected and autonomous driving” and of the “Smart healthcare” Integrated Systems.

For several Integrated Systems, there exists a clear public interest justifying to mandate one single Digital Industrial Platform

Public authorities are the only ones that can solve the “coordination problem” (= prevent the “penguin effect”), by mandating the uptake of a given technical-economic solution by a given date. Thereby, they provide all economic players with the certainty of the evolution, and on the date at which this evolution must be made – all elements necessary for private investment to take place.

The legitimacy of public authorities to perform such an intervention in the market could be grounded on the economic and technical efficiency benefits provided by solving this “coordination problem”. However, in the current political context when public intervention in the market remains contested in its principle by some political and economic players, it makes sense to concentrate public intervention on those fields where a clear and undisputed public interest exists, beyond the economic and technical benefits outlined above.

This public interest, justifying public authorities such as the Commission to mandate a single, interoperable Digital Industrial Platform, exists for several Integrated Systems:

• for “Smart non-food industry”: the Circular Economy
• for “Smart electric power grids”: security and continuity of supply
• for “Connected and autonomous driving”: safety of Vehicle to Vehicle interaction
• for “Smart food industry”: traceability for consumer protection
• for “Smart health care”: EU-wide portable Electronic Health Records.

There is therefore a strong rationale for the European Commission to take the opportunity of these clear public interests to mandate the fast convergence of European Digital Industrial Platforms around a single set of technical specifications, and thereby the competitiveness of European industry. These technical specifications should be accessible and usable by all players under Fair, Reasonable and Non-Discriminatory (FRAND) economic and legal conditions.

Specific observations on the document

These specific observations are in addition to any one that would be a consequence of the general observations outlined above. Proposed additions to the text are in bold, proposed removals are stroke through.

Executive Summary

p.5: Participants emphasized the associated legal, economic and social aspects which are not trivial. For example, privacy concerns regarding the sharing of data – probably most explicit in Europe in the Health and Care domain, but also for worker-related data – need to be taken seriously into account as they can easily create barriers to user acceptance.

Chapter 3. Overview of the Strategy in "Industrial Data Platforms"

§3.3.1, p.15: The insistence of the whole §3.3.1 on the “benefits of variety” makes very little sense, and should in my understanding be the purpose of a very thorough re-evaluation and discussion within the Working Group 2. Indeed, it is in contradiction with the overall target of end-to-end inter-operability and of seamless integration of systems.

Chapter 6. Overview of the Strategy in "Industrial Data Platforms"

§6.3.3, p. 44: The (international) possibilities and limitations regarding the use of data about workers (worker-related data) need to be clarified. Rules would need to be clarified at EU level regarding minimum standards on security, confidentiality, access to data by the worker, allowable processing of data and international transfer of worker-related data (specifically: outside the EU).

Chapter 7. Overview of the Strategy in "Internet of Things"

§7.2, p.47: To support this, standards for semantic and format interoperability are needed to connect different proprietary platforms (be they proprietary, community-based or open-source) and federate cloud services.

§7.2, p.48: It was noted that the rise of the Circular Economy may offer opportunities for linking the consumer market to industrial processes, so as to capture use information on the product life cycle for maintenance, repair, upgrade / retrofit / refurbishing, dismantling and recycling. Within the industrial IoT domain GE is a strong player and is already in discussion with Bosch and Microsoft, all having joined the Industrial Internet Consortium (IIC). It was highlighted that if the large companies drive the market, the dominant solutions of the future may not match EU goals or be in the best interest of the EU. Open standards or APIs, accessible and usable under Fair, Reasonable and Non-Discriminatory (FRAND) economic and legal conditions are essential to give SMEs  and industrial players in general access to new technology and leverage the dynamics of new ways of making business as proposed by start-ups.

§7.3.3, p.50: There is a need to address interoperability between commercial or non-commercial platforms, e.g. focusing on semantics and ontologies.

§7.3.3, p.51: Here the emphasis should be on supporting convergence to fewer but open platforms, accessible and usable under Fair, Reasonable and Non-Discriminatory (FRAND) economic and legal conditions. As highlighted Europe is strong in connectivity, which creates value that is distributed along the whole supply chain, but weaker in the area of Internet platforms and Big Data analysis where the value is more concentrated, and thus more visible there is potentially more added value.

§7.4.2, p.53: Here the EU should support a middle way that promotes the interests of European commercial or non-commercial platform developers, reduces dependencies for SMEs and start-ups, and allows seizing the benefits of new business models and innovative services across societal challenges, by making sure that the platforms can be accessed and used under Fair, Reasonable and Non-Discriminatory (FRAND) economic and legal conditions.

§7.4.3, p.53: The aim should be to promote openness to avoid lock-in, to support fairness in the sharing of the value added along the supply chain, to prevent the dominant position of one player while monitoring global competition, and provide supporting standards, regulation and policy.

[1] beyond mere wishful thinking or the implicit reference to the contestable markets fallacy, specifically in §3.3.1, p.15 “Competition between platforms is generally a good thing and switching costs for users of the platforms should be kept low. Lock-in should be prevented” and p.16, “there is a strong tendency for […] a oligopoly or even monopoly […]. It is therefore important that a playing field is preserved that continues to be interesting for new platform providers”). Once network monopolies are established, the investment time and cost necessary to contest them makes this contestation – or even the threat of this contestation – completely illusory.