Top Sector High Tech Systems & Materials

Roadmap Smart Industry Version 1.0 date: May 12th 2016

Roadmap team: Jan Post (Philips Health Tech, University of Groningen) Gregor van Baars (TNO Technical Sciences) Evert van den Akker (TNO) Herma van Kranenburg (NWO/STW) Timo Meinders (University of Twente) Wilbert van den Eijnde (ConsumersVoice)

1 | 34

Content

Summary ............................................................................................................................................. 3 1. Introduction ................................................................................................................................... 4 2. Importance of Smart Industry for the Dutch economy ................................................................. 5 2.1 Ambition and strategy.............................................................................................................. 5 3. International setting ....................................................................................................................... 6 4. SWOT.............................................................................................................................................. 9 5. Challenges ..................................................................................................................................... 12 5.1 Additive manufacturing .......................................................................................................... 12 5.2 Advanced Manufacturing........................................................................................................ 13 5.3 Robotics and Mechatronics..................................................................................................... 14 5.4 High precision equipment ....................................................................................................... 16 5.5 Condition-based (Predictive) Maintenance ............................................................................ 17 5.6 Cyber-physical systems ........................................................................................................... 18 5.7 Integrated life-cycle management .......................................................................................... 19 5.8 Human Technology interaction............................................................................................... 20 5.9 Mass customization ................................................................................................................ 21 5.10 Production management ...................................................................................................... 22 5.11 Smart Design and Engineering .............................................................................................. 23 5.12 New Business models............................................................................................................ 24 6. Smart Industry and the national and international context ........................................................ 26 6.1 Link to other roadmaps in Top Sector HTSM .......................................................................... 26 6.2 Link to other Top Sectors ........................................................................................................ 28 6.3 Link to international initiatives ............................................................................................... 28 7. Investments .................................................................................................................................. 31 8. Documents used (References) ..................................................................................................... 32 9. Contact information Smart Industry Roadmap team ................................................................... 33 Appendix 1: Organizations and stakeholders Smart Industry .......................................................... 34

2 | 34

Summary To face the challenges of remaining fit for the future, other countries have been developing strategies to build their industries to a level where they can compete within the global economy in the years to come. Most relevant for the Netherlands is the German Industry 4.0 initiative, but Belgium, Denmark, the United States and South Korea are also developing their own manufacturing strategies. In the Netherlands, this strategy is called “Smart Industry”. This document is the HTSM Roadmap for Smart Industry. It addresses (short-, long and) mediumterm industrial needs and actions for the development of Smart Industry by summarizing strategic focal points for innovation. The Dutch approach in Smart Industry is to prepare our Industry for the future to remain competitive on a global scale. Impact is expected in the fields of production technology, the products themselves (becoming smarter and providing data to gain even more value), and new business models. The document consists of an introduction to Smart Industry, analyses of the position of Dutch Smart Industry in the European and global context, the most relevant challenges industry is facing in the years to come, and what is expected in terms of investments for the future. This roadmap has been drawn up in close cooperation with a group of stakeholders from both Industry and science in the Netherlands. The main purpose of this roadmap is to serve as a vision document. The document is not intended to be fully complete in all details because the world of Smart Industry is developing fast. Smart Industry is not a stand-alone theme, which means that this roadmap is interconnected to many other roadmaps in the domain of HTSM, e.g. , Advanced Instrumentation, High Tech Materials, Embedded systems, Components and Circuits, Aeronautics, Automotive, ICT and Security, and outside HTSM, e.g. Smart-Farming and Energy. This roadmap focuses on the technologies needed and not directly on the application areas, which are covered by the application roadmaps. The market size of this sector is considerable. In the Netherlands, there has seen a rapid development in the industry for High Tech Systems in recent decades, with annual turnover totaling over EUR 27 billion. The ambition is at least to double this amount in the coming 10 years. This growth could come from reshoring, as Smart Industry makes European and Netherlands industry more competitive globally. To reach this goal, it is very important to act in a comprehensive way, together.

3 | 34

1. Introduction Smart Industry is an approach to initiate a common strategy to develop the industry further so it will be fit for the future (see http://www.smartindustry.nl). Industry is defined as our value creation activities, that ultimate result in value created for a customer. It is about future-proof manufacturing systems; these are smart and interconnected and make use of Cyber-Physical Systems. Digitization and new manufacturing technology are drivers for this. Smart Industry has to be seen as a strategic vision and action plan of our future industry. Smart Industry is built on three pillars: 1. High quality, network-centric communication between organizations, humans and systems, in the entire value network, including the product or service used by the end-users. 2. Digitization of information and communication among all value chain partners and at all levels in the production process. 3. Granular, flexible, and intelligent manufacturing technologies, adjustable on the fly to meet highly specific end-user demands. In the coming decade, a network-centric approach to production will replace linear production processes with intelligent and flexible network approaches. These networks will interconnect parts, products and machines across production plants, companies and value chains at a highly granular level. The network-centric approach will radically optimize production in existing value chains and, more importantly, the notion of network-centric production finally spells the end of the ‘value chain’ and the birth of the ‘value network’. One of the key enablers of the third industrial revolution was the digitization of information and communication. The Internet was instrumental in this, as was further software development. Digitization is raised to another level within Smart Industry. Not only will it enable communication between all partners in the value chain, but digitization of, for example, product quality, user-characteristics and production parameters based on sensor systems (Internet of Things) will also be crucial to new innovations in the production process, products and services and business models. Smart industry is about the next generation of technologies. New modular approaches, as well as (next generation) robotics, new ways of manufacturing ( for example 3D printing) and ubiquitous sensors will enable cost-effective flexible manufacturing to meet the specific demands by customers. Within the Smart Industry domain, Mechatronics and Manufacturing are essential to tackle the big challenges our society is facing. Proper design of machines are necessary for production and manufacturing, semiconductor fabrication, healthcare, etc. For example novel robot technologies, precision motion systems, or energy efficient drive techniques can constructively help to address problems we are facing in Climate change (environmental monitoring, but also more efficient production), Energy (efficient design of machines), Health (novel diagnostic or robotic intervention), Mobility (coordinated intelligent transportations) and Security (Monitoring and Intelligent prevention or Screening). The roadmap starts with information about the economic importance and international position and a SWOT analysis of the position of Smart Industry. The major part of the roadmap is the description of the challenges, the international setting of the Smart Industry and the expected investments in the future. The roadmap has been developed and reviewed by a multitude of persons from industry and academia representing key stakeholders in the Top Sector HTSM (for details of the stakeholders, see appendix 1). 4 | 34

2. Importance of Smart Industry for the Dutch economy For decades, the economic growth of the Netherlands, and thereby the prosperity and welfare, has been driven by population growth (baby boom, immigration), raw natural resources (coal, oil and gas), financial constructions (investment bubbles, housing bubbles) and increases in productivity in, for example, agriculture, industry and services. The first growth drivers are either exhausted or no longer desirable, which means that the main driver for maintaining prosperity and welfare in the future is a further increase in productivity, new business models and further strengthening of our competitive power. Industry is the backbone of the economy of the Netherlands when looking at its exports and earning capacity. The Dutch economy is strong. It belongs to the top 10 of strongest economies worldwide, with top sectors that can compete worldwide. The Dutch manufacturing industry is especially strong in delivering customized high-end professional products based on intensive relationships with its customers, but also a number of Dutch companies excel in mass production. The Netherlands has several companies that are frontrunners in digitization and advanced production technologies. However, behind this first tier there is still a lot to achieve. A Smart Industry survey executed by the Chamber of Commerce has shown that a large number of entrepreneurs (86%) are insufficiently aware of the imminent digital revolution and the consequences or potential it may have for their company. This is in line with the figures of the World Economic Forum. The Global Information Technology Report 2014 shows that the Netherlands is doing quite well withICT, but that it could do better in the adoption of new technologies. What is also striking is that companies use ICT more frequently in their contacts with consumers and less often in their business-to-business transactions. The survey and the many workshops that were recently organized show that entrepreneurs see various challenges ahead. The three most prominent challenges for companies are: 1. to be at the forefront of the ICT and production technology, 2. to collaborate effectively and organize themselves into chains and networks that make optimal use of business opportunities, by the use of big data and distributed production processes, 3. to develop new Smart business propositions using new and state-of-the-art technology and knowledge. 2.1 Ambition and strategy The ambition is to create a strong Dutch industry that generates growth and jobs. Industry digitization offers Dutch companies great opportunities to remain competitive in the face of increasing worldwide competition. Statistics show that ICT is currently the major driver for productivity improvements. Companies should therefore show the ambition to be more innovative and make use of these new (ICT) technologies to improve their competitive edge. The aim is to make Dutch industry more competitive by responding faster and better to opportunities provided by ICT and advanced and/or innovative production technologies. However, the benefits go beyond the businesses themselves. Strong and innovative industries generate growth and new jobs. That is the real objective.

5 | 34

3. International setting The purpose of this section is to map international trends that are comparable to Dutch Smart Industry. The amount of attention from other national governments shows that this is an important subject. However, it is not easy to make a comparison because the initiatives are diverse and the availability of information varies. Nevertheless, a number of conclusions can be drawn from the information collected. •

•

•

•

•

•

•

•

Industry digitization receive very substancial attention worldwide: Worldwide there is an increasing interest in and a growing promotion of policies to support the digitization of the manufacturing industry or the industry at large. It is clear that the German ‘Industrie 4.0’ concept was a major initiator. The view that ICT will cause disruptive innovation throughout the industrial complex is widely shared. Scope usually wider than ‘Industrie 4.0’: The German ‘Industrie 4.0’ concept mainly focuses on reorganizing the value chain. Many countries, including the Netherlands, have a wider scope directed at strengthening industry as a whole with ICT, new technologies as major drivers and new business models. Policy often aimed at getting back industry through innovation: The policy often aims to get back or revitalize industry that relocated or is about to relocate, with a program focused on innovation, flexibilization and new business concepts. Much attention for supporting valorization: An interesting observation is that many countries focus on supporting the transformation of knowledge into concrete applications (pilot production). The US Innovation Manufacturing Initiatives are a good example and are comparable to the Dutch Fieldlabs. A similar development is visible in Finland, Austria, the United Kingdom, Belgium and Germany. China is a big player, but other major players also invest significant sums: As can be expected, the big players in absolute figures are China, the United States, Germany, the United Kingdom, Spain and France. China in particular invests a lot of money (more than €500 billion) in strengthening its industries through ICT applications. The other big players also have a significant systematic government-funded investment policy running into the billions of euros to support their manufacturing industry. Very limited funding in the Netherlands for Smart Industry: In the Netherlands, there is very limited specific funding for Smart Industry. The local Dutch Fieldlabs are financed by local funding regimes like EFRO, which is a European fund for local development, and they are not financed from national funds. All countries that are comparable to the Netherlands have a structural policy: Virtually all countries that are comparable to the Netherlands have a structural national policy for industry digitization and strengthening the manufacturing industry (action agenda), involving investments of hundreds of millions or billions of euros in public funds. The Walloon Marshal plan with €2.5 billion earmarked for the fourth industrial revolution is a good example, but so is the public funds (more than €1 billion) the Austrian government has already been making available for this purpose for over a decade. The European Committee also has a strategy for industry digitization: There is also increasing attention for industry digitization at European level. Since 2013, the European

6 | 34

Committee has its own strategy (I4MS1: €77 million), which was recently linked to the KETs2 strategy (€6 billion). The coupling of the H20203 and the structural funds (ERDF4 generic innovation: €100 billion) is crucial. A part of this analysis is devoted to comparing the Netherlands with other countries, even though this is not easy due to the differences in the figures available. Nevertheless, it is possible to make some important observations and draw some conclusions: •

•

•

•

•

•

1 2 3

Dutch policy takes a broad interpretation of industry digitization: Dutch Smart Industry policy is in line with the broad scope followed by many countries. In the Netherlands too, systematic industrial innovation based on ICT and new manufacturing technologies are the central focus. Smart industry has a wide scope: Compared to other countries, the Dutch policies aim at a wide group of Smart Industry Technologies, for instance in its manufacturing, agricultural, chemistry and design Fieldlabs. Some other countries focus more on Advanced Manufacturing, one of the examples being the USA. The Netherlands is one of the leading countries in attention for valorization: With its Fieldlab approach, the Netherlands is following a worldwide trend to pay attention to valorization and scaling-up problems in its industry digitization policy. This approach places the Netherlands in a leading group that includes Belgium, Germany and the USA. The Netherlands is lagging behind in operationalization of the policy: Looking at comparable countries, it is clear that the Netherlands is a follower and that other countries (e.g. Austria, Belgium, Denmark, Finland, Sweden) have already made more progress in developing this policy. The differences are small and the gap can still be bridged, especially because the institutional collaboration between business and knowledge infrastructure established by the Top Sector policy is available. The Dutch government mainly focuses on adjustment and use of existing instruments: Many countries have a systematic overall policy for industry digitization, such as iMinds and Marshallplan 4.0 in Belgium, Production 4.0 in Sweden and ICT of the future in Austria. In the Netherlands, Smart Industry is the guiding concept and the focus is on adjustment and usage of existing policy instruments. Adjustment and use of existing instruments takes time: In the Netherlands, there are a number of important public funding resources for Smart Industry: WBSO5, TKI6, TO27 funds, NWO8 funds and the ERDF funds. WBSO is a generic instrument and the opportunities to support Smart Industry are limited. TKI (Top Sector policy), TO2 and NWO currently allow

I4MS is an innovation support program for SMEs. KETs stands for key enabling technologies.

H2020 are the Horizon 2020 innovation subsidies of the EU.

4

ERDF stands for European Regional Development Fund (ERDF), which is a fund allocated by the European Union. WBSO stands for ‘Wet Bevordering Speur- en Ontwikkelingswerk’ and is a fiscal instrument of the Dutch government to subsidize R&D through tax reduction. 6 Top sectors for Knowledge and Innovation (TKI) 7 TO2 stands for “Toegepaste Onderzoek Organisaties”, which is a federated collaboration of applied research organizations TNO, DLO, NLR, ECN, Deltares and Marin. 8 NWO is the Dutch organization for scientific research that funds thousands of research projects at universities and knowledge institutes. 5

7 | 34

•

•

•

•

small scale funding of Smart Industry initiatives, but this requires adjustment of the existing focus, which takes time. Unlike other countries, most instruments for Smart Industry are generic: Major funding sources for Smart Industry are WBSO, ERDF and TKI. These generic instruments can also be used for Smart Industry. Other countries, on the other hand, have developed specific, theme-based programs to support industry digitization. Limited and uncertain access to ERDF funds for the Netherlands: The European Committee offers substantial ERDF funds. As these funds mainly focus on strengthening economically weak regions, the chances of the Netherlands receiving much funds are small. There is enough money in the ERDF funds, but it is a generic program not specifically dedicated to Smart Industry. This means that funding requests for Smart Industry projects have to compete with projects for other purposes. A major part of the funds available have not been made formally available: A substantial part of the public funding for Smart Industry is not yet formally available, such as TOF9, ERDF and the intended financial support from SNN10. International cooperation and coordination is very important: International cooperation is essential for the development of Smart Industry. Other countries show a lot of initiative in this field, but it is also clear that cooperation and mutual coordination is necessary to make the best of the opportunities available.

9

TOF stands for ‘Toekomstfondskrediet Onderzoeksfaciliteiten’, which is a credit facility set up by the Dutch ministries of Economic Affairs (EZ) and Education (OC&W) to fund research projects 10 SNN stands for ‘Samenwerkingsverband Noord Nederland’, which is the North Netherlands Cooperation Agency in which the three northern provinces of the Netherlands work together to support and stimulate the regional economy

8 | 34

4. SWOT The whole domain of Smart Industry is huge and it is not only about process technology. The domain comprises process innovation and product innovation and, in the future, probably a combination of these two. The overlap always comprises ICT-related issues like data, numerical modelling, sensors, and control. This kind of ICT can be used in both processes and products. Another dimension of Smart Industry is the social context, because Smart Industry will have a big influence on people’s lives in the future because it leads to new products and a different working environment, which is therefore an aspect that needs to be addressed in this roadmap and will play an important role. Smart Industry also influences the business value chain. A variety of suppliers will supply or cosupply parts of the whole complex network, influencing the economic business model and raising major issues like ownership, data exchange and security, but also interoperability and life cycle design. Smart Industry also requires humans and robots to work together, making human-machine interaction and behavioral sciences important for Smart Industry as well. The challenges of these topics are described in the next section. In the 2015 report on Smart Industry, Science Agenda 20152025 by TNO-STW, the strengths of Dutch sciences is analyzed and described. Figure 1 summarizes the important scientific topics relevant to Smart Industry.

Figure 1: Overview of scientific topic related to Smart Industry

With a large team of stakeholders in Smart Industry (see appendix 1) an analysis was made of Dutch Strengths, Weaknesses, Opportunities and Threats. The results are shown in figure 2.

9 | 34

SWOT analyses January 2016 Opportunities 1: high mix, low volume, high tech 2: more advanced manufacturing 3: improve co-operation 4: new business opportunities 5: more service based business 6: faster time to market

Threats 1: higher market complexity 2: new international standards 3: Ned connection to new standards 4: catching up of other countries

Strengths

Weaknesses

1: good eco system 2: good technology base 3: multi disciplinary 4: Dutch good in complex systems and series of ‘one’

1: fragmentation 2: amount of financial support 3:conservative sector 4: lack of decisiveness 5: small country

A B Smart Industry Connection of Local “Fieldlabs”, co-operation to an Technology based international innovation & setting research C D International Connection to EU Innovation in Co-operation New Smart products

Figure 2: The SWOT analyses of the Smart Industry and future development areas

The SWOT analysis shows a number of important areas for co-operation: A: Opportunities & Strengths: One of the strong points of the Dutch labor force is that it is flexible, multi-disciplinary, willing to collaborate and entrepreneurial. These are all ingredients required to address the challenges in smart industry, based on a sound social and technological basis. The combination of research and technology and the existence of FieldLabs enables forefront technological innovation on different Technology Readiness Levels (TRL). The low and midrange TRL levels can be covered by fundamental research and the high TRL level valorization is covered in the Smart Industry “Fieldlabs”. A strong coupling between fundamental research and the Fieldlabs is a key enabler for the Dutch smart Industry initiative. B: Opportunities & weaknesses: Great opportunities are available in high-mix, high-tech, low-volume and mass customization production. Companies have the entrepreneurial spirit and the willingness to collaborate. However, the low level of financial support available in the Netherlands for Smart Industry and sometimes the lack of decisiveness to launch large initiatives makes the Dutch position within the Smart Industry initiatives around us more vulnerable. One of the opportunities is to intensify the connection to the Smart Industry initiatives within EU and the global world by creating interaction between regional/national co-operation and EU co-operation. C: Threats & Strengths: Process development and innovation is only a part of the Smart Industry agenda. Product innovation based on Smart, ICT related, concepts will also be an important part of the agenda.

10 | 34

Market-driven innovation first at product level and secondly at process level is essential here. In the EU, there is a discussion about Pre Commercial Procurement (PCP) to overcome this Threat. A realistic threat in adopting Smart Industry is high market complexity, but Dutch companies should be fit to address this complexity, as they are multidisciplinary. D: Threats & Weaknesses: Flexible production systems and the valorization of ICT and big data in Smart industry will be accompanied by standardization and IP issues about ownership of data, and this will have influence on the innovation process in time. In combination with the fact that the Netherlands is not a big country (and not a big player in international standardization), this could be a realistic threat for the future. Teaming up with other countries and international institutes can overcome this problem. The Netherlands could facilitate this discussion to help find the best standard, IP contract etc.

11 | 34

5. Challenges Smart innovations in the manufacturing industry are required to secure the competitive position of the Dutch industry and the welfare and wellbeing of Dutch society. Smart machines and robots communicate and interact with each other and their users, track errors and solve problems, and carry out on-demand maintenance. Smart Industry optimizes the human-machine interaction, yields faster, cheaper and more sustainable production, and enables mass customization at the highest quality. Business models will change, leading to the fusion of industry and services. In doing so, the industry will to enable better, personalized products and services, and improve efficient, adaptive, and flexible production, provisioning, and supply-chain processes. This chapter addresses the challenges in the upcoming 5-10 years for making the Dutch manufacturing industry fit for a sustainable future, so that it can contribute to economic growth and safeguard and strengthen employment. 5.1 Additive manufacturing Additive manufacturing (AM) is a promising and rapidly progressing field that provides unsurpassed design freedom and opens up many favorable possibilities at system architecture level when combined with design optimization. AM has numerous advantages compared to conventional subtractive manufacturing. It enables efficient manufacturing of complex, personalized and customized products built up layer-by-layer with high precision, resource efficiency (near-net shaping) and cost effectiveness. AM offers the possibility to create multi-material products (combination of, e.g., metals, polymers, composites, ceramics) and parts with material gradients. Typical AM technologies are 3D printing, tape placement, braiding, laser cladding, friction stir welding, etc. Integration with design tools and CAD software will allow AM to have a significant impact on both time and cost savings, as well as weight, storage, tooling, assembly, transportation, supply chain management and maintenance. By utilizing numerous state-of-the-art technologies such as pick and place, dispensing of viscous materials, sintering, etc., additive manufacturing can leap itself from merely producing bespoke dump parts to building smart objects printed locally. AM will be an enabling technology for many applications, such as embedded and smart integrated electronics, complex high-tech modules and submodules and human-centric products like dentures, prostheses, and implants. AM requires additional technology development to improve on cost, speed and quality, and therefore developments are needed in the field of new concepts for multi-material, multitechnology digital manufacturing, and high-speed continuous AM technology. Multi-scale computational material and process-level models are required to capture textured and multiproperty functionality. New models need to account for the influence of printing process parameters on the resulting mechanical behavior and functionalities. Different length scales will be relevant for proper macroscopic characterization. These can be used to investigate aspects such as threedimensional topography, surface texture and porosity, as well as residual stresses and delamination to improve AM accuracy and quality. AM equipment should enable first-time right manufacturing, higher throughput, better precision, larger dimensions and more versatile processes. These process developments should be accompanied by progress in advanced materials, with an emphasis on the related areas of laser and printing technologies, real-time in-line metrology, control technologies 12 | 34

and machine learning protocols. Besides standards for materials, new design programs, machine processes, and qualifications for built parts have to be developed to turn AM into a mature production technology. Also required are protocols for intellectual property rights in part designs for a digital workflow in many locations (encryption, standard file format, security for defense). 5.2 Advanced Manufacturing Advanced manufacturing technology contributes to the realization of three major trends in production systems, i.e. increased efficiency, quality and reliability. It requires process monitoring and modelling approaches, associated with novel optimization and maintenance strategies. Improvements in manufacturing technology will be data-driven and can be based on measurements or models (deep-learning techniques, statistics, and physically based models).

Figure 3: Integrated computational modelling approaches to improve manufacturing processes.

Research on integrated computational engineering in the past decades has resulted in many complex models, reaching increasingly high levels of accuracy and maturity. In most cases, these models are used to create a better understanding of the constitutive behavior of the material in question and the related production processes. However, reduced tolerances on product properties require higher accuracy of the current (simulation) models, whereas a higher level of maturity is required as well to make them useful on the factory floor as part of the control system. The final properties of many products are the result of a sequence of production steps, where preceding steps influence the current production step. Full-process-chain simulation is still in its infancy. The design of highly accurate multi-stage manufacturing processes requires the development of efficient full-process simulation tools. Automatic (mathematical) optimization based on process models is reaching a level of maturity that is acceptable for industry. After optimization, processes tend to perform better, but also become more critical. Therefore, robustness of processes has to be included in virtual process optimization, taking account of natural and uncontrollable variation in material and process properties. A challenge for the next decade is to integrate the evolution of variation in product properties in a full-process simulation for multi-stage manufacturing processes. Production tools are already equipped with a multitude of sensors and this will increase further in the near future. Current control strategies control the tool movements, temperatures, etc., but not got the state of the product. Advanced model-based control algorithms be developed to control the product properties directly and create a zero-defect manufacturing system. Standard metrology and feedback control is not yet able to adapt to high frequency (product-to-product) variations that are observed in practice. Data processing and translation into corrective actions, adjustments and active control must be improved to bring the desired performance improvement. Improved metrology includes accurate and absolute reliable measurements, measurement set-ups and measurement methods. Relations between the measured signals and product properties are extremely nonlinear and are described by the process models. Since these process models are usually very time13 | 34

consuming, lower order models are required for application in control loops. These reduced-order models can then be used to optimize the process or can be incorporated into a control system in a factory platform. A number of issues are very important, i.e. the simulations platform must be very robust and stable, data processing must be fully automatic and standardization plays an essential role to create platform robustness. Advanced model-based control can contribute to all three main trends: accuracy, flexibility and efficiency improvement. Enhanced control will drive defects to near zero levels. These technological developments enable industrial principles such as Zero Defect, Lean and Just-In-Time manufacturing to reach their full potential, while dramatically reducing cost and impact on the environment. This approach can be used in zero-defect production of high-volume production but is also very useful in flexible robotics to speed up the implementation on the factory floor. The adaptive learning process of the coming flexible robotics could be based on data both from physical modelling and sensors, in combination with adaptive control. This combination will lead to more innovation speed in process development and implementation on the factory floor. 5.3 Robotics and Mechatronics Mechatronics integrates electrical, precision mechanical, sensor, thermodynamic and control engineering and software for the design of products, systems and manufacturing processes. It relates to the multi-physics and multidisciplinary design of systems, devices and products aimed at achieving an optimal balance between all basic disciplines. Expertise levels are high, based on many years of development in various application areas such as semiconductor equipment, healthcare systems, printing systems, but also mass production equipment, consumer product design, scientific instrumentation, and automotive systems. Smart Industry themes such as high mix, high complexity, low volume manufacturing, introduce new challenges to robotics and mechatronics. The added value of mechatronics and robotics innovations is potentially very big, e.g. through integration of many sensors, wireless networks, and Figure 4: Integration of more information technology across the industrial environment feedback/ feedforward control (but also other sectors such as food processing, and smart approaches into the manufacturing agriculture) This typically leads to integration of more environment by increasing use of feedback/feedforward control approaches and production sensors, wireless networks and automation/robotics technologies into the manufacturing information technology. and assembly environment. In addition, handling technology related to gripping, manipulating, complex assembly, and precise component placement can be of great value, but also, and adaptive/learning or robust control loops. To achieve zero-defect manufacturing more data needs to be available at all steps in the manufacturing and assembly process. To begin with, this implies integration of many sensors, e.g. 14 | 34

for in-line inspection of parts, supervising correct processing, placement, assembly etc. Fast communication of the resulting data and measurement signals will be required to have a clear status overview of the manufacturing and assembly process. An intelligent processing platform and decision-making system is required to initiate corrective actions to specific production equipment, manual repair actions, compensation at other stages of the manufacturing and assembly line, etc. This calls for development of novel sensor technologies and metrology, vision integration, in-line inspection and monitoring, fast data processing and transport. Adaptive and learning control strategies will also play an importing role. To enable flexible manufacturing of high-mix, high-complexity, low volume products in a competitive way, it is essential to switch fast from producing small series of one product to producing the next product. Re-programming, unproductive ramp-up, and similar production-time loss factors destroy the competitive position and need to be eliminated through smart innovations. Effective flexible manufacturing will be enabled in both a feedforward and feedback manner. Feedforward in the sense that a priori product information (e.g. CAD data, design documentation, assembly instructions) directly leads to the optimal configuration of the manufacturing and assembly environment, including all internal communication and reprogramming of robotic manipulators for handling new parts for a new type of product (self-configuration). A self-learning process evolves when feedback derived from continuous monitoring of critical production parameters through continuous sensing and metrology leads to adaptation of machine settings or replacements of tools when quality parameters start drifting (self-learning). In Smart Industry scenarios, smart robots will take on a range of production tasks (production, inspection, transportation) and will behave as smart production entities based on local intelligence that reacts to data from sensor-rich production environments. However, the state-of-the-art task control strategies in manufacturing facilities still lack the flexibility needed for this future scenario. Research should focus on decentralized and mixed control strategies that will enable maximal flexibility and extensibility of systems of cooperating production robots. Mobile robots need to move in 2D and 3D through known and unknown, static and dynamic, structured and unstructured environments. Besides, they must be able to deal with unfavorable conditions for sensing, mobility and manipulation, like varying light conditions, water, dust, mud, etc. This relies on the robot’s observation of the world through its sensors and data acquisition through other robots and systems, such as surveillance cameras. They need to localize themselves, navigate to target destinations, while avoiding obstacles in a safe and efficient way. Intelligence and autonomy are key in this sense. Physical interaction between users and robots is getting increasingly important. Tele-operation and haptic feedback are examples of robotics technology to deal with more complex, more diverse robots, such that they can be safely be controlled by non-trained, non-professional users. A robot needs to be able to adjust itself to changing environments and changing tasks to work efficiently. Creating flexible and intelligent robots that are able to use and update databases and knowledge about their environment requires developments in many areas, in hardware, as well as in control software. Robots need to be able to learn from humans, their environment and from other robots. Especially robots working on repetitive tasks, which is often the case, can benefit a lot from learning, while optimizing their performance. There is a need for reconfigurable systems allowing self-adjustment, learning and adaptation, correction, and control as well as networking to bring about a significant impact on changeover time/cost, tooling, programming and energy usage of 15 | 34

those systems. Research should include aspects such as improved methods for engineering processes, communication structures, and generic resource description for ‘plug and play’ machine integration. Software development for integration and control of machine controller software is often timeconsuming and expensive. Considering robots alone, the costs of integration are three to five times the cost of the robot hardware alone. The reuse of robotic software artifacts is a key issue in decreasing the integration costs and can be promoted by domain engineering, components, frameworks and architectural styles. The interoperability of hardware and software components for robotics is also important in forcing a breakthrough in the development of robots. Business and consumer interests and technological advancements will lead to wide diffusion of robotic technology into our everyday lives, from collaboration in manufacturing to services in private homes, from autonomous transportation to environmental monitoring. Building an early awareness of the resulting ethical, legal, and societal issues will allow timely legislative action and societal interaction, which will in turn support the development of new markets. 5.4 High precision equipment To survive global competition, high-precision and high-quality products need to remain a differentiator, which will demand continuous improvements in process control and system accuracy in many aspects. The existing mechatronics competence base needs to be brought to the next level in the area of precision motion and handling systems, and thus requires significant progress in control systems theory, dynamics, thermal management, sensor technology and precision metrology, fast and efficient actuation, advanced control theory, motion control implementation platforms for high bandwidth control and data processing. This holds for both motion systems and robotic manipulators for picking and placing components and handling subassemblies. Distributed actuation, identification and control are mechatronic challenges in high tech systems with a high numbers of carefully selected distributed sensors and specially designed electromechanical actuators, with both continuous and discrete dynamics, and with systems and control technology that is able to handle this high level of complexity. Also driven by the availability of massive computing (massive parallel systems) new avenues for control become viable. Systems may possess many sensors and actuators and all information passing through the control can be used to estimate performance and disturbances at different time and spatial scales simultaneously. Multi Input- Multi Output control and systems that are adapting to disturbance or system variations will become industrially relevant. But also distributed control approaches to deal with the ever increasing complexity of high tech systems and their control architecture. Such distributed systems will allow multi-rate control solutions to be designed, lifting some of the limitations in present day equipment. Diverse types of measurement data must be combined to take the right decisions and actions. Research into the numerical processing, merging 1D, 2D, and 3D metrology, data fusion, and wireless transmission of this information will also be needed. Technologies have to be developed for mass reduction and increased speed of operation, while maintaining accuracy. New systems have to be really lightweight, able to cope with deformation (e.g. quasi-static, dynamical, thermally induced), extended actuation and metrology topologies, and operating under extreme conditions.

16 | 34

Higher speeds and accelerations are required to increase productivity, leading to large driving forces introducing more disturbances and heat loads. This calls for high force-density actuation, efficient power conversion technologies advanced drive electronics, distributed magnetic structures, highprecision and high-power switching amplifiers, and alternative actuation principles. Furthermore, structural stability should be guarded by advanced thermal control, spatial and deformation metrology, advanced materials with favorable properties, wireless machines, both for data and for power transmission, eliminating the parasitic influences of cabling. The field of systems and control is a strong enabling technology that ensures robustness to uncertainty of many feedback control systems in high tech systems applications. Also model reduction for multi-physics systems and hybrid system theory are relevant topics. Modeling interconnections of multi-domain (physical /chemical mechanical) dynamical systems in one and the same framework is needed to derive new concepts to exploit the combination to the full benefit of the system performance. A similar aspect is found in the integral optimization of mechanical design, topology, disturbances and controller solutions for high performance systems. In such cases developments of novel mathematic approaches or complete new paradigms will be needed. Solving multi-criteria, complex design problems will be the key to really exploiting the potential in novel system architectures. This probably will quickly go beyond human mental capacity. Shape and topology optimization provide a very promising enabler in this respect to find breakthrough solutions. On the longer term, this will call for methods to design and manufacturing of multimaterial systems. The mechatronic competence base for high precision manufacturing systems is highly developed and broad. For more in-depth elaboration, see the roadmap supporting position paper on mechatronics [ref www.xyz.] 5.5 Condition-based (Predictive) Maintenance Maintenance is vital in ensuring the availability, reliability and cost effectiveness of high-tech systems. Industry requires smart maintenance strategies to address the challenges posed by productivity, aging assets and servitization. Traditional maintenance concepts that are still commonly applied in industry rely on pre-determined fixed intervals for maintenance tasks. However, the degradation of systems is a dynamic process, governed by changes in both the system and its environment. The consequence is that many systems are either maintained too early, thereby spoiling a (significant) fraction of the system service life, or fail unexpectedly when the system is operated more intensively than anticipated. Condition-Based Maintenance (CBM) is therefore crucial to save on maintenance costs and increase system availability. Production facilities (and public infrastructure) are subject to aging. Towards their original end of life, these assets need to be upgraded, built anew or scrapped. CBM offers the opportunity to extend remaining useful life or improve sustainability and ecological footprint. Servitization is the trend that capital goods, such as complex machines, installations, and vehicles, are not just sold as a product but also offered as a physical component of a service: the availability of the equipment is what is being sold, often in a performance-based contract. If such an arrangement is to work and to be profitable for the party offering this service, condition monitoring and maintenance (and operations) based on such monitoring is essential.

17 | 34

The challenge is to achieve just-in-time maintenance. Such a predictive maintenance concept can be realized by following a multidisciplinary approach, combining disciplines ranging from failure physics (failure modelling, life prediction) and structural health and condition monitoring to data analysis, maintenance process optimization and logistic challenges in resource planning. The combination of data collection through smart sensor networks and advanced analysis of the collected data has great potential. Since many sensors are already available for process control, the specific challenge is to find out how that data can also be used for predictive maintenance. Datamining techniques like machine learning may be useful for that purpose, but methods based on modelling system behavior and failure mechanisms are expected to perform better in this context. Moreover, additional sensors may be needed when critical parameters cannot be derived from the process-control systems. Intelligent health and condition monitoring systems, e.g., based on advanced vibration analysis, can then be added. Ultimately, the objective would be to completely monitor the (production) system health and performance from a remote control room. Moreover, the production and its degradation process should be fully captured by models and the operational and environmental conditions should be controlled so that the number of human interventions is minimized and their predictability is maximized. This will lead to a reduction of costs and an increase of system availability. The most visible parts of Condition-Based Maintenance may not be the most crucial ones in ensuring a much higher acceptance of this new innovative way of maintaining assets. The most visible parts of the so-called data enrichment chain are data capture (through sensors) and data analysis (data handling and data analytics). A major challenge in data capture is the fact that data comes from different organizations and departments, and these may have conflicting goals and constraints towards sharing these data. The challenge in data analytics is to reconcile the perspective of the domain experts with their failure mode analyses with the perspective of the data analysts with their correlations between environmental and internal factors and degradation behavior. After that, the challenge becomes that of organizational collaboration in managerial decision-making, where financial, sales and other non-technical considerations have to be reconciled with the technical risks and opportunities. Next, the challenges come from planning and scheduling of the chosen course of action, and also from the organizational deployment of these decisions, partly through uniformly and adequately trained staff (e.g., with the aid of virtual reality and simulation) and smooth and robust IT-enabled workflows (e.g., with the aid of handhelds and augmented reality tools). Finally, the actual maintenance actions are performed in a standardized manner, with automatic logging of actions, and systematic evaluation afterwards of the effectiveness of the interventions so as to close the data enrichment cycle. 5.6 Cyber-physical systems A cyber-physical system (CPS) is a system featuring integration of, and coordination between, the system’s computational and physical elements, the system’s environment and driven by application constraints. Smart Industry involves integrating ICT in products, processes and factories and connect them to each other to achieve better quality and to adapt automatically and instantly to, for instance, changing material conditions or customer demands. Semantically interoperable systems collect and process detailed data about embedded states, events and processes, where data ownership and privacy are important. The resulting integrated approach towards design and implementation allows an increase in the overall CPS’s adaptability, autonomy, scalability, efficiency, 18 | 34

performance, functionality, predictability, reliability, and safety. To integrate networking, computation and physical processes, wireless networked systems for sensing and control are of key importance. Examples of CPS include communicating manufacturing systems/lines, systems to track and analyze emission, communicating (wireless) sensor systems, and systems to provide situational awareness. Internet-enabled decentralized monitoring and control algorithms, using wireless sensor systems, are required to improve process and product performance and enable proactive maintenance strategies using local and global information. The CPS will utilize effective, reliable, real-time and secure data collection and data analytics under industrial conditions. Modularity of CPS will play an important role in enabling a seamless interconnection of new CPS and/or removal of existing CPS from the network. Future CPS will have plug-and-play components, both the physical and cyber elements, where the CPS control must function properly under reconfiguration and can cope with scaling-up of the networked systems. Independently developed subsystems with their own goals and internal application need to support collective applications. Challenging performance requirements can only be met by making platforms aware of applications and vice versa. The current development cycle of CPS that follows a consecutive design of physical systems, of control algorithm and of information technology has limited the potential of CPS as disruptive technology. Bringing out the potential of CPS requires an integrated co-design method within an engineering science that brings together complex control algorithms, communication protocols, and computational platforms that can guarantee safety, performance and robustness. CPSs are complex distributed systems. Management of both hardware and software (distribution of time-critical tasks, locations of processing) during their life cycles need much better mechanisms and support than currently available. Future CPS can be operated in
safety- or mission-critical environments where failure in some of the components is not acceptable at any cost. The design of automation, control and information technology must therefore provide assurance of safety, reliable operation under normal and abnormal conditions, while guaranteeing performance. Self-monitoring and self-repair techniques are important. A big challenge to be addressed is that of security. Different stakeholders with varying business interests and hierarchy will access the product, production, and customer data outside enterprise boundaries to accomplish various manufacturing operations. Security mechanisms are required to protect privacy, safety, and the business cases of all parties involved. 5.7 Integrated life-cycle management Over the past years, the requirements made of industrial manufacturing companies have increased enormously due to ever changing market situations. Companies have to respond to external changes (e.g., increasing globalization, increasing market orientation, growing model variance, increased quantities, shorter product cycles, decreasing target costs) and internal changes (e.g., growing product complexity, increasing modification frequency of parts). New technologies like cloud computing, 3D printing etc. as well as the influence of cyber-physical systems will intensify this trend towards the future. New and complex products will be developed for a customer-oriented market. The production of highly customized products with short life cycles addressing volatile markets will require new structures and operational strategies from their supply chains. Future supply chains will need to reconfigure dynamically as customer-specific products will be based on an increasing 19 | 34

number of specific components. This calls for new technologies, structures and ICT systems to establish ad-hoc supply, manufacturing and de-manufacturing networks for customer-specific products, which support decision-makers in finding and establishing the best possible supply chain solution for any specific order. New supply chains that address globalization and the integrated offering of product with service will demand new approaches that take into account movement of material, exploitation of clusters of manufacturing excellence alongside an ability for local customization. Process design is gaining in significance. In the ever tighter global competition, it is critical to the long-term success of a manufacturing company. Processes are dynamic and call for an adaptation to the changing external and internal requirements as well as for the integration of the new technologies and the complexity caused by the cyber-physical systems. The goal is flexible and continuous processes that support the entire product life cycle from product planning and generation in engineering, procurement and production to distribution, service and recycling / end of life. In this context, information develops from a production factor with increasing meaning into an important success factor. 5.8 Human Technology interaction Personalization of products may offer great benefits for future users of products and services but, at the same time, it implies changes at multiple levels in the design process. Producers need to design and produce flexible solutions rather than one single solution, which will require adjustments in production processes and in dealing with personalized orders. To design new personalized products and product-service systems, research should employ empirical methods and tools to explore new forms of innovation through co-creation of value, system thinking, human-technology interaction and scenario-based, user-oriented product design, involving all stakeholders (users, designers and manufacturers).

Figure 5: New service systems based on innovative Human Technology Interaction in an industrial environment.

HTI in an industrial environment is becoming more implicit, and therefore the contribution of social sciences is of key importance. Co-evolution of flexible human and machine/robot systems will become possible by enabling collaborative environments to integrate and support a seamless flow of information between human and machine agents, person-tailored user interface for workermachine/robot interaction, production systems and robots that recognize worker’s requirements and behavioral patterns and robots that compensate for age- or inexperience-related worker limitations. Modern manufacturing system design builds on an optimal and continuous distribution of tasks between humans and machines for higher performance, adaptability and quality. To provide insight, (simulation) methods and tools for dynamic distribution of tasks (handling, inspection, interpreting and decision) between system/robot and human concerning productivity, quality, physical and cognitive workload and job satisfaction should be further developed. Adaptive operator 20 | 34

support systems are required to support fast and reliable reconfiguration of production systems and to minimize workload and learning time. Service and interaction robots will replace classical industrial robots. This implies paradigm shifts from stiff and precise movements to mechanical and socially compliant robots, from high-speed operation to low-speed operation, from a structured well-known environment to an unstructured, unknown environment that changes continuously, from robot “hard” skills to “soft” communication skills, and from robots isolated behind cages to freely interacting robots. Sensors are required to allow robots to react to their environment. They rely on vision, touch, sound, conductivity, and many other sensing techniques. These sensors often lack robustness, functionality and performance, and are very expensive. Combining the needs of multiple application domains can enable the required innovation to improve the state of the art in sensing. Moreover, the intelligent interpretation of sensor data, perception and sensor fusion of different sensing principles is very important to achieve these goals. Ubiquitous sensing has become a fact. There is a huge interest in going a step further and to make sense of the multitude of data that is being produced by sensors, cameras and microphones. Human-behavior understanding, affective computing and social-signal processing are some of the fields of research that investigate smart systems that understand what people are doing, how they are feeling and how they are interacting with each other and with technology. Further progress in these fields will allow a closer collaboration between people and robotic systems. Viewed from the other end, it is important to study how people will interact with these increasingly smarter systems. It is easy to see how trust will play an important role in this interaction. Questions about responsibility, accountability, control and the user in the loop will become prominent. Human Technology Interaction with products will also change drastically because products get smarter and become connected through the Internet of Things. Using sensors has become widespread in logistics, and smart products will increasingly present their users with information on their operation, giving usage, maintenance and repair instructions. 5.9 Mass customization In mass manufacturing, the optimization of production processes is the primary driver for price competitiveness. Drawbacks of this approach are that it leads to “one-size-fits-all” products with standardized components, conservative product designs, limited shapes, rigid supply chains and pressures to minimize product variety. If customer satisfaction is not primarily driven by price but also by the variety to choose from, a fast but more responsive manufacturing infrastructure is needed. Modern industry has to produce smaller lot sizes efficiently, enabling more variety in products against an acceptable price level. The globalized market and new business models require the ability to launch streams of new products with a high degree of personalization, for instance adapting to an individual’s biometric parameters or satisfying specific customer preferences. Customization is a game changer in high-value manufacturing and requires a much closer integration of design with manufacturing. At a local level, higher responsiveness needs to be developed to meet customer demands, and mass customization should be implemented in such a way that high product variability can be offered with a flexible combination of limited sets of modules. To achieve this, the manufacturing companies and their production systems must combine flexibility and efficiency. Future factories will be smaller, closer to their customers and increasingly modular. Manufacturing 21 | 34

equipment is modularized in units of the size of containers and even smaller (tabletop factories). Each unit contains a smart, Internet-connected piece of manufacturing equipment like a 3D-multimaterial printer, a CNC machine, assembly robots, test equipment, etc. With these units, one can produce all kind of products in small series. Most important for the suppliers/manufacturer is that the production system consisting of these units can manufacture different product orders without the need of changing the physical configuration. If the product mix or order volume changes substantially, the system has to be reconfigured and/or extended with more or different units. Product and production information can be uploaded to the production facility by the customers themselves or by their solution providers. The suppliers/manufacturers are able to offer a larger variety of parts/assemblies or even complete products in smaller batches. Previously, they got the order if they offered the lowest price. Nowadays they have to invest in flexibility. Any moment orders can change, product designs change and, at the same time, zero-defect quality is required. The future factory is situated in a personalized customer-centric world. It combines high performance and quality with cost-effective productivity, allowing small-series customization at large-series manufacturing cost level. The limit to customization will no longer be set by technology but by the extent to which customers are willing to be involved in the design of their own products. The challenge for suppliers/manufacturers is to make their factories smarter but, above all, more flexible and closer to the market. It is expected that ultimately smart factories will be smaller than before, integrated in metropolitan areas: “Industry as a friendly neighbor”. Economies of scale is no longer the paradigm, but the scale economies of networking become the rule of the game. Nextgeneration applications require technological breakthroughs like new printing technologies, new materials, multi-materials and advanced manufacturing concepts, and a high-level of integration. Future manufacturing systems deal with big sets of personalized data caused by the network centric approach for single-product/small series manufacturing, IP and copyright protection. The integration of Internet and Industrial automation has created a bright future for the smart factories of the future. 5.10 Production management Production management primarily addresses the decisions made for configuring a production system, either green or brown field. It considers both structural decisions (production capacity, facility design, production equipment selection and configuration) and infrastructural decisions (quality polices, supply chain logistics, production planning and control, production scheduling). The challenges for the coming 10 years are a direct consequence of three following trends. Firstly, producers of industrial equipment and production technologies are integrating an increasing number of hardware (i.e. sensors and automated calibration technology) to prepare their equipment to handle Smart Industry production paradigms. Secondly, markets are becoming much more dynamic, pushing companies to offer mass-customized products, production services systems at faster time to markets rates. Furthermore, emerging markets are expected to surpass established markets in developed economies in size and capacity in the coming 10 years. Thirdly, industrialized nations are tightening environmental regulations in a coordinated way in order to achieve a large reduction in CO2 emission and a substantial increase in the material reutilization rates. These trends require production systems to become much more flexible and resilient to uncertainties in terms of production volumes and product characteristics. In order to achieve this, the following challenges need to be addressed. First, more flexible and intelligent methods for 22 | 34

production planning, control and scheduling need to be researched and developed in order adapt to changing demand patterns, small production batches and disruptive market events. One challenge consists of developing methods that combine available sensor data and optimization algorithms while keeping human decision makers in the loop. Furthermore, this also calls for improved coordination between producers and suppliers, requiring more flexible contracts and increased information sharing. Secondly, new technologies to enable reconfigurable production cells have to be developed to enable fast changes in production layouts and to maximize the product variety by minimizing investment in production technologies. Such technologies include hardware solutions that (1) enable setting-up production cells by manipulating basic production equipment building blocks, (2) control-engineering/mechatronic solutions that minimize set-up times by automatically determining process parameters and machine controlling protocols, and (3) design automation tools to quickly determine feasible production cell configurations and layout characteristics. Thirdly, create symbiotic relations between the inputs and outputs of different production systems to optimize energy and material utilization rates. The main challenge is how to integrate the facility layouts and material flows of different companies producing different products in a way that waste energy and material can be reutilized. Finally, educate a technical workforce so that they have the capacity to operate and maintain this increasingly complex production environment. In this context, both engineers and technicians need to be trained constantly so that they can cope with production systems that are not only becoming more complex, but also much more susceptible to sudden changes. Solving these challenges will result in competitive and sustainable production plants using innovative technology-based approaches that drastically change rigid supply chain mechanisms and product-based business models into collaborative and robust production networks capable of delivering innovative products and services in time in a very dynamic and unpredictable, global environment. Solving multi-criteria, complex design problems will be the key to really exploiting the potential in novel system architectures. This will probably go quickly beyond human mental capacity. Shape and topology optimization provide a very promising enabler to find breakthrough solutions. 5.11 Smart Design and Engineering An integrated part of the Smart Industry roadmap is the Smart Design and Engineering challenge. The challenge addresses the lifecycle phase of design and engineering, responsible of translating customer requirements to product and manufacturing process specifications. With increasing variations of customer demands and increased automation of the factory the design and engineering organization has to be able to address more input variations and deliver more information to the factory, while pressure on time-to-market drives them to deliver faster than ever before. Secondly, the trend of work in industry is moving towards design and engineering from conventional manufacturing (“Praetimus, ‘Productization of supply chain companies’, white paper, 2016”). One of the key conclusions is that the added value of the manufacturing function is reducing while the added value of the design and engineering function is increasing. The way design and engineering is performed will change considerably due to the increased complexity and reduced lead-times. The processes will become highly integrated and automated. The challenge is to extend the current state of the art in design engineering with smart capabilities. The focus of this roadmap is on adding intelligence and internet to existing solution in design and engineering capabilities. Four fundamental areas of improvement are identified. Firstly, intelligent add-ons to existing models need to be developed to realize the model-as-a-service while meta 23 | 34

languages will be developed to support rapid (re)develop and (re)configuration of models. Secondly, system integration and standardization of interfaces will become of key importance. Methods are required to flexibly integrate and reconfigure models across disciplines and organizations. Standards for data exchange and communication to support flexible, integrated, reconfigurable processes across disciplines and organizations are needed as well. Thirdly, new control paradigms are necessary to enable experts to control operations of these complex and highly automated systems, including integrated verification and validation methods. Finally means need to be developed to reuse data and models beyond the original scope of application. Develop methods to extend data and model with meta data to support future reuse. This challenge has relations with other challenges in the roadmap. Integrated lifecycle management is related to this challenge as the application lifecycle becomes shorter the design and engineering organization has to focus on the knowledge lifecycle behind the application. Human Technology Interaction is required to enable users to address the far more complex and dynamic design and engineering process. New business models support the new collaboration strategies required to enable the integrated design and engineering process across organizations. 5.12 New Business models Smart Industry challenges the validity of established business models. Reconfigurable, adaptive and evolving factories are needed to face the uncertain evolution of the market or the effect of disruptive events. Manufacturing enterprises are pushed to take metropolitan actions, i.e., thinking globally but acting and staying economically compatible with the local market. Increasing customermarket orientation and the resulting problems of varying diversity and product complexity correlates to the complexity of business processes in manufacturing companies. The management of complexity in product and processes as well as the decentralization in case of smart factories is a real future challenge. The speed of corporate reactions towards changes to obtain or maintain a stable process situation depends on product, process transparency and open interfaces. The development of integrated, scalable and semantic virtual factory models will enable the implementation of decision support tools and optimization methods to address strategic decisions such as the location of new production sites, production technologies, and the selection of products and services to be offered in the market. These models should allow a fruitful interaction among all the relevant stakeholders in the design of manufacturing strategies. Besides, new business models that provide a fully closed loop circular economy rather than a linear economy approach need to be designed, i.e., models that reduce, reuse, remanufacture, recover, recycle and redesign. To successfully combine the use of new production technology, digitization and a network approach, companies are challenged to adapt the four interlocking elements of their business model, i.e., customer value proposition, profit formula, key processes, and key resources. The establishment of a network-centric production system that spreads throughout the entire asset life cycle may lead to the emergence of new forms of collaboration that are characterized by a cocreation approach to value creation. Customers play a predominant role among the collaboration partners. They become an integral part of the smart industry by providing information on their individual needs and use, which is critical input for optimizing existing and creating new networkcentric production systems. Future key research lines are the design and implementation of new business models that allow the active integration of customers, to evolve business models into efficient and effective mechanisms based on co-creation competences, customer intimacy within an 24 | 34

information-based business approach and the contribution of co-creation approach to the creation of customer intimacy. Future manufacturing enterprises collect customer requirements, analyze them and make the right product and service model. Enterprises are also expected to offer a comprehensive range of after-sales product services. New tools, methodology and approaches for user experience intelligence (i.e., social networks, crowd sourcing, social science methods, qualitative and quantitative, to generate insights, models and demonstrations, etc.) need to be explored and used. Smart Industry is on the agenda in many countries, hence, managing inter-company relationships in coopetitive settings, i.e., contexts in which competition and cooperation merge together, have to be advanced. Since competition and cooperation may not be considered as secluded spheres, the development and adoption of new technologies required for smart industry solutions (e.g., automation, digitization, flexibilization, etc.) may benefit from a network approach to innovation among various competitive firms. Existing business models will change because of the introduction of big data and new applications using this data. Innovative business models are based on a dynamic network of companies, continuously moving and changing in order to afford more and more complex compositions of services. In such a context, there is a strong need to create distributed, adaptive and interoperable virtual enterprise environments supporting these ongoing processes. The establishment of coopetitive settings to foster the development of a smart industry have to be pursued. Business models have to evolve into efficient and effective mechanisms of value creation within a smart industry. The technological advancement and adoption of technological standards will be impacted by the coopetitive setting and the pursuit of coopetitive strategies.

25 | 34

6. Smart Industry and the national and international context Smart Industry is not only a technical roadmap; it combines high-tech engineering disciplines with natural sciences, social sciences, humanities and behavioral sciences. Smart Product and Smart Manufacturing will affect our lives in the near term, and even more in the long term. This means that we have to prepare the Dutch community for the future. This preparation will concern new technology but also new business models, security issues, employability, etc. An important aspect is the interaction of humans with technology and the rapid developments in technology, e.g. in identifying the optimal conditions for humans, human skills, and their development through their life span, in a highly automated and digital environment. Security, ownership of data and the models based on this data are of big importance. In addition, standardization will become more and more important when this smart way of working lands on the factory floor. Currently, NWA route is defined around Smart Industry. This national agenda addresses long-term challenges in a broader context, considering transdisciplinary research themes from social, economic and behavioral sciences next to technological themes. The NWA route will deliver its first stable document mid-2016. Other documents describing Smart Industry on a national level include: • Smart Industry: Towards 21st century skills; Science Agenda 2015-2025, April 2015, ‘STW Agenda’. • Actie-agenda Smart Industry, Dutch industry fit for the future, November 2014, ‘FME Agenda’. • Agenda voor Nederland, inspired by technology. Chapter: ‘Dutch Industry, smart(-est) industry’, page 39. Below a summary is given on the crosslinks between Smart Industry and other Top Sectors and roadmaps. 6.1 Link to other roadmaps in Top Sector HTSM Advanced Instrumentation: Advanced manufacturing and predictive maintenance is achieved by, amongst others, a high level of control that requires advanced instrumentation (e.g., non-contact sensor technologies, strategies for improved control). This topic is explicitly addressed in the roadmap Advanced Instrumentation. Aeronautics: The competitive fields of the Dutch aeronautics manufacturing and maintenance sector are aerostructures, energy subsystems and components, maintenance and overhaul, aircraft systems and future concepts. Topics in the Smart Industry roadmap like advanced manufacturing, additive manufacturing, mechatronics, and mass customization are closely related to the roadmap Aeronautics. Automotive: The automotive roadmap has four pillars of which two are clearly linked with the Smart Industry roadmap, i.e. ‘Green Mobility’ (when it concerns the production of powertrains or components)’ and ‘Efficient, Sustainable and Flexible Factories’. Relevant topics in the Smart Industry roadmap are advanced manufacturing, mechatronics and robotics, and production management

26 | 34

Components and circuits: There is a continuous drive for smaller, better, cheaper and more efficient electronic components in the world. The link with the Smart Industry roadmap is mostly in the production of these components, on topics like advanced manufacturing, cyber physical systems, mass customization and high-precision equipment. Embedded Systems: Embedded systems comprise platforms of hardware and software that allow decision power and intelligence to be added to products, systems, infrastructure or services. The production of these smart products are linked with topics in the Smart Industry roadmap like CyberPhysical Systems, predictive maintenance, mechatronics and high-tech equipment. Healthcare: This roadmap comprises the R&D for healthcare systems and associated ICT, equipment, instrumentation and technical models. Topics in the Smart Industry roadmap that are linked to this roadmap are additive manufacturing, mechatronics and robotics, mass customization, human technology interaction and cyber physical systems. High-Tech Materials: High-tech materials focuses on the design, research and development of innovative materials. The Smart Industry roadmap addresses the production of these advanced materials and production of products using these smart materials. The most relevant topics are advanced manufacturing, production management and additive manufacturing. Lighting: The link to the Smart Industry roadmap is the advanced production of (O)LED-based products and life cycle management. Nanotechnology: One of the key enabling technologies for advanced manufacturing systems is nanotechnology. Topics in the Smart Industry roadmap that are clearly linked with this roadmap are advanced manufacturing, additive manufacturing, robotics and mechatronics. Photonics: The link between the Smart Industry roadmap and the Photonics roadmap is the use of advanced sensor technology for advanced manufacturing and high-precision equipment. Printing: Industrial printing (versatile production method to produce 2D and 3D structures) is one of the main challenges in this roadmap. The link of the Smart Industry roadmap to this roadmap is in topics like advanced manufacturing, mechatronics and 3D printing. Semiconductor Equipment: This roadmap focuses on the production equipment of advanced integrated circuits. The production of this equipment links to topics in the Smart Industry roadmap like advanced manufacturing, mechatronics, mass customization and high-precision equipment. Solar: The roadmap Solar focuses on solar equipment, processes and materials. The link with the Smart Industry roadmap is in the advanced production of components of solar equipment. Space: Three focus areas are defined in this roadmap of which ‘High Tech Space Instrumentation’ and ‘High Tech Space Systems and components’ are best linked with the Smart Industry roadmap for the production of components for these space applications (advanced manufacturing and predictive, remote maintenance) Chemistry: The link to the Smart Industry roadmap is Predictive and Condition-based Maintenance.

27 | 34

6.2 Link to other Top Sectors ICT: With everything being digitized and interconnected in Smart Industry, the cross-link to the ICT roadmap [https://www.dutchdigitaldelta.nl/actieplan] or the cross-cutting aspect of Top Sectors is apparent. Furthermore, the data analytics in Big Data is shared as a research topic. Agri&Food: In robust food production, sustainable farming and in food-consumer chains, there are cross-links with enabling technologies in Smart Industry. Chemical Industry: Like in Smart Industry, chemical industry has complex processes and complex production chains. Sustainability, re-use of materials, clean production methods, smart materials and cradle-to-cradle concepts are part of the new chemistry challenges. Creative Industry: Elements of the human-technology interaction and the customer as actor in the design process (low-volume series) are cross-linking this Top Sector with Smart Industry. Horticulture and starting materials: Robots and smart systems are the cross-link between this Top Sector and Smart Industry. Energy: In general, Smart Industry contribute in achieving higher energy efficiency. More specific, condition based maintenance of windmills and the development of smart grids are the cross-link between the Top Sector Energy and Smart Industry. Logistics: The digitalization and seamless interfacing of logistic processes in the digital factory including automated warehousing systems and realtime track&trace systems are the cross-link between the Top Sector Logistics and Smart Industry. 6.3 Link to international initiatives The next important dimension is Smart Industry in the international context. Almost all technologically important countries have a Smart Industry agenda. Next to this, there is an important EU agenda on this subject, for example the FoF (Factory of the Future program) 2020. This is already a co-operation with the EIT-ICT program. According to the EU-Factory of the future program: “The Factories of the Future PPP identifies and realizes the transformations by pursuing a set of research priorities along six research and innovation domains. Each of these domains embodies a particular aspect of the required transformations towards the factories of the future. The research and innovation activities undertaken within the domains should focus on a concrete and measurable set of targets, described as the manufacturing challenges and opportunities.“ Addressing these challenges and opportunities is at the core of what the Factories of the Future PPP is determined to achieve. R&D investments only become economically and socially relevant when they generate value for companies and society, namely via the creation and exploitation of new products, services, business models and processes, the development of new knowledge and skills and also the creation of new and better jobs. These objectives are particularly relevant in an industry-led initiative such as the Factories of the Future PPP and when public funds are involved.

28 | 34

Figure 6: Overview of TRL levels

The challenge for the next period is to take this trend even further, creating the conditions for projects and consortia to bring their results to a pre-commercial/exploitation stage, namely by developing pre-series, installing pilot production lines, etc.” The consequence of this European approach is that the Smart Industry program has to operate in a European context and that it has to cover most of the TRL levels. The lower TRL levels are connected to the more fundamental science programs. The higher levels around 3/4 are covered by applied science programs. The level around 5/6 is covered in the European context by, for example, the FoF program, defining demonstrator PPP’s and, of course, the Industry itself is strongly involved in TRL 8 and 9.

Figure 7: The European Valley of death

The European Iinnovation community and research domain is very strong in the lower TRL levels leading to a very strong research agenda but lagging behind in valorization and industrial innovation, leading to too less economic value. The missing link between fundamental research and the market place is called “the Valley of death”. To bring innovations to the market there have to be activities across all TRL levels. A possible way to solve this is Pre-Commercial Procurement (PCP) where the innovation is based on business cases from the market place or Industry.

29 | 34

For Smart Industry we also need activities across all TRL levels and these activities have to involve the EU domain in Smart Industry as well. What kind of PPP do we have available: •

•

•

On the low TRL levels: o EU: Fellowship program Madam Curie o National: NWA (Nationale Wetenschaps agenda) o Regional: local co-operation with universities On the mid TRL levels: o EU: FoF(Factory of the Future), the EU Eureka Cluster Artemis/Escel o National: STW (Stichting Toegepast Wetenschappelijk onderzoek), NWO (Nederlandse Organisatie voor Wetenschappelijk Onderzoek), TKI (Topconsortia voor Kennis en Innovatie) o Regional: EFRO based: Local Smart Industry Fieldlabs On the High TRL levels: o EU: Demonstrator project in FoF o Nat: o Reg: EFRO based: Local Smart Industry Fieldlabs

A multi-funding approach could cover all regional, national and European TRL areas. The only area where there is not much cover is the National High TRL level. The Dutch Government used to have the ‘peaks in the delta’ (Pieken in de Delta) approach for this.

30 | 34

7. Investments At this point, a budget estimation can be presented, although it should be noted that this estimation is very much under development because of ongoing discussions with companies and expected response from broad communication about the draft roadmap draft within the above-mentioned communities and platforms. The numbers presented below should therefore be considered as an indication and not a commitment. Given the nature of this roadmap, it is clear that Smart Industry-related topics are part of all and any development project within the Dutch industry. Double counting is therefore inevitable. The present status of the industry commitment to the Smart Industry roadmap is derived from the signed Letters of Intent from 2012, and these are being processed to get the correct numbers. The table with the EU-related project will be completed later this year. Table 1: Forecast of national investments related to the Smart Industry roadmap* Roadmap 2016 2017 2018 Industry 10000 15000 25000 TNO 2800 5000 10000 NLR 100 100 100 NWO 4500 4500 6000 Universities 5000 5000 5000 Departments (excluding TKI) Regions 10000 12000 14000 total 32400 41600 52100

2019 30000 15000 100 6000 5000

2020 35000 20000 100 6000 5000

16000 72100

18000 82100

2019 35000 30000 100 6000 5000

2020 40000 25000 100 6000 5000

10000 86200

10000 86200

*All figures in thousand € per year (cash and in-kind value), exclusive the NWA

Table 2: Forecast of international related investments within the Smart Industry roadmap* EU within roadmap 2016 2017 2018 Industry 30000 30000 35000 TNO 3800 9000 18000 NLR 100 100 100 NWO 4500 4500 6000 Universities 5000 5000 5000 EZ co-financing of EU programs European commission 10000 10000 10000 total 53500 58700 64200

*All figures in thousand € per year (cash and in-kind value), including national investments, exclusive the NWA

The above is a very rough estimate, partially based on the guideline that approx. 1% of the industries total R&D funds will be spent on this roadmap, based on the current situation and extrapolated to up to 2020. The tables will be updated when more accurate information becomes available. Due to apparent overlap between various HTSM roadmaps, it is uncertain which roadmap will be applicable for specific innovation. The nature of many Smart Industry topics for innovation is that it supports many different applications (semiconductor equipment, healthcare, printing, automotive, etc.) so although the innovation topics are clearly addressing this Smart Industry roadmap, it may turn out that the activities will be collected as innovation projects in other roadmaps that exploit Smart Industry knowledge and technologies for their application.

31 | 34

8. Documents used (References) The information shown in this roadmap is based on various contacts and meetings with stakeholders in the Smart Industry domain and on publications like roadmaps and other written information. A selection of these documents is given below: • • • •

• • •

• • • • • •

• • •

•

Nationale Wetenschaps Agenda (NWA). Nov 2015 https://www.wetenschapsagenda.nl/ COMMIT2DATA: https://www.dutchdigitaldelta.nl/big-data Actie agenda Smart Industry: Dutch Industry fit for the Future, TNO/‘FME Agenda’, November 2014 Smart Industry: Towards 21st century skills; Science Agenda 2015-2025, April 2015, ‘STW Agenda’. http://www.smartindustry.nl/wp-content/uploads/2015/06/Scientific-agendaSmart-Industry-2015-2025.pdf The Global Information Technology Report 2014. World Economic forum 2014, http://www3.weforum.org/docs/WEF_GlobalInformationTechnology_Report_2014.pdf Brainport Industries: http://www.brainportindustries.com/ High Tech Systems Center: http://www.tue.nl/en/research/research-institutes/topresearch-groups/high-tech-systems-center/ E. Bruins, D. Koser, A. Stokking. Smart Industry: Towards 21st century skills – Science Agenda 2015 – 2025TNO-STW. April 2015: http://www.smartindustry.nl/wp-content/uploads/2015/06/Scientific-agenda-SmartIndustry-2015-2025.pdf EFRA. EFRA Research Priorities. www.effra.eu Holland High Tech: http://www.rijksoverheid.nl/bestanden/documenten-en-publicaties/rapporten/2011/06/17/holland-high-tech/rapport-Top Sector-high-tech.pdf Robotics: Market Survey and Development for Brabant Roboned: http://www.roboned.nl/sites/default/files/RoboNED%20Roadmap.pdf Smartindustry: http://www.smartindustry.nl/ Internationale verkenning beleid digitalisering van de industrie,: https://www.tno.nl/media/6272/internationale_verkenning_beleid_digitalisering_van_de_ industrie.pdf Factories of the Future. Multi-annual roadmap for the contractual PPP under Horizon 2020. EFRA. 2013: http://ec.europe.eu/research/industrial_technologies Roadmap supporting position paper High Tech Mechatronics, G. van Baars et al. [http://www.hollandhightech.nl/nationaal/innovatie/roadmaps/smart-industry/xxxx] Agenda voor Nederland, inspired by technology. Chapter: ‘Dutch Industry, smart(-est) industry’, https://www.tno.nl/nl/over-tno/meer-over-ons-werk/agenda-voor-nederland10-inspirerende-essays/. High Tech Systemen en Materialen: Kennis en Innovatie Agenda 2016-2019

32 | 34

9. Contact information Smart Industry Roadmap team Industry Jan Post

TNO Gregor van Baars

NWO Herma van Kranenburg

Philips Consumer Lifestyle Oliemolenstraat 5. 9203 ZN Drachten The Netherlands Tel::+31512 599 111 [email protected] www.philips.com

TNO Technical Sciences De Rondom 1 5612 AP Eindhoven PO Box 6235 5600 HE Eindhoven Tel : +31 (0)8886 64348 Mob: +31 (0)621134540 [email protected] www.tno.nl

Technology Foundation STW Van Vollenhovenlaan 661, 3727 JP Utrecht Postbus 3021 3502 GA Utrecht Tel : +31 (0) 30-6001 308 [email protected] www.stw.nl

33 | 34

Appendix 1: Organizations and stakeholders Smart Industry The following persons have contributed to this document: Constantijn Cox Wim Symens John Blankendaal Patrick Bruijn Steven Soederhuizen Willem van Endhoven Joost Schut Bert Thuis Jeffrey Zitter Herma Kranenburg Jan Post Evert van den Akker Gregor van Baars Erik Ham Erwin Meinders Gu van Rhijn Egbert Jan Sol Bayu Jayawardhana Yutoa PEi Loes Jansen Martijn Wisse David Abbink Johan Lukkien Maarten Steinbuch Iddo Bante Juan Jauregui Becker Ton van den Boogaard Marco Groll Boudewijn Haverkort Jörg Henseler Fred van Houten Timo Meinders Stefano Stramigioli Tiedo Tinga Henk Akkermans

- 247 Steel - ASML - Brainpoort - Festo - Fokker - HHT - Keworks - NLR - NLR - NWO/STW - Philips - TNO - TNO - TNO - TNO - TNO - TNO - RUG - RUG - Tata Steel - TUD - TUD - TUE - TUE - UT - UT - UT - UT - UT - UT - UT - UT - UT - UT - Universiteit van Tilburg

34 | 34