Nanotechnology Foresight: How can we exploreEmployment and Skills Implications?

Ian Miles, Manchester Institute of Innovation Research(Ian.Miles@mbs.ac.uk)

Foresight, Roadmaps and Indicators for Nanoindustry InternationalWorkshop at NANOFORUM ’09, Moscow, on Oct. 7-8 2009

ABSTRACT

Policymakers responsible for creating the skill baseneeded for future economic development need to havebetter understanding of the implications of radicaltechnological change. In the long-term, such change canupset many of the forecasts based upon standardapproaches to modelling employment and occupationaltrends. This study reviews some approaches to dealingwith major new technologies, and then explores theusefulness of using familiar ideas from innovationresearch – diffusion and industry life-cycle analysis,thinking about technological paradigms and trajectories,for this purpose. Nanotechnology, in part because of itsrelative infancy, in part because of the intersection ofnumerous lines of knowledge development at the nano-scale, and in part because of the many contested claimsabout the scope and speed of its evolution, posesparticular challenges. The paper assesses how confidentwe can be about skill and employment projections relatedto nanoindustries - and industrial applications ofnanotechnology more generally – and proposes some ways inwhich we could increase our confidence and provide more

policy-relevant intelligence.

Nanotechnology Foresight: How can we exploreEmployment and Skills Implications?

Ian Miles, Manchester Institute of Innovation Research

1 Introduction: Foresight and ForecastingForecasting the development of major new technologies isnotoriously difficult. We may be able to delineate majortrajectories, but timescales are often mis-estimated, andwhich of several competing platforms and standards islikely to succeed is often determined by nontechnologicalfactors. Outlining the implications of technologicaldevelopment for skills and training adds a further levelof complexity.

One thing that we can be sure of, is that “business asusual” is likely to be disrupted. Institutions andpractices will need to change, the nature and content ofoccupations and skills will evolve. ContemporaryForesight practice (cf Georghiou et al, 2008) recognisesthat it is usually inappropriate to focus on a “mostprobable” future. When we confront disruptive change,conventional ways of estimating likelihood break down.It is important to explore alternative futures, to gain abetter grasp on the range of contingencies which mightplausibly emerge.

Contemporary Foresight also puts a great deal of stressupon a participative or interactive element. Thisreflects awareness that future-relevant knowledge andcapacity to act are distributed phenomena. They are notentirely decentralised, let alone evenly distributedacross societies, of course: each is concentrated inspecific locations. Not only is it that knowledge andcapacity to act may be localised in different sites.Different sorts of knowledge and capability are often,also, be concentrated in different ways. Knowledge isfragmented across numerous disciplines, professions,

practitioners, and economic agents for example.(Consider knowledge of fundamental sciences, of variousfields of technology application, or ethics andregulatory frames, of markets and marketing, and so on.)Capacity to act varies widely, of course. It ismanifested in various spatial levels and sectoralagencies of governments, business enterprise, and civilsociety networks. (Consider access to finance, to civilsociety, public opinion and media, to regulators, forexample.) If Foresight is to be well-grounded and to beeffectively acted upon, these conditions have to berecognised: which means establishing frameworks thatengage broader communities. Classical deskworkforecasting and modelling have their place, but they areonly some of the inputs to a wider Foresight process thatwill need to access knowledge that is widely distributed.

Few organisations possess sufficient knowledge about thewhole range of significant factors whose interplay shapesthe main contours of future possibilities. Theconventional response – convening an expert panel – canbe invaluable, as long as the panel is constructed so asto capture a sufficient range of knowledge: this oftenmeans going beyond the standard pool of expertise drawnon by an organisation. This conventional response canfail to be sufficiently persuasive to a wide range ofstakeholders, who may dismiss recommendations as flowingfrom the “usual suspects” with their establishedframeworks and sectional interests, too. The implicationis twofold: panels need to be constituted in appropriateways, both in terms of perceived legitimacy and practicalcapability; and techniques to engage wider pools ofparticipants in providing information, creating visions,and exploring implications, should also be implemented.Among the latter techniques are Delphi and relatedsurveys (polling large numbers of people), consultationprocesses and workshops, scenario and other “visioning”workshops, and the like. These are primarily engagementrather than being mainly dissemination activities.

As we have pointed out, capacity to act is alsodistributed across societies. Engagement is valuable

for reasons that go beyond enhancing the input ofrelevant information to the Foresight process, and beyondproviding the legitimacy that comes from a wider (if notactually democratic) sampling of perspectives. It meansthat the participants in the process will have a fullerunderstanding of the Foresight exercise, and of theissues that it is addressing. This means that theyshould have increased awareness of important facets ofthese issues, and thus have greater preparedness to dealwith contingencies as they arise. They should also havea better understanding of the perspectives and likelyresponses of other participants, and thus be more able toexplore possible collaboration and cooperation – orsometimes to achieve a better competitive positioncompared to others. They will, furthermore, be able toact as “carriers” of the messages from the Foresightexercise – they will know more about it, and more aboutthe logic and evidence on which the results andprescriptions are based. Compare the extent of learningachieved from taking part in group discussions anddialogues, with having a report explained to you, tosimply reading it on your own.

This general philosophy of Foresight was articulated,especially during the 1990s, in the course of developingthe approach to grapple with questions of Science,Technology and Innovation (STI) policy. Many earlyForesight exercises focused on determining priorities forR&D and related funding, for example (though issues ofskill and training often emerged in the course of suchexercises). Much subsequent Foresight has had a widerbrief, though it is no surprise that in an epoch of rapidand disruptive technological change, STI questions haveoften loomed large. Indeed, this change has been a majorcontributor to the widening, in recent years, of muchstrategic analysis based largely on conventionalforecasting activity to include more Foresight elements.Analysis of skills, training, and employment is noexception here: we see similar grappling with Foresightperspectives in fields as disparate as environmentalstudies, health planning, criminal justice…. Large-scale, fully-fledged Foresight exercises are thus

imperative if skills and training agencies are toconfront the challenges of technological change.

The present essay, however, is not about the design ofsuch a Foresight process. Rather, it is a more modestexploration of one type of input into such a process. Itis itself deskwork, exploring the potential andlimitations of further deskwork for forecasting theskills implications of nanotechnology (SKIN), andconsidering where wider Foresight activities will beessential. This will explore some approaches to thinkingthrough SKIN, in order to inform subsequent forecastingand the design as well as the content of any Foresightexercise. It will draw extensively on innovation studiesand what these have to say about major technologicalchanges. Even so, it will be somewhat schematic, becausethis is a large and expanding body of literature. Thepresent essay should be seen as merely a starting pointfor more detailed analysis.

2 Innovation Processes and Trajectories2.1 Innovation Research

Before turning to skills and employment, it is worthintroducing a number of observations about major shiftsin technology. Studies of the introduction and diffusionof new technologies demonstrate that these often takeconsiderable time to diffuse (e.g. David, von Tunzelmann,1978). It is also clear that successful innovation willtypically require “product champions”, influential peoplewho will continue to support and promote the innovationeven when times are tough.

There are cases of rapid take-off. As diffusion studiesindicate, this is most likely to happen:

when the new products are highly visible and theirfeatures can readily be appraised;when the benefits of these features are large andclearly demonstrable;when their purchase costs are relatively low;

when the efforts in terms of learning to use themeffectively are perceived to be low;and when the organisational adjustment (change inwork, jobs, routines) is seen as being relativelyunproblematic.

Often such conditions will only be achieved after a greatdeal of effort on the part of suppliers. There may thenbe a dramatic take-off after a long period of disinterest(other than on the part of a few pioneers). Examples arethe extremely fast uptake of microelectronics (NC Machinetools, robots, etc. in manufacturing; word processors andPCs in office work) in the 1980s, and of the Web andmobile phones in the 1990s. Interestingly, while someof these new applications of IT had been widely heralded,the success of cellular telephony - and text messaging(SMS) services within this – came as a great surprise tomany commentators.

The Web is a good case of a “design paradigm” thatestablished a market – especially once the web browserand search engine services were introduced. Mostobservers had anticipated a take-off of Internet use oncePCs were widely available. But until the Web burst ontothe scene (following the innovation of browsers andsearch engines) there was little sense that commonformats and standards were required. In consequence,much effort was wasted on developing unique and obsoleteinterfaces when people were seeking to develop new onlineservices. The Internet had been in use for decades, butit took improved PCs and networks, and visibly useful anduser-friendly services, to promote widespread uptake.This rapid adoption of the Internet and Web serviceshelped underpin the “dot com bubble” of the end of the1990s. And what does it take for a design paradigm toestablish a market? It needs to be compatible withsocial practices and available skills, or at least notchallenge these too dramatically. (Thus fax wasinitially preferred to email in part because it could betreated as an extension to secretarial letter-writing:email took off once professionals were themselves used tousing PCs to prepare documents, graphics, and dataanalyses, and could see the advantages of electronic

transfer of information.) It needs to offer substantialadvantages without too much inconvenience or need forrecourse to expert support. A critical mass of usersmay be helpful to establish economies of scale (e.g.costs of equipment and software), network externalities(other users to communicate with, large data resourcesonline), and the conditions for the emergence ofcompeting suppliers (driving down costs) andcomplementary products (increasing functionality): asuccessful design paradigm establishes such a criticalmass, while adopters may be reluctant to invest in non-standard products.

Discussion of diffusion of innovations is inevitablyassociated with the familiar subject of logistic S-curvesas ways of thinking about innovation and diffusionprocesses. Though familiar, the subject is worthrevisiting, with the diffusion, product cycle, andindustry cycle versions all being relevant to the presentessay. Figure 1 presents the classic S-curve.

In diffusion analysis, the uptake of an innovation isoften modelled in terms of S-curves, which highlightseveral issues. The diffusion of many innovationsfollows the basic pattern, wherein a period of slowuptake is succeeded by one of much more rapid growth,which in turn is succeeded by slower growth as mostpotential adopters have acquired the new product. Thegrowth of sales may not mirror the growth in theproportion of the population using the product, becausesome products, at least, can be acquired repeatedly –automobiles and TV sets, for example, are not onlyreplaced, but many households own more than one.

Figure 1 The S-curve

S-curves in diffusion analysis draw attention to thecharacteristics of early and late adopters. At theindividual level, early adopters are typically morehighly educated, more affluent, more linked in to massmedia, etc.; at the firm level they are typically larger,in more metropolitan regions, linked to internationalfirms (The classic reference here is Rogers, 1995) Ifwe consider the diffusion of major new technologies whichcan be applied to numerous ends, we can also considerthere to be a pattern of diffusion across sectors, withsome being early adopters, some later. For example, atsector level early adopters tend to be high-tech sectors,and probably those with closer links to the sectors inwhich the innovation was first developed and/or applied.

Early Adopters/Early Designs

Mass Markets/ Mass Production / New Applications

Take Off/ Design Stabilisation

Market Saturation/ Product Differentiation

Time

Diffusion Level/Product Maturity

Such S-curves also point to critical uncertainties inforecasting diffusion. For instance, if a take-off isexpected, how rapid will it be? how long will it bebefore take-off can be determined to have occurred? Andagain, where is there likely to be a ceiling? What isthe potential population of adopters, and is this thesame thing as the potential scale of incidences ofadoption (what is the relevant population?1 Even if weare restricted to one application, might there be more ofthe products in question diffused than there areindividuals?2 Should we treat different applicationsinvolving the same basic technology as discrete cases ofinnovation?)?

In terms of using these approaches for forecastingdevelopments, it is important to note some deviationsfrom the general picture. While S-curves often do proverather good fits to empirical data, in practice they areoften interrupted by such events as wars and economicdownturns. A competing technology may substitute for agiven innovation before it has been adopted by all of itspotential population – some firms have leapfrogged faxand gone directly to email, for example.

But S-curves are also used in another – closely relatedbut distinctive – context: the industry life-cycleapproach and that of the product life-cycle. Theindustry life-cycle approach was initially developed inthe context of discussions of international trade and thechanging international division of labour. It was a toolfor understanding the ways in which production in certainindustries, and of certain products, was shifting acrossthe world. (Vernon, 1966, etc) The basic idea here isthat industries are liable to “mature” over time. The

1 We may find that we have mis-specified the population – amongconsumers, the estimates may have missed out children, in theemployment data they may miss foreign workers, among firms, theinformal sector, and so on.2 The number of mobile phones in use in some countries exceeds thepopulation of these countries; and it is not surprising that thereare more phones used in business than there are firms, nor that thereare many more instances of use of common tools than there are offirms in the user sectors.

products they create become more standardised, theirprocesses become routine, and the production technologiesthat they require becoming cheaper and/or simplified.They are less dependent on a highly-skilled workforce.

Less mature industries need to be based in technicallyadvanced regions, where they can access high skills,technical support, and innovative equipment and componentsuppliers. As they mature, however, they can more easilybe located in less-advanced countries: transnationalcorporations may establish offshore productionfacilities, indigenous firms in developing countries mayemerge as low-cost competitors. This approach was usedto account for the shift of some established industriesout of the more advanced countries – and to recommendthat these countries should increasingly specialise innewer high-technology industries. The approach has itscritics – how far does it apply to individual products,product classes, firms, industry sectors; how to accountfor the persistence of many established industries insome advanced countries; how to handle the phenomenon of“dematuration” and shifts away from Fordist productionsystems? - but again it points to some underlyingdynamics of successful product innovation.

Empirical support for the approach is less conclusivethan for S-curves in diffusion analysis. In part this isbecause many industries display continuing change andinnovation, such that important shares of activity can beretained in the original high-tech countries.“Dematuration” may take place, with technological andorganisational innovation undermining the comparativeadvantages of low wage economies. For instance, newprocess technology may make labour costs less relevant,and new skills may be required in the labour force.Additionally, closeness to markets, and flexibility ofresponse to more volatile markets, may be important; somesectors may not migrate because of political or corporatestrategy. The developing countries themselves may becomepioneers in the sector, rather than imitators ofestablished technology – and there has been little signthat sectors are moving around the world to increasingly

less developed countries. Finally, it is striking howmany industrial activities have been offshored, with manyforms contracting out specific assembly and serviceactivities overseas. Some analysts have argued thatintensified international division of labour would meanthat latecomers would be restricted to less knowledge-intensive components of value chains, and be unable toacquire capabilities to play much of an autonomous rolein industries because they are only experienced inworking on small fragments of extended productionprocesses.

Still, the basic ideas of the industry cycle approach dohave continuing relevance. There are, and are likely tocontinue to be, shifts in the sites of production shiftsand changes in the international division of labour. Thedynamics of new technologies requires examining more thanone country in isolation in terms of the location ofproduction processes as well as the pace of adoption indifferent regions. Latecomers can take important rolesin industrial production – and in some cases importantnew products can emerge from them. Pioneers cannotassume that their own products will continue to dominatein other regions of the world.

Moving from the industry life-cycle to the product lifecycle approach, further points arise that areparticularly relevant to considering the dynamics of newtechnologies. This approach draws on both the diffusionand the industry cycle models, together with theaccumulated results of many innovation studies. (seeUtterback, 1996 for the Abernathy-Utterback model) Thefocus here is not so much on how the market or sectorevolves, but on the nature of the innovation itself.

Put very schematically, the Product Life Cycle approachsuggests that early versions of innovations aretypically, even is technologically sophisticated, verybasic and rudimentary as compared to their laterincarnations. The early versions of the innovationappeal to relatively few users, not just because themarkets are underdeveloped. Additional factors include

the lack of widespread knowledge of their current andpotential capabilities on the user side, and the lack ofknowledge on the supplier side of just what capabilitieswill be valued and how they may be used. This underpinsthe stress placed by many innovation researchers andpractitioners on good communication links between usersand producers, which informed the development of the“systems of innovation” approach. It also leads toconsidering that in the early stage of productdevelopment and diffusion there may be poor linkagesbetween the innovators and many potential users – andtypes of user (i.e. applications of some sorts may bequite well established, while others are much less matureor even yet to be created). More generally, theorganisation of upstream supply chains and downstreamsystems for distribution and training, are liable to berudimentary. A successful product life cycle will seethese features changing as product development anddiffusion coevolve. At the early stage of development,complementary goods and services are not available, ornot widely available, The key products themselves areliable to be expensive (and often unwieldy and/orunreliable), as compared to later versions of similarproducts. They early versions of products are liable torequire considerable technical skills for theirproduction and use, while these may become much moreroutine in later stages. Skills that were rare to beginwith, too, may have become much moiré commonplace astraining agencies and employees themselves see the needfor them.

However, if the products do gain a foothold, thenawareness of them increases, investment in them grows,experience as to how to apply them accumulates, and theirmarkets expand. This is in line with the diffusionaccount, but crucially we here see that the innovationitself is changing. The product is redesigned so that itbecomes more robust, easier to use, capable of moreefficient production. The early suppliers who havedemonstrated the potential for a substantial new marketare joined by new entrants, who may offer new designs forthe innovation. Often there is a competition between

different types of design, with one becoming the dominant“design paradigm” to which others should conform. If theearly suppliers are innovative small firms, large firmsare liable to substantially change the nature of thecompetition in the market (perhaps acquiring the smallfirms in the process); with superior marketingcapacities, the large firms may accelerate the diffusionprocess and hasten the development of a dominant design.Then, as the market grows, the focus of innovationtypically moving away from basic product/designinnovation, toward improving product quality, and then toprocess innovation enabling larger-scale, cheaperproduction. (This resembles the industry cycleanalysis.) The product becomes cheaper to use, with morecomplementary goods and services, more functionality,less requirement for high skills in its use. It can beacquired more readily, with replacement, repair andperhaps second-hand markets. One version of this accountdescribes the product cycle as involving a shift fromtechnology-push to demand-pull, though this is somewhatoversimplified. But what the product cycle approach doesdraw attention to is a pair of learning processes. Onthe one hand, successful innovations themselves evolvethrough an extended development process once they are onthe market (not only when they are precommercialprototypes in R&D laboratories), as suppliers learn whatusers require. On the other hand, users learn about theproduct and about how to use it effectively. The latterpart of the user learning process may be very protracted– for example, it is believed that the “productivityparadox” of new IT was only resolved once users hadlearned how to apply their large IT investmentseffectively. (Joergenson et al, 2008) There is arelated message here: since users play an important rolein shaping innovations, it is unwise to assume thatinventors and developers of innovative products have aclear view of how these products will be used.

This leads us beyond the literature on S-curves, towardsthose on organisational learning, on technologicalconfiguration, and on “reinvention”. Adopters andcreators of complementary innovations may effectively

modify the innovation – for example, by finding ways ofusing it, and areas of application, which were notanticipated by the inventor or even the by a supplieryears after the first introduction of the new product.While by the early 1980s PC manufacturers were alreadybecoming aware that consumer markets were developing forcomputers on which to play videogames, for example (whichwould have surprised computer pioneers!), few would haveanticipated the emergence of large scale music and videodownloads, of Web 2.0 social networking sites, and so on.

What is also implicit in these approaches is that thesocial and technological networks of which the economy iscomprised themselves can evolve during the innovationprocess, with changing linkages between producers andusers of innovations, with changing intermediaries(including regulatory and training organisations) and soon. The social networks that are required for successfuldiffusion of an innovation may themselves take time todevelop, and they may be stimulated or impeded by policyaction.

2.2 Fundamental Innovations: Beyond S-curves

It has been necessary to introduce the various themespresented above in order to provide a reasonably firmground in innovation analysis for the discussion ofnanotechnology. It would be too easy otherwise to becarried away by the hyperbole that has accompanied therecognition of the potentials of doing things at nano-scales.

The common ground between the various perspectivesexamined is the anticipation of continuing change andgrowing capabilities. Whether what we will witness willbe a technological revolution, or a loosely connectedseries of radical changes in technology system3 across a3 Freeman (1974 – see also Freeman and Louca, 2002; Freeman andPerez,1988) noted that the significance of innovations in industrial

number of industries (manufacturing, health, etc.), thereare still liable to be substantial implications foremployment, skills, industrial structures, and the like.Experience with earlier major changes in technology canbe a useful guide to thinking about these implications ofNT. In developing such “historical analogies” (ref Bell,1964) we should always be mindful that both specificitiesof technologies, and contextual changes such as

processes varies considerably. Some new process technologies arerelatively minor modifications of familiar devices used in specificapplications . These usually require little retraining andreorganisation of activities, and they may only reduce product costsor increase quality by a fractional extent. At the other extreme,some new process technologies represent dramatic changes: they canbe used to modify the methods of production in many sectors, and areliable to promote considerable rethinking of work organisation andeven of the nature of the end-products of user organisations.Freeman thus proposed a typology of innovations, which are, in termsof increasing significance: Incremental Innovations. These occur more or less continuously,

including in later stages of the product life- cycle; they usuallyinvolve small modifications with only minor implications fortraining, working practices, organisation; they often derive frominventions and improvements originated by engineers, workers, oruser suggestions.

Radical Innovations. These involve more substantial change, whichmay lead to new training, new organisation of work andorganisation practices - one "rule of thumb" for distinguishingradical from incremental innovations is whether new manuals havebeen written. Radical innovations frequently derive from formalR&D activities, usually on the part of the supplier; they may wellinvolve product innovation as well as changes in productionprocesses and organisational arrangements, and thus they maycreate new markets or extend old ones - setting off new productcycles, or "dematuring" established products.

New Technology Systems. Here, beyond a specific branch or sector,basic inventions (perhaps the result of R&D in supplier industriesthat have heretofore had little contact with the users) areforming the basis for reorganising a whole set of productionprocesses. Both radical and incremental innovations based on thenew technology emerge, in a range of related sectors. Newtechnology systems are liable to develop over a relatively longperiod, since they require substantial organisational innovationand the generation of new markets.

Technological revolutions. In this case fundamental basicinventions constitute new "heartland technologies" that can beapplied to transform production - and generate new products -across a very wide spectrum of activities, across many sectors of

globalisation and economic competition, economic up- anddownturns, and even wars and other crises (environmental,health, security, for example), may reduce the scope forgeneralising from past experience. Such factors may slowdown the pace of change, or speed up particulartechnological developments. They are liable to createmarkets (e.g. public procurement, more or less luxuryproducts), and thus to focus innovative efforts

3 Nanotechnology as Revolutionary Technology

3.1 Which Nanotechnologies?Many commentators hail nanotechnology (NT) as the nexttechnological revolution. In some accounts this will bethe successor to the well-established InformationTechnology (IT) revolution and the still-acceleratingbiotechnology (BT) revolution. Many commentators aresceptical of many particular claims as to future NTpotentials – they argue that nano-robots and atom-by-atomassembly are at best long-distant prospects. But evensceptics are liable to be impressed by the wide range ofemerging developments and applications that go under theNT label. One common line of criticism is that NTdevelopments are too multifarious to be classifiedtogether as a coherent body of knowledge development –indeed, that many activities have been reclassified as NTfor reasons of attracting funding, and these may move toother terminology once fashions change. NT involvescontrol and restructuring of matter at the nanoscale –usually taken to mean less than 100 nanometres.. (Simplybeing able to examine or measure nano-scale phenomena isnot itself NT). But this says little about what iscontrolled, how, and in what ways.

the economy. In effect many new technology systems are developingsimultaneously. A long period of diffusion and development islikely as many social, institutional and organisational changesare initiated in different areas.

The case for NT as revolutionary is an intriguing one.The argument is that there are major increases in ourtransformative capacity that are imaginable – and in somecases already realisable – by working at nano-scales. Formillennia human beings have been producing and applying“new” materials, and many materials that are only familiarin recent decades - for example plastics – actually have arelatively long history. There are extremely intriguingand completely novel (or unknown) substances that haverecently been developed – for example, the fullerenes,carbon nanotubes, and metcars. NT means new approachesto Advanced Materials (AM) - not just a particular classof new material, nor even application of a particulartechnique or set of instruments to producing materials.NT knowledge, as applied in economic activities atpresent, is largely a matter of applying a range of newprocesses to be applied to the production, shaping, andconfiguration of materials, so that effectively theproperties of AM or of familiar materials processed in newways (e.g. “chips” with circuitry engraved upon them) canbe defined at very great detail, at nano- and indeedatomic or molecular levels.4

Mike Roco (2007) of the US National NanotechnologyInitiative, argues that there will be four generations ofNT development, as depicted in Figure 2:

First generation, turn of the century – “passivenanostructures” and is typically used to tailormacroscale properties and functions.. These arematerials designed to perform one task, whosespecific behaviour is stable in time. Examplesinclude: nanostructured coatings, dispersion ofnanoparticles, and bulk materials - nanostructuredmetals, polymers, and ceramics.A second phase, beginning in the early years of thiscentury - “active nanostructures” for producingmechanical, electronic, magnetic, photonic,biological, and other effects. These are typicallyintegrated into microscale devices and systems.Examples include new transistors, components of

4A helpful discussion of this point in the context of AM is providedby Cohendetet al (1991 ).

nanoelectronics beyond CMOS, amplifiers, targeteddrugs and chemicals, actuators, artificial “muscles”,and adaptive structures A third generation, beginning in the second decade ofthis century - “systems of nanosystems with three-dimensional nanosystems”, constructed with thousandsof interacting components, by means of varioussyntheses and assembling techniques such as bio-assembling; robotics with emerging behaviour, andevolutionary approaches. A key challenge associatedwith construction of such systems is enablingnetworking to take place at the nanoscale, anddesigning hierarchical architectures.. Roco mentionssome research areas that will underpin this:heterogeneous nanostructures and supramolecularsystem engineering. This includes directed multiscaleself assembling, artificial tissues and sensorialsystems, quantum interactions within nanoscalesystems, processing of information using photons orelectron spin, assemblies of nanoscaleelectromechanical systems (NEMS) and convergingtechnologies (nano-bio-info-cogno) platformsintegrated from the nanoscale.Fourth generation - “heterogeneous molecularnanosystems”. These are seen as integratednanosystems, with hierarchical systems within systems(functioning much like the cells of complexorganisms). each molecule in the nanosystem has aspecific structure and plays a different role.Molecules will be used as devices: their engineeredstructures and architectures will permitfundamentally new functions. This will lead tonanoscale machines, nanosystem biology forhealthcare, and new human-machine interfaces at thetissue and nervous system levels, for example.

Figure 2 displays Roco’s estimates of the developments ofNT markets.

Figure 2 Roco’s forecast of generations of Nanotechnology

Source: Roco (2007) andhttp://crnano.typepad.com/crnblog/2006/03/new_risks_new_f.html

accessed 27/03/09

Figure 3 Forecast of Market size of Nanotechnology

Source: Roco (2007)

There does not seem to be one “heartland” NT, in the way thatmicroelectronics has been for information technology, or genesequencing/manipulation for biotechnology. It is possiblethat one may emerge, that one of the existing techniques willprove to be the basis for numerous rapid developments. Thediversity of approaches to NT - from microelectronics,materials science, biotechnology, pharmacology, etc. – maymake this unlikely. However, the IT example is one where,despite microelectronics being an obvious case of a heartlandtechnology, there are numerous related technologies, somecomplementary (e.g. software) and some more independent (e.g.optronics and photonics). There is limited conceptual andempirical analysis to tell us how “tight” the cluster ofcore underpinning technologies in a technological revolutionneeds to be.

In any case, nano-transformation is a set of technologieswith extremely wide applicability. For example, take the AMcase. Barker5 indicates the potentially great pervasivenessof new materials technology by noting that the materialssectors of industrial economies may well account for 5-10% ofoutput, and that materials can account for as much as 60% of5 Barker (1990)

the cost of manufactures. AM currently constitute a smallproportion of the total materials markets - and thus the AMrevolution, if such a thing exists, has a long way to go.But their role in many strategic applications (aerospace,etc.) is already substantial. Barker identifies a series ofcharacteristics that distinguish new from traditionalmaterials technology: these are summarised in Table 1, andare clearly of such importance to modern industry as torequire close attention by producers and users of materialsof many kinds. The new properties that AM possess may wellcreate opportunities for new products and processes.

3.2 Innovation and Diffusion

The usual S-curve of diffusion was originally introducedto describe the pattern of uptake of a particularartefact or process across individuals, regions, orenterprises. Using it to describe patterns of uptakeacross sectors is an obvious step, and one that is usefulfor thinking about skills and employment implications ofa new technology. A problem with SKIN, however, is thatNT is now just one technology, and is potentiallysomething that can be embodied in many types ofapplication and artefact. It will be more difficult toassess (and predict) diffusion trends for NT than it wasfor, say, microelectronics – where surveys did enquireinto the share of products and processes employingmicroprocessors, for example, or produced or controlledby robotics.

Table 2 presents one listing of applications of NT thatcan fairly readily be translated into sectoral areas ofapplication. An effort to fit a prospective chronologyto NT applications was made in a study for Texastechnical colleges by Vanston and Elliot (2003). Theiroverview is presented in Table 3. A recent and moresectoralised forecast from Harper (2008) features inFigure 3. . The trends in this forecast seem to refer toend-uses, at least to a large extent. Thus, electronicsdoes not appear as an application area in its own right -

though nanoscale engineering is already a reality in thissector. Notable is the expected shift from high-techmilitary/aerospace applications to those focusing onsocial concerns such as health and nutrition.

Table 1 Characteristics of New Materials Technology 6

6 Based on Barker, 1990, op cit.

Information Content R&D, Processing and design expertise are much higher proportion of total costs, energy and raw materials lower.

Complexity Greater control of materials microstructure, soAM often constituted by a series of phases yielding a desired microstructure with specific properties. Requires multidisciplinary knowledge inputs.

Integration of Function Packing of more performance characteristics into smaller areas and volumes, reduced steps in manufacturing process.

Added Value High unit prices related to information content and level of processing required.

Variety Broad and diverse range of materials, reflecting variety of manufacturing methods and raw material inputs, and amount of scientific and engineering knowledge, now available; so scope for more customisation to user requirements.

Market Size Already having impact on nearly all sectors ofmanufacturing industry, especially high-tech sectors; likelyto have multiplier effect across whole economy.

Market Growth Whereas many traditional materials have mature or saturated markets, AM display rapid growth.

Life Cycle Apparently short, reflecting increased competition among continually evolving materials, and shorter life cycles of products in which used.

Table 2 Applications of Nanotechnology

1 MEDICINE1.1 Diagnostics; 1.2 Drug delivery; 1.3 Tissue engineering2 CHEMISTRY AND ENVIRONMENT2.1 Catalysis; 2.2 Filtration3 ENERGY3.1 Reduction of energy consumption; 3.2 Increasing the efficiency of energy production; 3.3 The use of more environmentally friendly energy systems; 3.4 Recycling of batteries4 INFORMATION AND COMMUNICATION4.1 Memory Storage; 4.2 Novel semiconductor devices; 4.3 Noveloptoelectronic devices;; 4.4 Displays; 4.5 Quantum computers5 HEAVY INDUSTRY5.1 Aerospace; 5.2 Construction; 5.3 Refineries; 5.4 Vehicle manufacturers6 CONSUMER GOODS6.1 Foods; 6.2 Household; 6.3 Optics; 6.4 Textiles; 6.5 Cosmetics

Source: Wikipedia “List of nanotechnology applications”at

http://en.wikipedia.org/wiki/List_of_nanotechnology_applications accessed 27/03/09

Note: this is simply an illustrative list. The author isaware of NT applications being developed across a muchwider span of industries – for example, nano-sensors foragriculture and environmental use – and of many moreapplications within the sectors listed above (e.g.nanosystems for fuel cells and batteries).

More elaborate forecasting of skill requirements willneed to have more detailed and better-grounded estimatesof NT uptake in various application areas and sectors.Probably the best step would be to identify a key set ofNT applications and then seek to estimate thee extent towhich they will be taken up, at what speeds, in differentsectors, countries, types of firm, etc.

Table 3 Vanston and Elliot’s Forecast of Commercial Realisation of NT

Short-term (0-3years)

Mid-term (3-5years)

Long-term (5+years)

Modest Commercial OpportunityInstrumentation, tools and computer simulation

Instrumentation, tools

Instrumentation, tools

Nanomaterials (metal and ceramic nanopowders, fullerenes, carbon nanotubes)

Nanomaterials (metal and ceramic nanopowders, fullerenes, carbon nanotubes)

Nanomaterials (metal and ceramic nanopowders, fullerenes, carbon nanotubes)

Important Commercial Opportunity- Life sciences

(diagnostics)Life sciences (diagnostics, screening and taggingtechnologies)

- Electronics/IT/optical devices

“Smart” nanomaterials

- Computer simulation -Large Commercial Opportunity

- Life sciences (drug delivery)

Life sciences (drug delivery, design and development)

- Electronics/IT (data storage, microprocessors)

Electronics/IT/optical devices (datastorage, memory, optical devices, molecular and quantum

computing)

NEMS (nanoelectromechanical systems)

Source: based on Vanston, J and L Elliot (2003)

Figure 3 A Forecast of NT Markets

Source: Harper (2009)

In the absence of solid data, or even of expertjudgements that allow for sufficiently detailed analysis,all we can do at this point is to present a schematicdepiction of the sorts of development that are liable tooccur, if NT proves to be a generic technology, or set oftechnologies, with applications across many domains.This future seems entirely plausible, given the strikingwealth of applications that is already emerging fromfirst generation NT. Thus Figure 4 outlines the likelyemergence of multiple diffusion curves, as new nano-toolsare developed – applications of various kinds, adoptedacross wider or narrower sets of user sectors. It doesnot attempt to fit exact time lines or specify the leadand lagging user sectors – this is a matter for detailedexamination by experts (though it is possible to makesome generalisations about the types of firm and sectorthat are liable to be lead users). We are always liableto be surprised as products find applications well awayfrom those considered by the pioneering inventors andinnovators. We are also likely to find that the take-off

of some diffusion curves is slower than anticipated, asregulatory, technical or skills problems kick in. Thisis probably one reason for many forecasts of rapidgenerational change already appearing premature. (SeeNIA, 2007, for a striking and more realistic view ofgenerational change.)

4 Innovation and the Demand for Skills

4.1 Nano-skills?What do we know about the skill requirements associatedwith technological change? Clearly the picture will vary

Figure 4 Diffusion Curve of new Core Technologies

Niche applications

Multiple Applications:

successive mass

marketsTrigger and

enabling applications

Ongoing Qualitative Change

Time – may be DECADES

Adoption of

different

applications in

various firms,

sectors, etc.

Successive New Technologies

with improved characteristics, diffusing more

rapidly

Cascading Applications,

new platforms and potentials

(ceilings of curves may

differ)

Common Platforms

across the different sorts of innovation discussedearlier, so let us focus on major technological changes,affecting many sectors and processes, such as NT isbelieved to be.

The frameworks set out earlier give us some basicperspectives on how the nature of key technologiestypically changes (for instance, toward more standard andusable designs) as the market develops (typically an S-curve of diffusion and a proliferation of applications).These frameworks do not tell us directly about the speedof development in any given area – though we shouldcertainly beware the trajectory of hype! – but they doprovide some implications for skills and relatedinstitutional issues. We could for example, anticipate along-term growth (and for a long period acceleratinggrowth) in the requirements for specialist knowledge ofthe fundamentals of the new technology – in R&D and inemerging applications. Accompanying this, at first withsome lag but ultimately outpacing it by many times, willbe requirements for skills in the applications of the newtechnologies, fusing domain-related skills with goodunderstanding of the technology. With another lag, manymore workers will be affected in the sense that they willneed less specialised knowledge in the technology, thoughpossibly more practical experience will be required aboutuse of the technology in specific circumstances.Typically, we would anticipate the formation of newprofessional classifications, qualifications,accreditation, and associations and networks over thisperiod; the reconfiguration of training courses andpossible some new institutions.

This might be a process unfolding over a few decades, orover many. Predicting the actual pace of NT advance isbound to be difficult. Some recent commentators areanticipating major market take-off in the next few years,for example. Despite the current economic gloom, TimHarper (of the consultancy CMP Cientifica) anticipates "a big take-off in 2011, and by 2012 the industry isreally going to be booming" (Guardian (London) 26/03/09,Technology Supplement p1). Nanotechnology pharmaceutical

and healthcare products are forecast to be worth $3.2tnby 2012, with military-use nanotech products taking 14%of the total market ($40bn). NT products for the motorindustry are put at 4%, those for foodstuffs at up to 2%,and other products designed addressing pollution willalso be significant. China’s role in NT will becomparable with those of the EU and US. In the samenewspaper report, Richard Appelbaum (Center forNanotechnology in Society, University of California)considers that by 2014, some 15% of manufacturing outputwill consist of NT products. Such estimates run into allsorts of definitional problems – are we talkingnanomaterials, systems, or things that happen to havebeen themselves shaped by NT devices? But it isstriking that after some years where the hype bubblearound NT seemed to have burst, it is once again beingpromoted enthusiastically in various media outlets.

When it comes to discussing SKIN, the situation issomewhat less clear. Some NT proponents are happy enoughto stress the potential for job creation. Thus Roco(2007), citing input from industry and academic experts,projected that $1 trillion in products incorporatingnanotechnology and about 2 million jobs worldwide will beaffected by nanotechnology by 2015… Extrapolating frominformation technology, where for every worker another2.5 jobs are created in related areas, nanotechnology hasthe potential to create 7 million jobs overall by 2015 inthe global market. Indeed, the first generation ofnanostructured metals, polymers, and ceramics havealready entered the commercial marketplace.” (The word“affected” seems rather slippery here.)

But there are few studies that examine more specificskill requirements, especially if we are looking forstudies involving substantial research into occupationaltrends, job announcements, or even industry opinion.7

Only a couple of such studies have been located whilepreparing this paper.7 Roco (2007) noted that “Small Times reported 1,455 U.S.nanotechnology companies in March 2005 with roughly half being smallbusinesses, and 23,000 new jobs were created in small startup ‘nano’companies”.

One such is by Vanston and Elliot (2003), prepared toinform technical colleges. They stress the uncertaintiesof NT development, though much of the report considersforecasts that are in general rather optimistic.Because of the uncertainties, basic skills that could beapplied in various related fields are emphasised.Colleges were recommended to prepare programmes (in acoordinated way) but not expect to deliver them untildemand becomes more clear – which could be the case veryrapidly, once take-off is achieved.. As alreadymentioned, the development of commercial applicationsacross various fields are liable to follow a sequence, asdepicted in Table YYY. Thus it is possible to speculateabout some patterns of emerging SKIN. The area with mostgraduate employment opportunities in the near future isnanomaterials: technicians’ jobs include productionsupervision, quality control, equipment calibration andmaintenance, user education, responding to customerrequests. . One area that is strategically importantbut may involve relatively few (very high skill) jobs isinstrumentation, tools and computer simulation. Longerterm prospects are bright in electronic, optical andinformation applications, though limited immediateemployment will be created: skills may be analogous tothose in computers and microelectronics. Life sciencewill require higher level skills than those oftechnicians, and this may be a bottleneck; NT may impacthealth professions in general; NT applications inenvironmental areas will require data processing,equipment operation and repair, and administrativepositions. This is a promising approach – SKIN analysiswill be much enhanced by reliable appraisal of thesectoral diffusion of NT – though in practice the areasexamined seem fairly restricted.

The report presents a series of speculative jobdescriptions with notes about skill requirements – forexample a medical technologist to perform clinicallaboratory tests and procedures using NT, withappropriate advanced technical skills. Other positions

include mechanical and electronics technicians,machinist, mechanical designer, and so on.

In one very interesting study – which could well do withan update - Hendry (1999) considered three new technologyindustries – optoelectronics, biotechnology, and advancedmaterials. Interviews and literature reviews were usedto explore skill needs in these three areas, in the UK.Advanced materials (AM) is the closest to our field ofNT, but since it does not specifically focus on the nano-level, a range of other materials activities are involved(and of course, non-materials focused NT is notexamined). Advanced materials were defined aspolymers, ceramics, and high-performance metals, andcomposites and laminates of these. The materials area iswell-established in the UK, with, for example, anInstitute of Materials – optoelectronics on the otherhand was highly fragmented. Hendry noted that the take-up of AM was slower than expected, with low-cost high-value products dominating, mainly produced by SMEs aslarger firms had withdrawn from the area at the time ofwriting. Fairly rapid growth was largely being driven bydemand from aerospace and automotive industries.

Hendry, (1999, p6) noted three key sets of AM skills:

“(i) fundamental understanding of the(specific) materials concerned, with skills insynthesis, design, processing, and fabrication;

(ii) supporting infrastructure (generic)technologies such as ultra-precise measurementand testing techniques, modelling andsimulation;

(iii) project management skills andappropriate organisation to carry out concurrentengineering, in which the design of a product isdone in close conjunction with the design of themanufacturing process, and customers andsuppliers are brought into the design processearly on, in order to meet ever-decreasingproduct development cycles. “

Studies of skill requirements for materials engineers hadstressed the uncertainty involved in developing AMapplications. This uncertainty calls for closecollaboration with supplier firms, the researchinfrastructure, and customers, underpinning the need for“soft skills”. Soft skills are described by Hendry asencompassing: creativity, problem solving, proactivity,communication skills, business awareness, and the abilityto use and integrate other disciplines. The variety ofevolving applications calls for this latter,interdisciplinary or interprofessional ability. In theIT world we are increasingly hearing discussion of the“T-shaped” individual, who combines deep specialistknowledge of one area with awareness of the terminology,principles, and issues in other areas of business and/ortechnology. A similar appraisal emerges here. Hendryquotes Kaounides (1995) as describing materials scienceas a

“multi-disciplinary science requiring inputsfrom solid-state physics, chemistry,metallurgy, ceramics, composites, surface andinterface sciences, mathematics, computerscience, metrology and engineering…. rigidseparation of the different disciplines isbecoming inappropriate … barriers or boundariesbetween them are beginning to erode. Theexamination of elementary particles, atoms andmolecules cuts across materials whatever theirorigin....” (Kaounides, 1995:15)

Such discipline-spanning knowledge and skills are oftennoted in connection with innovative industries andsectors. This is one reason why Hendry concluded thatthe issue of skill shortages is rather complex. Oftenthe right skills might exist, but not necessarily theoptimum combinations of skills.

Abicht (2009) presents results from a survey of 178German nano-businesses. Finding that over half of theemployees of these companies hold University degrees, herecognizes that this implies that the firms are highlyresearch-intensive character. (The smaller the company,

the higher the proportion of employees with suchdegrees.) The remaining employees are skilled workers(20%), master craftsmen and technicians (10%), with justunder 10% in administrative or unskilled jobs. Interms of expectations, the firms thought that the shareof skilled workers would rise with the shift fromresearch/development to production/service. Asignificant growth in job numbers is projected, withpersonnel in small and medium-sized nanocompaniesgrowing from 27,300 employees in 2008 to 43,200 by 2013.(Most of this growth is expected to be in the next twoyears!) Half of the companies intend to meet theirdemand for further education through external educationalinstitutions. Abicht argues that cooperation withuniversities and R&D institutes is an essential means ofknowledge transfer for nanotechnology companies, becauseof the need for interdisciplinarity. But the companiesreported that few knowledge providers have successfullyadapted to the demands from this new industry. It wassuggested that providers of further education have torespond in particular to the requirements of smallercompanies.

Training and other courses have difficulties indetermining how to combine more fundamental and morespecific knowledge of the technologies and applicationsinvolved, and combining these with skills in projectmanagement and collaboration. The academic communitymay stress inter-disciplinary knowledge, while industryseeks to deploy application-specific knowledge (andcomplains about the limited development of skillsspecific to particular subsectors). Training bodies needinformation on the extent to which a particular depth ofknowledge in each of the three sets of skills will berequired by employees at different levels. (Hendrynoted, for instance, the importance of intermediateskills at technician level.) Industrial success willrequire skills at all levels involved, and to the extentthat there is a problem, it seems to be one of achievingan optimum balance of skills.

4.2 Skills and Product Cycles

The studies of nano-skills that have so far beengenerated are necessarily providing us with snapshots ofSKIN at an early stage of industrial development. Theskill requirements that we see now may provide onlylimited insight into future quantities andconfigurations.

Here we can recall the discussion of product cycles.Tether et al (2005) discuss skill requirements over threephases of the life cycle. (1) In the early, ‘fluid stage’of the product or industry, the key skills are those ofentrepreneurs (often combined with scientific ortechnical specialists, and skills in marketing).Production skills tend to be more general and adaptable,rather than specific and rigid, as production workersadapt to rapidly changing technologies and demand. (2) Ina later ‘transitional stage’, with its shift from productto process innovation, management skills become morefunctional and ‘scientific’, whilst those of theworkforce become more specific, with a more precisedivision of labour. New specialist equipment may atfirst augment the skills of skilled workers. But overthe course of the mature phase of the industry the labourforce becomes increasingly deskilled as equipment becomesfamiliar and routine. (3) In the last phase of theproduct cycle, production is increasingly ‘offshored’, tocountries with lower costs of production (especiallylabour) – this offshoring pattern was classic formanufacturing industries, and seems to apply to someelements of modern services, too, as new IT can be usedto link service suppliers and users. Some high valueadded knowledge intensive activities – design, ongoingR&D (e.g. for product differentiation and new products),marketing, and strategic management) – may well remainbased in the home country. The skills involved aremanagerial command & control skills and for a small“elite”, while the bulk of the workforce has with low orunspecific skills (we would add, sales and logistics tothis).

This approach has considerable virtues, but when we areconsidering a set of products such as those ofnanotechnology – where the products of today’s nano-industries may well be critical inputs for a wide rangeof user industries – it makes sense to differentiatebetween the industries innovating core nanotechnologies,and those that are making use of these in applications.These latter “user” industries may themselves often beinnovators, creating new applications and nano-enabledproducts and processes. Tether’s account is mainlyfocused on the core nanotechnology innovators and thoseof the user industries that are themselves highlyinnovative users of nano-capabilities.

Jones et al (2007) went further in explicating the skillsrequired to drive innovation through the various steps ofthe emergence of a new product or process. The innovatorneeds skills for:

Sourcing and selection of ideas – skillsrequirements here are connected centrally with theidentification, collection and filtering of ideasfor innovation . Innovation managers (and employees)will ideally have an awareness of existing sourcesfor innovation both within and outside theirorganisations, and an ability to ‘scan the horizon’for - and develop relationships that will lead to -new sources of ideas and stimuli for innovation. Anability to interpret data (from market, consumer andcompetitor research etc.) and to evaluate theviability of innovation ideas is also crucial.Knowledge of and an ability to apply relevant IPprotection mechanisms constitutes a furtherimportant skill. Once an innovation idea is selectedfor possible progression to development stage,skills in arguing for its viability and potentialvalue – often in the face of strong competition fromcompeting projects – become paramount.

Development of innovation ideas – upon securingfinancial support for progression of an innovationidea, attention is directed to the practicalities ofdevelopment. Here, skills connected with theassemblage of development teams, allocation and

management of budgets and resources, generation ofappropriate spaces and conditions forexperimentation, sourcing and specifyingcomplementary inputs, and establishing networks andpartnerships are called into play. In thedevelopment of new artefacts/technologies, thesourcing of technical and design skills is often acentral concern.

Testing, Stabilisation and Commercialisation – a keyskill at the ‘stabilisation’ stage is evaluation ofrisks and benefits of continued experimentation.Cost effective innovation requires an ability torecognise the optimal point at which to call a haltto prototyping and the comparison of competingalternatives. It also requires a good knowledge ofthe preferences and requirements of the intendeduser/consumer base and an understanding of the waysin (and extent to) which an innovative product orprocess will meet with anticipated needs. Anunderstanding of the ability of potential users toderive benefits from an innovation (i.e., their‘absorptive capacity’) is also necessary.Stabilisation and commercialisation requires that aninnovating company has the skills in place to ensurereproducibility of an artefact or service at anacceptable cost and price (technical, engineering,design and marketing skills are often at the forehere). Commercialisation also requires thatattention is afforded to ‘capturing value’ from aninnovation – here, skills associated with managingrisk and deriving appropriate roll-out strategiesare foregrounded.

Implementation and diffusion – marketisation,implementation and diffusion are frequentlyunderstood to be connected intimately with projectmanagement and technology transfer skills. Beyondthese, skills in managing and coordinating value andsupply-chain relationships, and in evaluatinginnovation practice and performance are crucial.Reflexivity is becoming an increasingly importantcomponent of innovation practise as firms recognisethat collection and evaluation of data (i.e.,

knowledge management and intelligence generation)can result in the development of improved innovationprocesses.

These are pointers to the skills that are required formajor innovations – though in general terms, theimplications for nanotechnology development are easy todeduce. But it is apparent that the quantitatively mostsignificant SKIN emerge as these major technologies areemployed. For example, current occupational data suggestthat around half of Europe’s workforce are using PCs anda third are using the Internet, which means that at leastbasic skills in a range of standard computer applications(word processing, spreadsheets, databases, browsers andsearch engines, etc.) are being utilised. These jobshave been transformed – while some clerical jobs havediminished substantially. These are wide-rangingchanges, even if the new skills are ones that mostprofessionals would regard as fairly basic – and that arebecoming part of basic secondary education in mostadvanced countries. In contrast, more advanced ITskills, and the sorts of technical work that go withthese, are much more restricted: only a few percent ofEurope’s workforce are IT professionals, and only half ofthese are actually located in IT sectors rather than usersectors.8

So we can also see user industries SKIN as evolvingthrough the product cycle. Figure 5 presents a highlyschematic illustration of how this might be represented,with the lead users being very much user-innovators,requiring high levels of understanding of the nano-toolsas well as of their own application domains, and withusers at later stages of the product cycle becomingprogressively more numerous, and requiring quantitativelymore skill in applying these tools in the particulardomains.

8 CEPIS (2006) estimated that 4.2 million IT practitioners within theEU, while approximately 180 million people are using IT at work.

This is highly schematic, and it will be necessary todifferentiate between numerous occupational positions –researchers, other engineers, technical support workers,production staff, managers, and so on, in order to mapthe entire range of affected occupations. Manufacturingand other application sectors will be supported byknowledge-intensive business services (KIBS), and by thesupply of trained staff and knowledge inputs fromeducation, training and research organisations. .But thepoint of the earlier discussion of multiple productcycles is that we are liable to see a series of S-curvesas specific NT applications are developed and adopted inspecific industry sectors. As these unfold, sorequirements will be experienced through the (changing)occupational structures of these sectors, with various

Figure 5 Product Cycle: Skills and Employment

Time

Diffusion over

population of users

New Technology Users

New Technology ProducersA) Entrepreneurial.

Adaptive workforce.Scientific expertise.

D) User-innovators.

B) Efficient production & supply chain management.More routine workforce.

E) Fusion of domain and new technology knowledge.

New workforce skill

requirements.

C) Marketing skills.

Outsourcing.New development.

F) New users: as E) (skill

requirements reduced?)

Established users: new product

development skills

Major employment implications – new jobs, job

loss

Moderate employment implications – new jobs

combinations of the skill classes emerging in producerand user sectors. Figure 6 combines Figures 4 and 5,to illustrate this.

The content will vary across sectors and applications inrelatively predictable ways, but much of the balanceacross the three types of skill will be shaped by workorganisation strategies – including those involvingoutsourcing and offshoring. New technologies alone donot dictate skill requirements and work organisation.

5 Conclusions

A longer-term examination of SINT requires that we seethem in terms of a cascade of product cycle and diffusioncurves. A succession of job creation and skillrequirements are liable to emerge, as the corenanotechnologies are enhanced by complementarytechnologies and by integration into various applicationareas, and as dominant designs and common platforms andstandards are established. The larger employment effectsare liable to lie in these application areas, though thisis not to underestimate the potential significance of theheartland industries (consider the importance of IT firmsin today’s world). Skill requirements evolve assuccessive applications emerge; it is plausible thatapplications will be further combined into innovationsthat themselves form the basis for new products and

Figure 6 Diffusion of Multiple Applications of new Core Technologies, and Skill Implications

Niche applications

Multiple Applications:

successive mass markets

Trigger and enabling

applications

Ongoing Qualitative Change

Time

Adoption of

different

applications in

various firms,

sectors, etc.

Successive New Technologies with

improved characteristics,

diffusing more rapidly

Cascading Applications,

new platforms and potentials

(ceilings of curves may

differ)

AD E F

CB

AD E F

CB

Skill and employment

implications cascade as

applications multiply

AD E F

CB

AD E F

CB

industries.

“Calibrating” this approach requires estimates of thespeed of uptake and the extent to which transformativeprocesses are themselves transformed through NT. Inaddition to the estimation of which industries andapplications take off at which speeds, it will benecessary to confront the arguments about howrevolutionary NT will be – will it essentially be anextension of existing nanoengineering techniques, orembody the more radical visions of molecular engineeringand “bottom up” NT. Disputes about these perspectivesbecome rather important for forecasting beyond the mediumterm (5-15 years) – and even within it. Scenarioanalysis may be one response to such disputes, withalternative scenarios reflecting the realisation of oneor other viewpoint..

But equally important for medium and shorter-termanalysis is a dimension that this essay has largelyneglected – the issue of international development anddiffusion of technology. Countries and regions alsodiffer in terms of these features, and one of the mostsignificant tendencies in global development has been therapid “catch-up” of emerging economies in IT. Hugeinvestments in R&D in biotechnology and NT are apparentin several emerging economies, leading some commentatorsto anticipate that these will pose challenges to theadvanced industrial economies – they may be “latecomers”in many fields, but are at the forefront of several areasof new technology. It remains to be seen how the balancebetween regulatory and entrepreneurial cultures, and thelinkages between various elements of national innovationsystems, leads to these challenges being realisedrapidly.

With such complex issues in play, it makes little senseto pin one’s hopes on the accuracy of a single forecast.Again, we suggest that it is most appropriate to combinescenario analysis, in which some plausible alternativesare explored systematically, with horizon-scanning and“weak signal” methods, to provide early warning of how

far the key elements of various scenarios are beingmanifest.

Bibliography

Abicht, L. (2009) “Qualification Structure and Demandfor Further Education of German Nanotechnology Companies”in W. Luther and G. Bachmann (eds) nano.DE-Report 2009:Status Quo of Nanotechnology in Germany Bonn, Germany:Federal Ministry of Education and Research, DepartmentNanomaterials; New Materials

Barker, B. (1990), Engineering Ceramics and High-temperature Superconductivity: two case studies in theinnovation and diffusion of new materials, DPhil thesis,PREST, Manchester

Bell, D. (1964), "Twelve Modes of Prediction—a.Preliminary Sorting of Approaches in the Social.Sciences." Daedalus, Vol. 93, pp. 845-80

Beniger, J. (1986) The Control Revolution: Technologicaland Economic Origins of the Information Society, HarvardUniversity Press

CEPIS, 2006, Thinking Ahead on e-Skills for the ICT Industry in Europe Brussels: Council of European Professional Informatics Societies; available at: http://ec.europa.eu/enterprise/sectors/ict/files/thinkingaheadone-skillsineuropereport_en.pdf

Cohendet, P., M J Ledoux and E Zuskovitch, 1991 “TheEvolution of New Materials: a new dynamic for growth” inTechnology and Productivity: the challenge for economicpolicy Paris, OECD

David, P. A. (1991) "Computer and Dynamo: the ModernProductivity paradox in a Not-Too-Distant Mirror." InTechnology and Productivity: the Challenge for EconomicPolicy, Paris: OECD

Freeman, C., (1974) The Economics of IndustrialInnovation, Penguin, Harmondsworth

Freeman, C. and Louca, F. (2002), As Time Goes By\; Fromthe Industrial Revolutions to the Information Revolution,Oxford: Oxford University Press

Freeman, C., Perez, C., (1988). “Structural crises ofadjustment, business cycles and investment behaviour” inDosi, G., Freeman,C., Nelson, R., Silverberg, G., Soete,L. (eds.), Technical Change and Economic Theory. Pinter,London, pp. 38–66.

Georghiou, L., J. Cassingena Harper, M. Keenan, I. Milesand R. Popper (eds) (2008), The Handbook of TechnologyForesight, Cheltenham, UK and Northampton, MA, USA:Edward Elgar

Green, L., B. Jones, and I. Miles, (2008) Skills for Innovation INNOGRIPS Mini Study 03, Global Review of Innovation Intelligence and Policy Studies, at: http://grips.proinno-europe.eu/knowledge_base/view/222/grips-mini-study-on-skills-for-innovation/

Harper, T. (2008) „From Micro to Nanoelectronics :Disruption, Convergence or Evolution?“ presented atEmerging Markets Semiconductor Applications TechnologySymposium, 7 May 2008 Suntec Singapore available athttp://www.semiconsingapore.org/ProgrammesandEvents/cms/groups/public/documents/web_content/ctr_023570.pdf

Hendry, C. (1999), New Technology Industries, London,DfEE, Skills Task Force Research Paper 10

Jorgenson, D. W., Ho, M. S. and Stiroh, K. J. (2008), “ARetrospective Look at the U.S. Productivity GrowthResurgence“ Journal of Economic Perspectives Vol. 22 no.1, p3–24

Kaounides, L. C. (1995), Advanced Materials, London:Pearson Professional

Miles, I. (1988), Information Technology & InformationSociety: Options for the Future, London: Economic &Social Research Council, PICT Policy Research Papers No.2; revised/reprinted as, “The Information Society:

competing perspectives on the social and economicimplications of Information and CommunicationsTechnologies” in W Dutton (ed) (1996) Information andCommunications Technologies: Visions and RealitiesOxford, Oxford University Press

Naisbitt, J. (1982) Megatrends Warner Books

NIA (2007) NIA Forecast of Emerging Technologies, London:Nanotechnology Industries Association, available at:http://www.nanotechia.org/lib/tmp/cmsfiles/File/NIA_TechnologyForecast_June2007.pdf

Roco M.C. (2007), “National Nanotechnology Initiative -Past, Present, Future” in W. Goddard III, D. Brenner, S.Lyshevski and G. Iafrate Handbook of Nanoscience,Engineering and Technology, 2nd ed., CRC Press (Taylorand Francis)

Rogers, E. M. (1995),Diffusion of innovations, (4th ed.).New York: Free Press.

Tether, B., Mina, A., Consoli, D., Gagliardi, D., (2005)A Literature Review on Skills and Innovation. How DoesSuccessful Innovation Impact on the Demand for Skills andHow Do Skills Drive Innovation? CRIC report for theDepartment of Trade and Industry London: Department ofTrade and Industry available at:http://www.berr.gov.uk/files/file11008.pdf

Utterback, J. M. (1996) Mastering the Dynamics ofInnovation, Harvard Business School Press

Vanston, J and L Elliot (2003) Nanotechnology: atechnology forecast Technology Futures Inc. for TexasState Technology College, Waco, Texas

Vernon, R. (1966),”'International Investment andInternational Trade in the Product Cycle”, QuartertyJournal of Economics, Vol. 80, pp. 190-207, 1966,

von Tunzelmann, G. N. (1978), Steam Power and BritishIndustrialization to 1860 Oxford: Clarendon Press