next generation computing roadmap

98
Digital Agenda for Europe Next Generation Computing Roadmap FINAL REPORT A study prepared for the European Commission DG Communications Networks, Content & Technology by:

Upload: giannis-grig

Post on 10-Nov-2015

214 views

Category:

Documents


0 download

DESCRIPTION

FINAL REPORTA study prepared for the European CommissionDG Communications Networks

TRANSCRIPT

  • Digital

    Agenda for

    Europe

    Next Generation

    Computing Roadmap

    FINAL REPORT

    A study prepared for the European Commission

    DG Communications Networks, Content & Technology

    by:

  • This study was carried out for the European Commission by

    eutema GmbH (Austria) in co-operation with Optimat, EPCC and 451 Research

    Authors: Erich Prem Jrg Irran (eutema), editors; Mark Sawyer, Mark

    Parsons (EPCC), Csilla Zsigri (451 Research), Ian Morgan and

    Ashley Stewart (Optimat)

    Co-ordinator: Erich Prem, eutema

    prem [at] eutema [dot] com

    Internal identification

    Contract: 30-CE-0528423/00-42

    SMART 2012/0052

    DISCLAIMER

    By the European Commission, Directorate-General of Communications Networks, Content & Technology.

    The information and views set out in this publication are those of the author(s) and do not necessarily reflect

    the official opinion of the Commission. The Commission does not guarantee the accuracy of the data included in

    this study. Neither the Commission nor any person acting on the Commissions behalf may be held responsible

    for the use which may be made of the information contained therein.

    ISBN 978-92-79-37580-4

    DOI: 10.2759/4587

    European Union, 2014. All rights reserved. Certain parts are licensed under conditions to the EU.

  • About the authors

    Erich Prem is a research and innovation strategist based in Vienna, Austria. He is the CEO of eutema and has a background in Computer Science and lectures at the Vienna University of Technology and University of Vienna.

    Jrg Irran is a computer scientist working as a programme manager and IT consultant in Vienna, Austria. He is a graduated engineer and an experienced computing professional with many years of experience acting both as a researcher and IT consultant and being involved in several scientific and industrial driven research projects.

    Mark Sawyer is a High Performance Computing specialist with over 20 years of experience managing technology transfer projects between industry and academia. He has been involved in HPC road-mapping activities for the EC, and has acted as a technology advisor and consultant for numerous commercial and public organisations.

    Mark Parsons is a Professor of High Performance Computing at the University of Edinburgh and Executive Director of EPCC, the supercomputing centre at the university. Over the past 20 years he has led a wide variety of HPC and distributed computing projects ranging from technology transfer projects with industry to lead-edging research projects at the forefront of numerical and data-driven computing.

    Csilla Zsigri is a multi-skilled and multi-lingual business, management and technology consultant. She is Director of Consulting Services EMEA and member of 451 Advisors, consulting division of The 451 Group, a leading global analyst, data and professional services company.

    Ian Morgan is a technology strategy and innovation specialist that has carried out numerous technology fore-sighting and road-mapping studies in a variety of enabling technologies and including digital technologies, Internet of Everything and Big Data. Ian has also provides business diversification and strategic growth support to entrepreneurs and start-ups.

    Ashley Stewart is a strategy consultant based in Scotland, UK. She has both academic and commercial experience, researching and conducting a number of Digital Technologies and ICT projects.

  • Table of Contents

    1 Abstract _____________________________________________________________________ 1

    2 Executive Summary ____________________________________________________________ 3

    2.1 Summary _________________________________________________________________ 3

    2.2 Zusammenfassung _________________________________________________________ 9

    2.3 Rsum analytique ________________________________________________________ 14

    3 Background and Methodology __________________________________________________ 19

    3.1 Background ______________________________________________________________ 19

    3.2 Methodology ____________________________________________________________ 20

    4 Scenarios of the Future ________________________________________________________ 23

    4.1 Background ______________________________________________________________ 23

    4.2 Megatrends and scenario coverage __________________________________________ 23

    4.3 The Digital Citizen: Its all about me __________________________________________ 25

    4.4 The Digital Nation: Its all about us ___________________________________________ 27

    4.5 Intelligent Transport: Trains and other Vehicles with Brains _______________________ 30

    4.6 Education and Research: Connected brains ____________________________________ 32

    4.7 Future Healthcare: Health and happiness in the digital age _______________________ 35

    4.8 Living with scarce resources: Renewtopia _____________________________________ 38

    4.9 Future Manufacturing: At a factory near you ___________________________________ 40

    4.10 Technology needs _______________________________________________________ 43

    5 The RTDI challenges __________________________________________________________ 47

    6 State-of-Play ________________________________________________________________ 51

    6.1 Our computing environment today ___________________________________________ 51

    6.2 High-level European SWOT analysis __________________________________________ 56

    7 Research Priorities ____________________________________________________________ 63

    7.1 Policy options and recommendations _________________________________________ 64

    7.2 Research programme recommendations ______________________________________ 66

    7.3 Game-changing and disruptive technologies at the horizon _______________________ 69

    8 Roadmap ___________________________________________________________________ 71

    8.1 Scenario-specific roadmaps _________________________________________________ 71

    8.2 Combined European Roadmap ______________________________________________ 88

    Acknowledgements ______________________________________________________________ 89

  • NGC Roadmap Study 1

    1 Abstract

    Over the years computing has evolved to nearly always on web-based mobile computing devices. In the near future we can expect that hardware will become a commodity and the value will be in the software to drive it and the data it generates. The data deluge will require an infrastructure that can transfer and store the data, and computing systems that can analyse and extract value from data in real time. There are arguments suggesting that the computing sector will become increasingly polarised between small application-specific computing units that connect to provide system services, and larger more powerful units that will be required to analyse large volumes of data in real time.

    This report presents a vision of next generation computing for the next 10-15 years. It does this by developing a number of visionary scenarios covering key areas of everyday life. Starting from these scenarios, we present a series of technology roadmaps, associated research / development / innovation challenges and recommendations for Europe to exploit the opportunities offered by the next generation of computing.

    Seven scenarios were carefully developed to address critical aspects of society and economy. Describing how computing will evolve in each of the scenarios has allowed us to describe a series of technology needs that, by considering Europes current strengths and weaknesses in computing, we could translate into research and innovation challenges for Europe, and into value creation opportunities for the European industry.

  • NGC Roadmap Study 2

  • NGC Roadmap Study 3

    2 Executive Summary

    2.1 Summary

    2.1.1 Key Messages

    Parallel hardware is now mainstream, but parallel software is not. While all consumer CPUs are now multi-core, software is still designed as mainly sequential. The parallelisation of legacy code is very expensive and requires developers with skills in both computer architecture and application domain. European industry needs a new generation of tools for writing software, backed by innovative programming models. New tools should be natively parallel and allow for optimisation of code at run-time across the multiple dimensions of performance, reliability, throughput, latency and energy consumption while presenting the appropriate level of abstraction to developers. Innovative business models may be needed in order to make the development of new generation tools economically viable.

    High-performance computing meets cyber-physical systems. Applications in automation, aerospace, automotive and manufacturing require computing power which was typical of supercomputers a few years ago, but with constraints on size, power consumption and guaranteed response time which are typical of the embedded applications. This is a market opportunity to build upon the existing strength of European industry to develop a family of innovative and scalable technologies, powering computing devices ranging from the embedded micro-server to the large data centre.

    Internet of Everything is developing fast. Computing applications merging automation, real-time processing of big data, autonomous behaviour and very low power consumption are changing the physical world we live in, and creating new areas of application like e.g. smart cities, smart homes, etc Data locality is becoming an issue, driving the development of multi-level applications which see processing and data shared between local/mobile devices and cloud-based servers. European industry has the know-how and innovation capacity to be a leader in this area, where issues like interoperable interfaces, privacy and data sharing rules will play a very important role in the development of the market.

    2.1.2 Scenarios for next generation computing

    Over the years computing has evolved to nearly always on web-based mobile computing devices. In the near future we can expect that hardware will become a commodity and the value will be in the software to drive it and the data it generates. The data deluge will require an infrastructure that can transfer and store the data, and computing systems that can analyse and extract value from data in real time. There are arguments suggesting that the computing sector will become increasingly polarised between small application-specific computing units that connect to provide system services, and larger more powerful units that will be required to analyse large volumes of data in real time.

    This report presents a vision of next generation computing for the next 10-15 years. It does this by developing a number of visionary scenarios covering key areas of every day's life.

    Starting from these scenarios, we present a series of technology roadmaps, associated research / development / innovation challenges and recommendations for Europe to exploit the opportunities offered by the next generation of computing.

  • NGC Roadmap Study 4

    Stakeholders throughout Europe were involved in the study through direct contacts and through two separate workshops providing a validation and refinement of the recommendations arising from this study.

    Seven scenarios were carefully developed to address critical aspects of society and economy. Describing how computing will evolve in each of the scenarios has allowed us to describe a series of technology needs that, by considering Europes current strengths and weaknesses in computing, we could translate into research and innovation challenges for Europe, and into value creation opportunities for the European industry.

    Table 1: Overview of scenarios

    A common theme across all scenarios is the need for small low-cost and low-power computing systems that are fully interconnected, self-aware, context-aware and self-optimising within application boundaries.

    A key element of the value chain is in software and programming methodologies, and this builds on existing strengths in Europe. Moreover, with a particular know-how in industrial and embedded multicore systems, Europe is well placed to support the growth of the infrastructure needed to transfer, store and analyse large volumes of data in real time. This will form the backbone to support the growth of the Internet of Everything and the next generation of Cloud Computing.

    Scenario Focus

    Its All About Me Empowering the individual citizen

    Its All About Us Communities and how they collaborate

    Trains and other Vehicles with Brains Making transport more efficient

    Connected Brains Research, education and knowledge sharing

    Health & Happiness Health and social well being

    Renewtopia Sustainability, Energy and resource management

    At a Factory Near You Manufacturing in the future

  • NGC Roadmap Study 5

    2.1.3 Areas of opportunity

    In this context, we can identify several areas of opportunity for computing in Europe, where investment in research and development can generate significant economic value in terms of exploitation.

    Cyber-physical systems

    Building on existing strengths in embedded / cyber physical systems and increasing research in this area will allow maintaining European leadership. Efforts will be focused on implementing a vision of smart networked cyber physical systems, based on manycore low-power architectures and powered by natively parallel software.

    Software

    Research priorities should include autonomous systems together with dynamic and configurable computing including context-aware, self-optimising software and dependable systems. A strong effort is needed on programming models and tools for next-generation systems including native parallel programming and multi-dimensional optimisation (energy, throughput, response time, reliability, resilience). Productivity in parallel software development should be greatly improved, while limiting the need for developers to be skilled in both low-level computer architecture and high-level application domain. Innovative programming models and tools should provide the right level of abstraction to make parallel programming less expensive and more agile. This area will potentially provide strong economic value.

    Energy

    Europe is a global leader in energy efficiency and this will be a key requirement in next generation computing. Improvements will apply across the whole computing continuum, from the high performance data centres required to exploit the opportunities offered by big data to the small computing devices used in mobile and embedded applications powering the Internet of Things. High-performance computing will range from cyber-physical systems to industrial and scientific applications, with a variety of solutions scaling across different computing powers but sharing energy efficiency concerns. The economic value of energy efficient computing is potentially very high because it can enable applications which are otherwise not possible in very diverse fields like health, environment, and automation.

    Computer interfaces

    Advanced human-computer interfaces will become increasingly important and will support natural and immersive interfaces such as mixed-reality devices. These interfaces will require adequate high performance and real time computing power, as well as research into health, behavioural and psychological issues to humanise our relationship with this new ubiquitous computing landscape. Similarly, advances in security will need computing power to analyse large quantities of data in real time in order to identify threats and provide mitigation actions, and to guarantee the appropriate level of data privacy in different usage scenarios.

    In this world with large amounts of shared and open data easily available, there will be an increasing need for tools and methodologies to address privacy and security issues. Data and its openness,

  • NGC Roadmap Study 6

    consistency and governance will become an area of increasing importance that will either be a key enabler or a high barrier for the effective exploitation of next generation computing.

    Another cross cutting area will be the need for industry wide interoperable interfaces for data and services, which will be needed for the effective development of the Internet of Everything.

    There are also several other areas that are arguably outside the timeframe considered in this report, but have nevertheless the potential to become disruptive forces for the next generation of computing.

    Quantum computing although talked about for many years is still in its infancy but has the potential to solve large scale computational problems orders of magnitude faster than existing systems.

    New energy sources, computing architectures and energy harvesting and storage systems that allow cyber physical systems to works for years with no need for recharge; this will have revolutionary impact on computing, making always on devices economically feasible and opening the door to disruptive applications.

    Smart materials beyond implantable and wearable computing, technologies such as printed electronics, biomaterials and graphene have the potential to radically change both the human-computer interface and the way in which computers interact with the physical world.

    2.1.4 Policy support

    At a policy level, work is needed to ensure that the infrastructure and legislative environment in Europe provides the right conditions for uptake of computing technologies.

    Easily understood policies on data openness, governance, privacy and sharing, especially across borders, will have to be developed with the collaboration of all involved stakeholders (citizens, local and national governments, industry, SMEs).

    The Internet of Everything requires a fully connected society and further work is required to develop faster and cheaper internet access, especially mobile access, across Europe. The high costs for large data transfers, the limited coverage of fast mobile networks, and the cost of cross-border data roaming are not compatible with the development of the market.

    European technology and service providers should be encouraged to cross sell and share technologies across multiple markets to maximise convergence and address common challenges.

    European governments should lead by example in the openness and sharing of data to stimulate innovation in areas such as public services, energy, environment and Health, by promoting interoperable interfaces and a sustainable market place for services.

    Further work is still needed to build an innovation eco-system that brings together academia, industry, entrepreneurs and funding organisations, this should also consider cross sector / application collaboration.

  • NGC Roadmap Study 7

    Technological and societal changes will create opportunities for new business models, possibly based on open collaboration and on innovative ways of connecting the actors in the value chain. The infrastructure and legislative environment must be able to support and enable value creation both from traditional industrial actors and from new actors like citizens, non-profit organisations, local governments, prosumers, micro-enterprises.

    Europe is well positioned to benefit from the opportunities that will arise as we move towards the next generation of computing and the society that it will enable. But there will be an increasing challenge for Europe to put in place common communication standards, a policy on open and shared data, seamless cross-border mobile telecommunications services and privacy and security measures that will create the environment for next generation computing and its applications to flourish.

  • NGC Roadmap Study 8

    2.1.5 Combined roadmap

    As a summary, below you can find the combined roadmap for European research derived from the interviews, the desk research, the online-consultation and the experts input at the workshops. It expands the concepts and ideas which are briefly described in this executive summary; the methodology to build it, and its contents, are fully explained in chapters Research priorities and Roadmap of this report.

  • NGC Roadmap Study 9

    2.2 Zusammenfassung

    2.2.1 Kernaussagen

    Parallele Hardware liegt im Trend, nicht aber Parallele Software. Whrend die CPUs fr Konsumenten jetzt multi-core sind, wird Software noch immer vor allem sequentiell entworfen. Die Parallelisierung von Altcode ist sehr teuer und erfordert Entwickler mit Kenntnissen auf dem Gebiet der Computerarchitektur als auch im Anwendungsgebiet. Die Europische Industrie bentigt eine neue Generation an Werkzeugen fr die Softwareentwicklung untersttzt durch innovative Programm-Modelle. Neue Werkzeuge sollten grundlegend parallel sein und die Optimierung des Codes zur Laufzeit ber verschiedene Leistungsdimensionen wie Schnelligkeit, Zuverlssigkeit, Durchsatz, Latenz und Energieverbrauch hinweg erlauben und zugleich die passende Abstraktionsebene fr Entwickler darstellen. Dabei knnen auch innovative Geschftsmodelle ntig werden, um die Entwicklung einer neuen Werkzeuggeneration wirtschaftlich gangbar zu machen.

    Hochleistungsrechnen und cyber-physikalische Systeme. Anwendungen in Automatisierung, Luftfahrt, Automobilbau und Produktion erfordern eine Rechenleistung, die vor wenigen Jahren fr Superrechner typisch war; allerdings mit Einschrnkungen hinsichtlich Gre, Leistungsverbrauch und garantierter Antwortzeit, die typisch fr integrierte Anwendungen sind. Aufbauend auf bestehenden Strken der europischen Industrie stellt dies eine Marktchance dar, um eine Familie innovativer und skalierbarer Technologien zu entwickeln, die Gerte ermglichen, die von integrierten Mikroservern bis zu groen Datenzentren reichen.

    Das Internet fr Alles entwickelt sich schnell. Computeranwendungen, die Automatisierung, Echtzeitverarbeitung von groen Daten (Big Data), autonomes Verhalten und sehr niedrigen Energieverbrauch verbinden, verndern die physische Welt in der wir leben und erzeugen so neue Anwendungsgebiete, wie z.B. Smart Cities, Smart Homes etc. Datenlokalitt wird dabei ein wichtiges Thema, das die Entwicklung bei Mehrebenen-Anwendungen vorantreibt, bei denen Verarbeitung und Daten zwischen lokalen/mobilen Gerten und Cloud-basierten Servern aufgeteilt wird. Die Europische Industrie verfgt ber das Knowhow und die Innovationskraft, um auf diesem Gebiet fhrend zu sein. Hier werden Themen wie interoperable Interfaces, Datenschutz und Regeln fr den Datenaustausch eine wichtige Rolle fr die Marktentwicklung spielen.

    2.2.2 Szenarien fr die EDV der nchsten Generation

    EDV hat sich ber die Jahre hin zu dauernd verfgbaren web-basierten mobilen Gerten entwickelt. In der nahen Zukunft knnen wir erwarten, dass Hardware ein Gebrauchsgut wird und dass der Wert in der Software liegen wird, diese zu betreiben, sowie in den dabei generierte Daten. Die Datenflut wird eine Infrastruktur erforderlich machen, welche Daten bertragen und speichern kann, sowie Rechnersysteme, die Daten analysieren knnen und Mehrwert aus den Daten in Echtzeit generieren.

  • NGC Roadmap Study 10

    Es gibt Argumente, die nahe legen, dass der EDV-Sektor zunehmend polarisiert wird zwischen kleinen anwendungsspezifischen Recheneinheiten, die Systemdienste zur Verfgung stellen knnen und greren, leistungsfhigeren Einheiten, die fr die Analyse groer Datenmengen in Echtzeit ntig sein werden.

    Dieser Bericht stellt eine Vision fr die EDV der nchsten Generation fr die kommenden zehn bis fnfzehn Jahre vor. Dies geschieht durch verschiedene visionre Szenarien, die wesentliche Gebiete des tglichen Lebens abdecken. Ausgehend von diesen Szenarien stellen wir eine Reihe von Technologie-Roadmaps mit den zugehrigen Herausforderungen in Forschung, Entwicklung und Innovation vor und prsentieren Empfehlungen fr Europa, um die Chancen, die sich aus der EDV der nchsten Generation ergeben, zu nutzen.

    Stakeholder in Europa waren direkt bzw. in zwei Workshops in die Erstellung dieser Studie involviert und haben so die sich ergebenden Empfehlungen validiert und verfeinert.

    Um die wichtigen gesellschaftlichen und wirtschaftlichen Aspekte anzusprechen, wurden sieben Szenarios genauer entwickelt. Die Beschreibung der Evolution der EDV in jedem der Szenarios hat es uns mglich gemacht, eine Reihe von Technologieanforderungen zu beschreiben, die wir unter Bedacht auf Europas derzeitige Strken und Schwchen in der EDV in Forschungs- und Innovationsherausforderungen fr Europa bersetzen konnten und damit in Chancen fr die Schaffung von Mehrwert fr die europische Industrie.

    Tabelle 2: berblick ber die Szenarien

    Ein gemeinsames Thema ber alle Szenarien hinweg ist der Bedarf an kleinen, gnstigen Rechensystemen mit geringem Energiebedarf, die vollverbunden, ich-bewusst, kontextbasiert und selbst-optimierend innerhalb der Anwendungsgrenzen arbeiten.

    Ein wesentliches Element der Wertschpfungskette stellen Software und Programmiermethoden dar. Dies baut auf Europischen Strken auf. Auerdem ist Europa mit seinem speziellen Knowhow in industriellen und integrierten Multicore-Systemen gut positioniert, um das Wachstum der fr den Transfer, die Speicherung und die Analyse von groen Datenmengen in Echtzeit bentigten Infrastruktur zu frdern. Dies wird das Rckgrat bilden, um das Wachstum des Internet der Dinge und der nchsten Generation des Cloud Computing zu untersttzen.

    Scenario Fokus

    Its All About Me Ermchtigung des einzelnen Brgers

    Its All About Us Gemeinschaften und wie sie zusammenarbeiten

    Trains and other vehicles with brains Effizienzverbesserung im Verkehr

    Connected Brains Forschung, Ausbildung und Wissen teilen

    Health & Happiness Gesundheit und soziales Wohlbefinden

    Renewtopia Nachhaltigkeit, Energie und Ressourcenmanagement

    At a Factory Near You Produktion der Zukunft

  • NGC Roadmap Study 11

    2.2.3 Chancen

    In diesem Zusammenhang knnen wir verschiedene Chancen fr die EDV in Europa identifizieren, in denen Investitionen in Forschung und Entwicklung bedeutenden wirtschaftlichen Vermarktungswert generieren knnen.

    Cyber-physikalische Systeme

    Auf bestehenden Strken in integrierten / cyber-physikalischen Systemen aufzubauen und Forschung in diesem Bereich auszudehnen ermglicht die fortgesetzte Fhrung Europas in diesem Bereich. Anstrengungen werden auf die Implementierung einer Vision von smarten, vernetzten, cyber-physikalischen Systemen gebndelt, und zwar auf der Grundlage von Mehrkern-, Niedrigenergie-Architekturen und getrieben durch inhrent parallele Software.

    Software

    Forschungsprioritt sollte auf autonomen Systemen liegen gemeinsam mit dynamischer und konfigurierbarer EDV inklusive kontextabhngiger, selbst-optimierender Software und zuverlssigen Systemen. Verstrkte Anstrengungen sind im Bereich von Programmiermodellen und Werkzeugen fr die nchste Generation von Systemen ntig inklusive inhrent paralleler Programmierung und mehrdimensionaler Optimierung (Energie, Durchsatz, Antwortzeit, Zuverlssigkeit, Ausfallsicherheit). Die Produktivitt auf dem Gebiet paralleler Softwareentwicklung sollte wesentlich verbessert werden, whrend die Notwendigkeit fr Programmierer, Kenntnisse in low-level Computerarchitektur und high-level Anwendungsdomnen zu haben, beschrnkt bleiben sollte. Innovative Programmiermodelle und -werkzeuge sollten das richtige Abstraktionsniveau zur Verfgung stellen, um paralleles Programmieren billiger und flexibler zu machen. Dieses Gebiet kann einen hohen wirtschaftlichen Wert darstellen.

    Energie

    Europa ist weltweit fhrend auf dem Gebiet der Energieeffizienz. Dies wird eine wesentliche Anforderung in EDV Systemen der nchsten Generation darstellen. Verbesserungen werden den gesamten Bereich der EDV betreffen, von Hochleistungs-Datenzentren fr die Nutzung von Chancen durch Massendaten bis hin zu kleinen Gerten, die in mobilen und integrierten Anwendungen eingesetzt werden und das Internet der Dinge ermglichen. Hochleistungsrechnen wird von cyber-physikalischen Systemen bis zu industriellen und wissenschaftlichen Anwendungen reichen, mit einer Vielzahl an Lsungen, die ber verschiedene Rechnerleistungsstufen skalierbar sind und zugleich dem Gedanken an Energieeffizienz gerecht werden. Der wirtschaftliche Wert energieeffizienten Rechnens ist potenziell sehr hoch, weil dieses Anwendungen ermglicht, die andernfalls in verschiedenen Gebieten nicht mglich sind, z.B. im Bereich Gesundheit, Umwelt und Automatisierung.

  • NGC Roadmap Study 12

    Interfaces

    Fortschrittliche Mensch-Computer Schnittstellen werden immer wichtiger und werden natrliche und sinnlich stimulierende Schnittstellen ermglichen, wie z.B. mixed-reality Gerte. Diese Schnittstellen erfordern eine adquat hohe Rechenleistung und Echtzeitverarbeitung; ntig ist aber auch Forschung im Bereich Gesundheit, des Verhaltens und der Psychologie, um unser Verhltnis zu dieser neuen allumfassenden Rechnerlandschaft menschlicher zu machen. In hnlicher Weise bedrfen Fortschritte auf dem Bereich der Sicherheit jener Rechenleistung, um groe Datenmengen in Echtzeit zu analysieren und so Bedrohungen zu identifizieren und Gegenmanahmen einzuleiten und um das richtige Ma an Datenschutz in unterschiedlichen Nutzungsszenarien zu garantieren.

    In unserer Welt, in der eine groe Zahl von Daten geteilt und offen frei verfgbar ist, wird es einen steigenden Bedarf fr Werkzeuge und Methoden geben, um Themen wie Datenschutz und Datensicherheit anzusprechen. Daten und deren Offenheit, Konsistenz und Regulierung wird ein Gegenstand von steigender Wichtigkeit sein, der entweder ein wesentlicher Faktor oder eine groe Barriere fr die effektive Nutzung der EDV der nchsten Generation sein wird. Ein weiteres Querschnittsthema stellt der Bedarf an interoperablen Schnittstellen fr Daten und Services fr die Industrie dar, die fr die effektive Entwicklung des Internet fr Alles ntig sind.

    Einige andere Gebiete liegen vielleicht auerhalb des anvisierten Zeithorizonts fr diesen Bericht, haben aber dennoch das Potenzial, Strkrfte fr die EDV der nchsten Generation zu werden:

    Quantenrechnen obwohl darber schon seit vielen Jahren gesprochen wird - steckt immer noch in den Kinderschuhen, hat aber das Potenzial, viele groe Berechnungsprobleme um Grenordnungen schneller zu lsen als existierende Systeme

    Neue Energiequellen, Computerarchitekturen und Energieernte- und Speichersysteme, die es cyber-physikalischen Systemen erlauben, ber Jahre hinweg ohne Aufladen zu arbeiten; dies kann zu Revolutionen in Computersystemen fhren und stndig verfgbare Gerte wirtschaftlich verfgbar machen, und damit die Tr zu neuen Anwendungen ffnen

    Intelligente Materialien jenseits implantierbarer und tragbarer Gerte, Technologie wie z.B. druckbare Elektronik, Biomaterialien und Graphen haben das Potenzial, um sowohl die Mensch/Maschine Schnittstelle radikal zu verndern als auch die Art, auf die Computer mit der physischen Welt zusammenarbeiten.

  • NGC Roadmap Study 13

    2.2.4 Politikuntersttzung

    Auf politischer Ebene ist die Sicherstellung von Infrastruktur und eines legislativen Rahmens in Europa ntig, welche die richtigen Bedingungen fr das Aufgreifen von Computertechnologie sicherstellen.

    Es sind einfach verstndliche Regelungen auf dem Gebiete offener Daten, Regulierung, Datenschutz und Datenaustausch - besonders ber Grenzen hinweg - in Zusammenarbeit aller involvierten Stakeholder (Brger, lokale und nationale Regierungen, Industrie und KMUs) zu entwickeln.

    Das Internet der Dinge erfordert eine voll vernetzte Gesellschaft. Weitere Arbeiten sind erforderlich, um den Internetzugang schneller und gnstiger zu machen, insbesondere der mobile Zugang in ganz Europa. Die hohen Kosten fr die bertragung groer Datenmengen, die beschrnkte Abdeckung durch schnelle mobile Netze und die Kosten von grenzberschreitendem Datenroaming stehen nicht mit der Marktentwicklung im Einklang.

    Europische Technologie- und Serviceprovider sind gefordert, Technologien ber mehrere Mrkte hinweg zu verkaufen und untereinander zu teilen, um die Konvergenz zu maximieren und gemeinsamen Herausforderungen zu begegnen.

    Europische Regierungen sollten auf dem Gebiet offener und geteilter Daten mit gutem Beispiel vorangehen, um so Innovation in Bereichen wie ffentlichen Diensten, Energie, Umwelt und Gesundheit zu stimulieren indem interoperable Schnittstellen und ein nachhaltiger Markt fr Dienste gefrdert werden.

    Weitere Anstrengungen sind ntig, um ein Innovations-kosystem zu schaffen, das wissenschaftliche Einrichtungen, Industrie, Unternehmer und Frderorganisationen zusammenbringt; dies betrifft auch die Zusammenarbeit ber verschiedene Sektoren und Anwendungen hinweg.

    Technologische und gesellschaftliche Vernderungen werden Chancen fr neue Geschftsmodelle schaffen, mglicherweise basierend auf offener Zusammenarbeit und auf innovativen Arten, die Akteure in der Wertschpfungskette zusammenzubringen. Die Infrastruktur und der rechtliche Rahmen mssen in der Lage sein, die Wertschpfung sowohl durch hergebrachte Industrieakteure als auch durch neue Akteure wie Brger, Nichtregierungsorganisationen, lokale Regierungen, Prosumer und Kleinstunternehmer zu ermglichen.

    Europa ist gut positioniert, um von den Chancen zu profitieren, die sich bieten whrend wir uns hin zur nchsten Generation von Rechensystemen und die Gesellschaft, die sie mglich macht, bewegen. Aber die Herausforderungen fr Europa nehmen zu: Es sind gemeinsame Kommunikationsstandards, Regelungen fr offene und geteilte Daten, nahtlose grenzberschreitende mobile Telekommunikation und Datenschutz- und Datensicherheitsregeln aufzustellen, welche die Bedingungen fr die EDV der nchsten Generation und fr florierende Anwendungen schaffen.

  • NGC Roadmap Study 14

    2.3 Rsum analytique

    2.3.1 Messages cls

    Alors que le paralllisme du matriel informatique est dsormais incontournable, les logiciels parallles ne le sont toujours pas. Alors que les CPU du commerce sont des processeurs multi-curs, le logiciel est encore conu de faon squentielle. La paralllisation des logiciels est trs coteuse et ncessite des dveloppeurs dots de comptences aussi bien en architecture informatique que dans le domaine des applications. L'industrie europenne a besoin d'une nouvelle gnration d'outils de dveloppement logiciel, soutenue par des modles de programmes novateurs. Les nouveaux outils doivent d'emble tre parallles et garantir l'optimisation des codes l'excution, en termes de performance, fiabilit, dbit de traitement, temps d'attente et consommation d'nergie, tout en prsentant le niveau adquat d'abstraction aux dveloppeurs. De nouveaux modles commerciaux peuvent s'avrer ncessaires pour rendre le dveloppement des outils de nouvelle gnration conomiquement viable.

    Le calcul haute performance au service des systmes cyber-physiques. Dans les industries de l'automatisation, de l'aronautique et de l'automobile ainsi que dans le secteur manufacturier, les applications ncessitent des capacits informatiques caractristiques des super ordinateurs d'il y a quelques annes, mais avec les contraintes de taille, de consommation d'nergie et de temps de rponse garanti typiques des applications embarques. Il s'agit l d'une opportunit pour dvelopper sur la base des forces existantes de l'industrie europenne une nouvelle gnration de technologies novatrices et volutives allant des micro-serveurs embarqus aux grands centres de calcul.

    L'Internet des objets se dveloppe de manire acclre. Les applications informatiques combinant la fois automatisation, traitement en temps rel de gros volumes de donnes, comportement autonome et trs faible consommation d'nergie permettent de changer le monde physique dans lequel nous vivons et de crer de nouveaux domaines d'application, l'instar des villes et des maisons intelligentes , etc. La localisation des donnes devient un facteur important pour favoriser le dveloppement d'applications multi-chelles permettant le traitement et le partage des donnes entre les appareils locaux/mobiles et les serveurs bass sur le cloud computing (informatique en nuage). L'industrie europenne dispose du savoir-faire et des capacits d'innovation ncessaires pour faire d'elle le leader dans les domaines o les questions relatives aux interfaces interoprables et aux rgles de confidentialit et de partage des donnes joueront un trs grand rle dans le dveloppement du march.

    2.3.2 Scnarios de l'informatique de nouvelle gnration

    Au fil des ans, l'informatique a volu, crant des appareils informatiques mobiles presque toujours connects et bass sur le web. Dans un avenir proche, on peut s'attendre ce que le matriel ne devienne qu'un simple produit et que la valeur ajoute rside dans le logiciel, tout comme dans les donnes qu'il gnre. Ce dluge de donnes ncessitera une infrastructure qui pourra transfrer et sauvegarder les donnes, ainsi que des systmes informatiques capables d'analyser et d'extraire en temps rel la valeur de ces donnes. Certains sont d'avis que le secteur informatique deviendra de plus en plus focalis sur de petits centres informatiques ddis des applications spcifiques pouvant se connecter entre eux pour offrir des services cibls, et sur des centres plus grands et plus puissants ncessaires pour l'analyse en temps rel de grands volumes de donnes.

  • NGC Roadmap Study 15

    Le prsent rapport prsente une vision de l'informatique de nouvelle gnration pour les 10 15 prochaines annes. Il est fond sur un certain nombre de scnarios visionnaires couvrant les principaux domaines de la vie quotidienne. Partant de ces scnarios, nous prsentons une srie de feuilles de routes sur les technologies, les dfis qui y sont associs en termes de recherche / de dveloppement / d'innovation, ainsi que des recommandations pour l'exploitation, par l'Europe, des opportunits offertes par la prochaine gnration de l'informatique.

    Des intervenants de toute lEurope ont t impliqus dans cette tude, par le biais de contacts directs et via deux sminaires, afin de valider et d'affiner les recommandations dcoulant de cette tude.

    Sept scnarios ont t minutieusement labors dans le cadre de la rsolution des principaux problmes sociaux et conomiques. En dcrivant l'volution de l'informatique dans chacun des scnarios, nous avons pu prsenter un ensemble de besoins technologiques qui, compte tenu des forces et faiblesses actuelles de l'Europe dans le domaine de l'informatique, peuvent se traduire par des dfis auxquels l'Europe est confronte en matire de recherche et d'innovation, et par des possibilits de cration de valeur pour l'industrie europenne.

    Tableau 3: Aperu des scnarios

    Nom du scenario Centre dintrt

    Its All About Me (Il ne sagit que de moi) Renforcement de lautonomie du citoyen

    Its All About Us (Il ne sagit que de nous) Communauts et leurs modes de collaboration

    Trains and other vehicles with brains (Transport et trains intelligents)

    Rendre les transports plus efficaces

    Connected Brains (Cerveaux connects) Recherche, ducation et partage des connaissances

    Health & Happiness (Sant et Bonheur) Sant et bien-tre social

    Renewtopia Viabilit, gestion de lnergie et des ressources

    At a Factory Near You (Dans une usine prs de vous) Fabrication dans lavenir

    L'un des thmes communs tous les scnarios est la ncessit de petits systmes informatiques peu coteux et faible consommation nergtique, entirement interconnects, sensibles au contexte et capables d'auto-optimisation dans les limites de lapplication.

    Un lment important de la chane de valeur rside dans le logiciel et les mthodes de programmation, bass sur les forces existantes de l'Europe. En plus, avec un savoir-faire exceptionnel dans le domaine des systmes multi-curs industriels et intgrs, l'Europe est bien place pour favoriser le dveloppement des infrastructures ncessaires au transfert, la sauvegarde et l'analyse en temps rel de gros volumes de donnes. Ceci constituera l'ossature de la mise en place de l Internet des objets et du cloud computing (informatique en nuage) de nouvelle gnration.

  • NGC Roadmap Study 16

    2.3.3 Domaines d'opportunits

    Dans ce contexte, nous pouvons identifier plusieurs domaines d'opportunits en matire d'informatique en Europe, dans lesquels l'investissement dans la recherche et le dveloppement peut gnrer une valeur conomique considrable.

    Systmes cyber-physiques

    L'utilisation des forces existantes dans le domaine des systmes embarqus / systmes cyber-physiques et l'intensification des recherches dans ce domaine permettront l'Europe de garder son leadership. Les efforts seront concentrs sur la mise en uvre d'une vision de systmes cyber-physiques en rseau, intelligents, bass sur des architectures multi-curs faible consommation fonctionnant avec des logiciels parallles natifs.

    Logiciel

    Les priorits de recherche doivent inclure les systmes autonomes ainsi que l'informatique dynamique et reconfigurable, y compris les systmes sensibles au contexte, auto-optimisants et fiables. De grands efforts doivent tre raliss dans le cadre des modles et outils de programmation pour les systmes de nouvelle gnration, notamment dans la programmation parallle et l'optimisation multidimensionnelle (nergie, dbit de traitement, temps de rponse, fiabilit, rsilience). La productivit en matire de dveloppement de logiciels parallles doit tre considrablement amliore, rduisant ainsi le besoin pour les dveloppeurs d'tre qualifis aussi bien en architecture des ordinateurs que dans le domaine des applications de haut niveau. Les modles et outils de programmation novateurs doivent apporter le niveau d'abstraction adquat pour faire de la programmation parallle une activit moins coteuse et plus flexible. Ce domaine sera probablement d'une valeur conomique considrable.

    nergie

    L'Europe est un leader mondial en termes d'efficacit nergtique, principale exigence pour l'informatique de nouvelle gnration. Des progrs seront apports sur l'ensemble du continuum de l'informatique, allant des centres de calcul haute performance ncessaires l'exploitation des opportunits offertes par les gros volumes de donnes, aux petits appareils informatiques utiliss pour les applications mobiles et embarques alimentant l'Internet des objets. Le calcul haute performance va stendre des systmes cyber-physiques aux applications industrielles et scientifiques, avec un ventail de solutions utilisant diffrents niveaux de puissance de calcul, mais ayant en commun les proccupations lies l'efficacit nergtique. La valeur conomique de l'informatique co-nergtique est potentiellement trs leve, car elle permet de nouvelles applications dans divers domaines, tels que la sant, l'environnement et l'automatisation.

  • NGC Roadmap Study 17

    Interfaces informatiques

    Le perfectionnement des interfaces homme-machine sera d'une importance considrable et rendra possible des interfaces naturelles et immersives, tels que les priphriques de ralit augmente. Ces interfaces ncessiteront une puissance de calcul leve et en temps rel, tout comme des recherches dans le domaine de la sant, de la psychologie et du comportement, afin d'humaniser notre relation avec ce nouveau paysage de l'informatique omniprsente. De mme, les progrs raliser en matire de scurit ncessiteront des capacits informatiques pour l'analyse en temps rel de grands volumes de donnes, en vue d'identifier les menaces, de prendre les mesures correctives ncessaires, et de garantir le niveau appropri de confidentialit des donnes dans les diffrents domaines d'utilisation.

    Dans ce monde dot de grands volumes de donnes partages et librement accessibles, le besoin en outils et mthodes de rsolution des questions lies la confidentialit et la scurit s'avrera considrable. Les donnes, ainsi que leur disponibilit, cohrence et gouvernance, deviendront un domaine de haute importance qui sera soit un catalyseur cl, soit un obstacle notoire pour l'exploitation effective de l'informatique de nouvelle gnration. Un autre domaine intersectoriel sera la ncessit d'interfaces interoprables pour les donnes et services l'chelle industrielle, lment cl du dveloppement effectif de l' Internet des objets .

    Il existe galement plusieurs autres domaines qui sont sans doute en dehors des dlais spcifis dans le prsent rapport, mais qui ont nanmoins le potentiel de devenir des forces perturbatrices pour l'informatique de nouvelle gnration.

    Linformatique quantique bien que dbattue depuis de nombreuses annes - en est encore ses dbuts, mais offre des perspectives prometteuses, pour rsoudre des problmes informatiques d'envergure, plusieurs ordres de grandeur plus rapide que les systmes existants.

    De nouvelles sources d'nergie, de nouvelles architectures informatiques et de nouveaux systmes de production et de stockage d'nergie permettront aux systmes cyber-physiques de fonctionner pendant des annes sans avoir besoin d'tre rechargs. Ceci aura un impact rvolutionnaire sur l'informatique, rendra conomiquement viables les dispositifs always on ( connexion permanente) et ouvrira la porte des applications novatrices.

    Matriaux intelligents : au-del de l'informatique vestimentaire et implantable, des technologies l'instar des systmes lectroniques imprims, des biomatriaux et des graphnes sont capables de changer radicalement aussi bien l'interface homme-machine que la manire dont les ordinateurs communiquent avec le monde physique.

  • NGC Roadmap Study 18

    2.3.4 Appui des politiques

    Au niveau des politiques, beaucoup doit tre fait afin d'assurer que les infrastructures et le cadre lgislatif en Europe fournissent les meilleures conditions pour l'adoption des technologies informatiques.

    Des politiques faciles comprendre relatives l'ouverture, la gouvernance, la confidentialit et le partage des donnes, particulirement au-del des frontires, doivent tre mises en place avec la collaboration de tous les acteurs concerns (citoyens, gouvernements locaux et nationaux, industries, PME).

    L' Internet des objets ncessite une socit entirement connecte et des efforts supplmentaires sont ncessaires pour rendre l'Internet haut dbit moins coteux et accessible, particulirement pour l'Internet mobile dans toute l'Europe. Les cots levs des transferts de grands volumes de donnes, la couverture limite des rseaux mobiles haut dbit et le cot des services de donnes en itinrance ne sont pas compatibles avec le dveloppement de ce march.

    Les fournisseurs de technologie et de services europens doivent tre encourags dans la pratique des ventes croises et le partage des technologies dans plusieurs marchs, afin de maximiser la convergence et de faire face aux dfis communs.

    Les gouvernements europens doivent montrer l'exemple dans le domaine de l'ouverture et du partage des donnes, de manire stimuler l'innovation dans des secteurs tels que les services publics, l'nergie, l'environnement et la sant, en promouvant des interfaces interoprables et un march durable pour les services.

    Des efforts supplmentaires sont ncessaires pour la mise en place d'un cosystme novateur qui rassemblera monde universitaire, industrie, entrepreneurs et organismes de financement, tout en prenant en compte la collaboration intersectorielle / inter-applicative.

    Les changements technologiques et socitaux offriront des opportunits pour de nouveaux modles d'entreprises, probablement bass sur une collaboration ouverte et sur des mthodes novatrices de connexion des acteurs de la chane de valeur. Les infrastructures et le cadre lgislatif doivent tre en mesure de prendre en charge et de favoriser la cration de valeur, aussi bien par les acteurs industriels traditionnels que par de nouveaux acteurs, comme les citoyens, les organisations but non lucratif, les gouvernements locaux, les prosommateurs et les micro-entreprises.

    L'Europe est bien place pour tirer profit des opportunits qui se prsenterons au fur et mesure que nous avancerons vers l'informatique de nouvelle gnration et vers la socit qu'elle engendra. Toutefois, l'Europe sera confronte un dfi majeur dans la mise en place de normes de communication communes, dune politique sur les donnes ouvertes et partages, de services de tlcommunication mobile transfrontaliers uniformiss ainsi que de mesures de scurit et de confidentialit, qui, ensemble, promouvront le contexte ncessaire l'informatique de nouvelle gnration, ainsi que ses applications.

  • NGC Roadmap Study 19

    3 Background and Methodology

    3.1 Background

    Computing is undergoing a significant transformation. The chart below provides a progressive view of computing from 1970 to 2020. This illustrates the evolution of different aspects of computing such as hardware, software, service, connectivity, among others, and in light of the governing megatrends and business models.

    Figure 1: The Evolution of Computing

    This figure demonstrates that computing is moving towards an always connected, ubiquitous networked society where citizens work collaboratively, have access to low cost and low power hardware and can customize their software and apps to meet their own needs. As such we are now living in the era of personalised smart computing where tablets, smartphones and other connecting devices take center stage. Additionally, the business models of the computing industry are also undergoing transformation, as there is a significant shift towards a service oriented approach rather than direct sales.

    Paradoxically, of the major changes in ICT research and its increasingly pervasive nature can sometimes hinder a clear view of where the grand challenges are and how the future may look. Although computing has undergone dramatic developments and radical changes in the last years there is no reason to believe that its dynamics will decelerate any time soon. Computing benefits from recent advances in nanotechnology, photonics, biochemistry and other disciplines, but it is also a major driver behind these fields.

    This facilitates the continued emergence of new generations of components and systems including new and unconventional approaches of a breakthrough character.

  • NGC Roadmap Study 20

    Moreover, computing related research is not limited to the ICT sector itself but rather widely spread among other industries, scientific fields and technologies and in addition linked to radical new ideas, future technologies and breakthrough/frontier research. From decoding the DNA to simulations in particle physics, computing has become a major driver behind scientific knowledge itself. This emphasises the importance of ICT research in general and research on the future of computing in particular.

    The last 40 years have been a golden era for computing; processing, connectivity and storage have all undergone relentless growth in capability and performance, while getting cheaper to procure. The impact in economic and societal terms has been profound, and has raised high expectations among the population at large. We are however reaching an inflection point. The technologies that underpin ICT are reaching fundamental limits that will have a profound impact on our ability to make progress, while demand for more performance and capability increases steeply.

    3.2 Methodology

    The methodology is designed to provide a technology roadmap for next generation computing (NGC) that will help to direct the research themes within Horizon 2020. The methodology, the aggregation of information and coverage of various input sources - expert workshops, stakeholder interviews, online questionnaire, desk research, SWOT analysis, value chain analysis etc. is structured into work packages as depicted in Figure 2: Methodology and Work package overview. Four main work packages are driving the study: consultation, analysis, forecast and strategy. The process flow of interconnecting existing data, various input sources and analysis parts is depicted in Figure 3 and described thereafter.

    Throughout the course of the study we conducted desk research. This literature review supported the market and SWOT analysis, identification of stakeholders, understanding of supply and value chains, provided forecasts of existing and future services and applications and also suggested potential scenarios. To validate the findings from market data and to explore the computing landscape further, we conducted a number of stakeholder interviews with industry, technology industry, academia and RTDI programme managers. The interviews were semi-structured and the discussion focussed on:

    Market trends and disruptive innovations

    Game changing products and services

    Future competitive strategies

    Unsatisfied needs

    European strengths and weaknesses

    Industry ecosystems

    Collaboration and cooperative competition

    We further conducted in-depth interviews with stakeholders that were primarily small and medium businesses with headquarters based in Europe. The stakeholders had the following areas of expertise:

    Datacentres (facilities & technologies)

    Cloud Computing,

    Telecommunications & Mobile Computing

    Hardware, Systems & Servers

    Virtualisation

    Software (enterprise, proprietary, open source, simulation & visualisation)

    Energy efficiency

  • NGC Roadmap Study 21

    The interviews supported the identification of stakeholders, value chains and supply chains across a range of associated industries, forecast of existing services and future services, SWOT analysis and research and innovation challenges.

    KEY

    Industry Interviews Desk Research

    RTD Programme Mgrs Academic interviews

    Industry Technology Interviews

    Academic Interviews

    TECHNOLOGY

    WP

    1 C

    on

    sult

    atio

    nW

    P 2

    An

    alys

    isW

    P 4

    Str

    ate

    gy

    TECHNOLOGY

    WP

    3 F

    ore

    cast

    INDUSTRY

    Key stake holdersand market players

    Supply and value chain

    TechnologyNeeds

    Strengths of European Countries (research & funding)

    DEMAND & EVOLUTION

    Forecase of Existing Services

    Future Applications

    Visionary Application Scenarios

    POLICY

    European SWOT Analysis

    Research, Development and Innovation (RDI) Challenges

    Programme Scenarios(5) plus roadmaps

    S + W O + T

    Verification

    Online Survey

    Workshop 1 Workshop 2

    Interviews

    Figure 2: Methodology and Work package overview

    We also conducted an online survey. The purpose of the survey was to reach a broader audience than the interviews, and presented respondents with an opportunity have their say on what is important for computing in Europe over the next ten years and beyond.

    The survey was completed by 173 respondents. The majority were from research/academic institutions (82%), and describe their organisations as European (50%). The respondents report that they support software (70%) and consultancy (65%) (more details can be found in the appendix online questionnaire).

    Workshops were also completed to validate findings and were one of the key elements of the comparative analysis of potential options for a roadmap-based initiative on next generation computing. The first workshop pursued two main objectives:

    1. The discussion of computing related mega trends that are most likely to shape the future of computing

    2. The development of several visionary scenarios.

  • NGC Roadmap Study 22

    31 delegates attended the first workshop and were from both industrial and research organisations. The workshop was very much an active and participatory led workshop and the goal of the plenary session was to discuss uncertainties for the next 10 years.

    Breakout groups were established: one group focused on megatrends and three additional groups focused on sketching scenarios that were influenced by the plenary discussion. The study team used the outcomes of the workshop to design visionary scenarios for next generation computing.

    The aim of the second workshop was to finalise the visionary scenarios, describe next generation computing, draft research and technology requirements and collect expert stakeholder views on a research roadmap that will prepare Europe for the future.

    Figure 3: Methodological process

    The outcome of the introduced instruments such as desk research, interviews, online questionnaire, expert workshops, etc. was the base for being able to construct scenarios of the future and corresponding roadmaps as well as to present recommendations for future work programme foci as presented in the succeeding sections. For additional and more detailed information on the outcome of the introduced modules please also refer to the annex document.

  • NGC Roadmap Study 23

    4 Scenarios of the Future

    4.1 Background

    In this section we present selected visions of the future to illustrate how next generation computing might enable applications in various walks of life. The purpose of this exercise is to deduce the RTDI challenges that would have to be overcome if the scenarios are to be realised. The scenarios are based on known current technology and market trends, but also inspired by some degree of imagination (as is essential when thinking about the future). There is the assumption that new and potentially disruptive technology can be developed and will find applications of the future. There is overlap between the scenarios, which is to be expected if there are to be in any way realistic.

    The scenarios should therefore not be thought of as strong predictions of the future, nor as the ideal towards which we should be aiming, but rather as a vehicle for us to understand the potential for NGC in a variety of contexts, and as a tool to identify important RTDI topics

    The scenarios we have constructed are as follows:

    The Digital Citizen: Its all about me. This scenario is focussed on the individual as a consumer of digital services in the future.

    The Digital Nation: Its all about us. This scenario is focussed on the responsibilities of a nation state and how these could be affected by NGC.

    Intelligent Transport: Trains and other vehicles with brains. This scenario deals with intelligent transport from the user and provider viewpoints.

    Education and Research: Connected brains. This scenario investigates how NGC will enable new modes of research and learning.

    Future Healthcare: Health and happiness in the digital age. This scenario deals with the provision of healthcare from both the patient and provider viewpoints.

    Living with scarce resources: Renewtopia. This scenario covers the general trend of scarcity of resources, in particular energy, with a special emphasis on renewable energy, its generation, storage and use, and how these will both drive NGC and open up application areas that are enabled by NGC.

    Future Manufacturing: At a factory near you. This scenario is one view of how manufacturing could change in the future, the opportunities for NGC to facilitate those changes and the associated challenges.

    4.2 Megatrends and scenario coverage

    When developing the scenarios, we took into account governing megatrends, as well as more specific technology trends. Intelligence gained through in-depth interviews, desk research and online surveys revealed the following technology trends (see also annex document):

    More with less - density, energy and cost

    Software driven world

    Cloud and hybrid

  • NGC Roadmap Study 24

    Mobile computing and Internet of Things

    Open and build your own approaches

    Converged and integrated systems

    In addition to these technology trends there are more generic, global and interrelated trends that will influence future developments. These include:

    Demographic change, a notable example being the growth in the proportion of older people in developed economies.

    Increased urbanisation.

    Greater mobility of the population.

    Scarcity of resource such as materials, water, energy and viable agricultural land.

    Climate change as a result of mankinds activities.

    Rapidly evolving business models

    How the scenarios were designed around these technological and more general megatrends to cover political, economic, societal and technological areas and challenges is depicted in Figure 4. By covering various areas and being based on the outcome of the various instruments used in this study the scenarios are representing a broad range of future applications and technology areas.

    Figure 4 Scenarios are designed to cover various areas

  • NGC Roadmap Study 25

    4.3 The Digital Citizen: Its all about me

    This scenario is driven by an increasingly mobile and (almost) permanently on-line population demanding a personalised digital experience and control of their privacy. Digital citizens will expect information and services available to enrich their lives they expect the cyber landscape to work for them, help them, make them feel safe and secure, and that opens up opportunities. They do not expect it to be intrusive, nor feel as if it is running their lives. They enjoy control and personalisation of their digital experience. Some of their interaction with the digital landscape is direct, by means of a personal device, and some is ambient through devices, sensors and interfaces that are built into the environment. The key aspect of the digital citizen of the futures interactions is that they can control their experience and privacy. For example the digital citizen may expect to have the right to be digitally forgotten in other words for some or all of their online activity to leave no permanent record, and to be informed if there is a breach of their privacy.

    4.3.1 The Scenario

    The citizens of the future expect to be connected at all times to online services and information. They will expect their connectivity to be continuous and seamless as they move around. The idea of having to find somewhere with WI-FI connectivity or worrying about roaming charges is outdated and alien. Connection to the future Internet is cheap, reliable, ubiquitous and invisible.

    The digital citizens are sensitive to the price of services and are confident and able to compare and select services. They have become conscious of lock-in, and seek services and systems that they can switch between easily. They are security conscious but not paranoid. They are the successor to the first generation of mobile internet users, for whom novelty was a selling point. The citizens of 2020 and beyond look more directly for benefits and costs of the services and systems that they use. They have almost no tolerance of services that are unreliable.

    Services that enhance quality of life and leisure time will be in high-demand, with gaming likely to continue as a massive market that will drive many aspects of technology from components, software, displays and augmented reality.

    Their position in the value chain is both as consumer of services, and generators of data. The latter may be directly by sharing a wide range of assets such as photos, restaurant reviews, creative works and so on, and also indirectly by allowing profile data to be collected and used by third parties. Willingness to allow such profile information to be gathered will be variable among users; some will value privacy highly, while others are willing to exchange access to their profile data in exchange, say, for lower price services.

    4.3.2 Application example: The Smart Stadium

    The way in which the digital citizen interacts with the cyber landscape can be illustrated by envisaging the future digital citizen at a stadium attending a major sporting event.

    You arrive at the stadium, where you are directed by a smart system run by the stadium to the right entrance. The stadium can identify that you have arrived because you have allowed it to recognise your smartphone (or something else that you carry) when you are in range of its network. Your social network, which you have also allowed to know your location, looks up whether any of your friends are also at the event, and informs you. It suggests that you have time to meet at the bar beforehand.

  • NGC Roadmap Study 26

    You tell the social network that you prefer to wait until half-time, and the social network, knowing your preferences, reserves a space at the bar easiest to get to from where you are all sitting and lets your friends know the plan.

    The stadium confirms the validity of your virtual ticket and you are allowed into the stadium. You are credited with access to commentaries, replays, highlights and analysis as standard. You can upgrade to premium services if you wish (some stadiums even have fleets of mini drones carrying tiny HD video cameras that you can try out, but you think this is a bit of a gimmick. Instead you prefer to use the camera-net application that allows you to see the views from other peoples devices in return for allowing the same on your phone for other users). If you dont have an account with the company that runs these additional services, your software agent suggests (via your smartphone) a range of additional services that you might want as a pay as you go option. As a digital citizen you have become used to this business model of a personal software agent suggesting what you might be interested in, and paying for things electronically. The payments are authorised using biometrics, either using a device you carry or a mobile paypoint in the stadium. Behind the scenes your bank is running complex anti-fraud software to protect both you and itself against cyber-crime. This has become increasingly important as so many transactions are done electronically from customers on the move. Cash still persists, because its anonymity still has attractions, but digital payments are by far the dominant type.

    Your seat environment contains all you need to enjoy the event, including digital glasses and headset that allow you to access the match programme, camera views and commentary, depending on your viewing package. Of course you can easily use your own display glasses if you want to (around you a few people can even be seen with old-style tablets).

    Its easy to set up the commentary you want and the viewing angles you might want to have as shortcuts. The system allows you to simulate a few views so that you can tune them and quickly switch as the game progresses. This is important, as even though you are surrounded by all this technology and you expect everything to be personalised, its a real match with a real ball played at

  • NGC Roadmap Study 27

    real speed. The real-time experience is still the main part of the occasion and the technology has to be up to the job.

    At half-time you join your friends at the bar for that meeting suggested by your social network. Your social network realises that you have made the meeting together, and suggests that, since it knows its been a long time since you last time met face to face, you might want it to make a dinner reservation according to your group preferences. You are comfortable with this type of interaction with social networks, provided you have control over levels of privacy and the level of interaction. Software companies have invested heavily in designing agents that will act on your behalf there is big money to be made from this.

    After the match finishes, the stadium empties efficiently. The designers used complex simulations of the physical design and information systems, coupled with varying models of crowd behaviour to come up with a stadium that allow efficient and safe movement of people. Simulations are also carried out in real time to ensure that safety problems will not arise.

    Outside the stadium a pool of self-driving electric cars is available, some owned by the stadium and some by taxi firms. The queues are managed efficiently, and a car is chosen for you depending on how quickly you need to get to your destination and whether you are willing to share. Payments are all made automatically. The traffic management in the stadium vicinity has been switched to a mode to cope with the traffic volume at the stadium. The traffic management system knows about the destination of each vehicle and directs each one accordingly, optimising the flow. Priority routes are available to those who can afford it and the traffic management system deals with all the authentication and authorisation issues. All this is unobtrusive; each vehicle receives its route instructions directly and in real-time, whether it is in self-driving mode or being driven by a human (which still happens, but is becoming increasingly rare).

    As you approach your home, your home management system knows you are arriving and starts the utilities it can predict that you will need (bearing in mind it already knows about your possible dinner arrangement). You gain access to your house using a biometric based security system, and tell your personal software agent to confirm the dinner idea it suggested earlier. Your agent knows that you live within walking distance of the down-town restaurant it has reserved and knows that you like to walk, so it tells you that it will just arrange a taxi home. As you prepare to leave, the agent politely reminds you that rain is forecast later, so taking an umbrella might be a sensible idea.

    4.4 The Digital Nation: Its all about us

    The scenario describes the interactions between individual and society (other individuals) and the state (national and EU government, law enforcement etc). The states responsibilities clearly interface to other areas such as education, healthcare and transport where state and private sector begin to overlap.

    4.4.1 The Scenario

    Governments will be replacing as much paperwork as possible with digital services. Smart environments will be prevalent (smart city, road, etc). Some members of society see this as a benefit, offering opportunities, information and security. For others it is seen as intrusive, invasive and untrustworthy. Privacy issues will be a major issue.

    Some citizens may resort to ways of subverting what they see as surveillance. The state may choose to outlaw some of these methods which may create further tensions.

  • NGC Roadmap Study 28

    Governments will appreciate that infrastructure projects will probably have greatest value if they can interact. The providers will need to develop interoperable services which will require common standards to be used, and which will require data-integration technologies. The service providers will drive the technology needs to be able to provide the services in a cost effective way.

    States will of course out-source many of these services. Government contractors (especially in IT) do not have a good reputation. Uptake of the services may be tarnished by this populations may decide they dont trust them before they have even been rolled out.

    Defence is an area where governments will need NCG to support conventional defence systems and digital threats. Major defence companies are likely to continue to be recipients of considerable investment for research. NCG will be needed both in the deployments (for example missile guidance and detection) and in the design tools used (for example HPC simulation tools).

    The need to defend against cyber-attacks may change the defence policy of states significantly. Whereas the digital criminals will generally be aiming to accumulate wealth through illegal means, cyber-attacks will be aimed at causing damage and disruption to national infrastructure. Sources of attack will include full scale attack by another state (unlikely), attack by terrorist/activist groups (highly likely) and the low-level hacker (the digital equivalent of graffiti, and an ever-present threat). Resilience against attacks will need to be an inherent property of all services and infrastructures, whether state maintained or not. Monitoring networks and detecting threats will be a major activity of the defence and security services

    There may be a demand for personal encryption systems. Management of encryption keys is the main issue here: webs of trust may evolve. Centralised Public Key Infrastructure (PKI) may not be regarded as secure by some privacy activists since the roots of trust might be seen as capable of being influenced, for example by government or big business. The proportion of people who understand the issues well is likely to be small, but how people will react will affect how governments can use NGC. Anonymity will still be attractive (cash for example will still be in use).

    NGC will be important in education in two ways; to enable advanced teaching methods and as a subject in itself. We have relied to date on life experience of using ICT; this should be taught properly to the next generation of students. Schools will be high-tech places with NCG used for many aspects of learning. Informed decision making, particularly about the realities of security, will help define the future use of digital marketplace and services.

    States will want to find a way to raise money from the new technologies and the economic activities they stimulate, but will want to do so without stifling economic growth or alienating consumers.

    Tax systems will come under scrutiny, as they have not caught up with modern ways of business. Companies have been criticised by using loopholes to avoid paying tax, and in the digital era this appears to be too easy for them. They will continue to do this, obviously.

    Governments will also want to use NCG to have a more productive workforce by enabling those who are currently not economically active to become so (for example more part time work for disabled, pensioners, part time workers, those in remote areas etc.). They are likely to do this by providing incentives for individuals and service providers (for example for getting good internet access into remote areas, providing training etc).

    One of the assets managed by the state is the electromagnetic spectrum. Licensing of this can raise money for states; conditions on the licensing deals can ensure that the licensees (for example

  • NGC Roadmap Study 29

    telecommunication companies) make the best use of it for national benefit, and do not squat. There need to be incentives (or laws) to ensure that operators use or lose their bandwidth.

    4.4.2 Application example: Emergency response

    A fire starts in a major public place, in this example a shopping complex. Within the complex, an intelligent network of sensors detects the outbreak of fire and track its progress. The network is installed and operated by a private company, but is subject to health and safety regulation part of which requires it to interact with public emergency services.

    When the public emergency response services are alerted, they immediately create an incident management centre to control the operation. The computing and communications infrastructures must allow for a control centre to be set up almost instantaneously to access the data and networking. Emergency teams would need to be able to predict the progress of the fire, which would need data on the buildings to be instantly accessible from a database, simulation tools and the necessary computing power, and the ability to rapidly evaluate different courses of action.

    The emergency response team is able to evaluate different strategies to carry out an effective evacuation and to tackle the fire. This will be made possible by the use of advanced simulation tools together with HPC available on demand.

    A key issue in managing the emergency will be the ability to detect and track people. Technologies to do so for their own security would be useful however there are potentially massive privacy issues. Locating people through their smart phone or other device might be possible, but mandating that people always carry such a device for safety reasons would likely be impractical and would attract fierce opposition from many citizens.

    Sensor technology that could detect whether people were trapped inside the building without needing any identity information is more likely to be used in the relief effort.

  • NGC Roadmap Study 30

    Control and management systems for other infrastructure (for example public transport) would need to interface with the emergency teams so that contingent disruption could be managed. Some of these infrastructures will be privately operated, requiring interaction between public and private systems.

    4.4.3 Application example: Cyber-crime prevention

    Digital crime will increase in the future. Fraudulent on-line transactions may replace burglary as the crime the public fears. Public will demand that devices such as phones, laptops etc (which we will most likely still be using) are not useful to criminals if stolen and do not pose risk to the owners if lost. In other words, theft of a phone or other personal device should not mean the thief can impersonate you on-line.

    Theft of data will be as serious a crime as a theft of tangible assets or money. This will apply to individuals as well as companies and government. It will be unrealistic to assume that intrusion will not happen, and perimeter models of security will be outdated. Intrusion detection will need to be highly sophisticated.

    Major international cybercrime organisations will exist in the future. There will need to be a big effort on international (and pan European) efforts to detect and prevent crime. State will drive the need for technologies that will support this. Cyber-crime prevention will require advanced methods of analysing behaviour to detect potential criminal activity. This will need major advances in modelling and simulation software, high performance computing and data analysis tools to detect intrusion, together with encryption technologies easily usable by the public to limit unauthorised use of data.

    Some technologies that will be needed in preventing crime may also be of use to the criminals themselves (face and image recognition, software to analyse behaviour etc). There may be a need to licence some these technologies. Enforcing this will be a challenge.

    The public needs to be educated about digital security and good practice needs to be understood. This education process will need to start at an early age schools will include this in the curriculum with the same level of importance that road safety had in the 1970s.

    4.5 Intelligent Transport: Trains and other Vehicles with Brains

    Future transport systems will need to be used much more efficiently in a response to increasing energy costs and the need to use existing infrastructures more efficiently, together with the demand for more travel as populations become ever more mobile. This scenario deals with the situation as it may be after 2020, under the influence of these megatrends.

    4.5.1 The Scenario

    Until around 2020 the railroad system was used very inefficiently. Although the railroads had a huge coverage over Europe, the concepts and technologies were not forward looking enough until pushed forward in 2015 by initiatives from the EC.

    Up to 2015 the basic concept of using railways had not changed since its invention. In comparison to roads, railways were used very differently and sparsely, for example regarding train length in relation to unused rail length in front and behind a train.

  • NGC Roadmap Study 31

    With the help of massive research programmes CPS and HPC approaches were developed that allowed a denser and on-demand usage of railroads. Today the autonomous trains featuring powerful sensors and fulfilling high security standards are being able to do their own planning, are able to communicate with each other and with several control instances to realise a seamless co-operation with autonomous cars and other vehicles.

    4.5.2 Application examples: Business trip of the future

    Stephanie needs to make a spontaneous business trip to a partner company which is 450 km away. Five minutes after deciding to meet her business partner in person she talks to her personal assistant avatar which she can see through her digital glasses. Her digital assistant starts to send requests to an international broker service, and starts to plan possible travel options. Since Stephanie needs to take a demonstrator weighting approx. 23 kg with her, the assistant neglects public transport options. Due to the urgency and time constraints, the assistant checks all other options for leaving the high-density traffic zone of the city even though it is aware that Stephanie usually prefers to avoid shared car options.

    Having worked out some options for the main part of the journey, Stephanies assistant avatar communicates with a local transportation broker, which in turn broadcasts the transportation request to pick-up vehicles in the area around the company. These autonomously driving cars are operated and maintained by private companies. The vehicles calculate the costs for a detour to pick-up a person at Stephanies company. Each estimate is done in real time, and has limited validity. The cars are programmed to gain the maximum profit without delaying the ride for the passengers that are already on board. It is crucial that they do an excellent job in calculating the risk and the estimated energy consumption of taking a detour versus bringing their passengers already on board to their target locations in time.

  • NGC Roadmap Study 32

    Using the low latency wireless megacities network (LLWMC) twelve vehicles reply to the broker including their bid to take the passenger and the desired target zones they are heading for. The broker forwards the offers to Stephanies assistant avatar, and they become the base for further calculations. Having a number of options for the initial part of the journey, the assistant starts planning the next sections of the trip. Three target zones are in the medium density traffic zones providing good starting points to far distance transportation vehicles.

    Since the final target - the company that Stephanie is going to visit - is in the countryside in a remote location, the availability of local vehicles is very low during business hours. Therefore the plan is to find a vehicle that is available for the rest of the day. Eight available vehicles are offered from three different providers via the vehicle auctions system CARbay. Four of them are close to the end of the auction, which allows a decision to be received in a short time frame. The digital assistant enters the auction based on its implemented and up-to-date company policy rules and is successful in grabbing a car for the rest of the day.

    The final part of planning the trip is to optimise the remaining 400km. To do so, the digital assistant directly connects to the control and navigation system of the car using the secure passphrase that is valid through the time of the rent. It deploys an optimisation task by handing over the target, the estimated payload, the preferences of Stephanie and the policies of the company. The car takes over the task of finding a route based on its charging level, cruising range, traffic announcements, charging stations including todays energy rates, the estimated payload and many other variables.

    The car starts to broadcast requests using the local zone broker service to find other cars with similar target zones for teaming up to use a shared train ride.

    Starting from a group of 8 cars there is a chance to order one of the autonomous car trains. After a few moments enough cars acknowledged which triggers the broadcast to the wide zone broker service (w-ETTBS). Similar to the mentioned shared car services the trains are operated by private companies and able to plan their trips based on energy prices, maintenance cycles, available routes, traffic status, etc. and are replying with their offers. If there are similar replies the ETTBS starts with a mini auction where the cars are included that are acting in that case as representatives of the interests of their passengers, for example I want to be there in time despite the costs vs it should be economical, I am in no hurry.

    Due to high bandwidth communications and powerful on-board multi core brains of the autonomous vehicles as well as the HPC centres running the ETTBS the whole planning process took only 55 seconds.

    Stephanie takes her coat and enters the pick-up vehicle followed by the autonomous transportation trolley carrying the 23kg demonstrator she wants to show her business partner. She leans back and starts to review her prepared presentation on her foldable and lightweight screen foil.

    4.6 Education and Research: Connected brains

    Research and development are undergoing significant changes and the need to innovate our education systems is almost universally acknowledged. An important megatrend driving development in research and development is the shift towards a post-scientific society in which the focus is on the development of novel services and products and thus on innovation. Individualization is a major trend in all modern market economies. With people looking for individualised solutions to their demands, the role of large markets in countries such as Brazil and China will continue to remain important.

  • NGC Roadmap Study 33

    This is a direct consequence of a globalised economy, but also of the global power shift to new and emerging markets in Asia and other regions of the world.

    Changing demographics is also driving the need to improve the ways in which we create innovation and the ways to educate the young but also ensure life-long learning for everybody. It will be important to ensure proper education opportunities not only in increasingly urbanised areas, but also in the more sparsely populated countryside where internet connectivity may remain significantly lower.

    Future researchers and scientists will rely on a rich set of tools for scientific discovery, collaborative working, research and innovation management. This scenario presents a vision of the future in which researchers collaborate on a massive international scale, driven by the trends described above and enabled by next generation computing.

    4.6.1 The Scenario

    Researchers in 2020 and beyond can quickly access a large variety of data repositories. Intuitive user interfaces support dynamic interactions with the data in an intuitive fashion to quickly find the answer to What if? questions or to discover new hypotheses. The scientist of the future can rely on a set of powerful computing resources that are dynamically reconfigured in response to changing load requirements. Tools are capable of finding a balance between prime services as a price premium, economic services with reduced data size or resolution and delayed services (overnight etc.)

    Data acquisition in laboratories happens to a large extent automatically with the help of robots and intelligent systems for data collection, validation, formatting and evaluation. Systems are able to detect anomalies and novelty in the data.

    Researchers easily run simulations before and after real-world experiments to speed up the time from the concept phase to research prototypes and implementation. New production tools such as 3D printing support the prototyping phase, but also the exchange of new ideas between researchers world-wide.

    The IT infrastructure supports the seamless connection of experts to support collaboration on complex large-scale research projects, but also to harvest microknowledge. The future experts personal IT system facilitates the experts contribution to a