The New Digital Economy and Development

UN CTA UNITED NATIONS THE «NEW» DIGITAL ECONOMY AND DEVELOPMENT UNCTAD Technical Notes ICT Development ˚8 UNCTAD, DIVISION ON TECHNOLOGY AND LOGISTICS SCIENCE, TECHNOLOGY AND ICT BRANCH ICT POLICY SECTION TECHNICAL NOTE NO8 UNEDITED TN/UNCTAD/ICT4D/08 OCTOBER 2017 ‘’ Digital Economy Development1 Abstract: technical note frames ‘’ Digital Economy (NDE) including, prominently: 1) advanced manufacturing, robotics factory automation, 2) sources data mobile ubiquitous Internet connectivity, 3) cloud computing, 4) big data analytics, 5) artificial intelligence. main driver NDE continued exponential improvement cost-performance information communications technology (ICT), microelectronics, Moore’ Law. . digitization design, advanced manufacturing, robotics, communications, distributed computer networking (.. Internet) altering innovation processes, content tasks, possibilities relocation work decades. , features NDE . , sources data, smart phones factory sensors, sending vast quantities data “cloud,” analysed generate insights, products, services. , business models based technology product platforms — platform innovation, platform ownership, platform complimenting — significantly altering organization industries terms competition range leading-edge industries product categories. , performance ICT hardware software advanced point artificial intelligence machine learning applications proliferating. features share reliance advanced ubiquitous ICT, embedded growing platform ecosystem characterized high levels interoperability modularity. NDE appears poised extend organizational geographical fragmentation work realms, including indivisible geographically rooted activities reside front global chains, &, product design, knowledge-intensive innovation-related business functions. impact jobs international competition crucially depend pace change ability organizations societies manage . technical note discusses NDE defined, explores implications location innovation manufacturing, notably involving developing countries. implications smaller developing country firms discussed, positive negative scenarios society general. 1 technical note prepared Dr. Timothy . Sturgeon, MIT Industrial Performance Center contributions guidance Torbjörn Fredriksson (team leader) Diana Korka UNCTAD. note commissioned ICT Policy Section UNCTAD background Information Economy Report 2017 (UNCTAD, 2017a). feedback Judith Biewener, William Bonvillian, Masud Cader, Mark Dallas, Stefano Ponte, Marshall Van Alstyne, Hugh Whittaker, Ezequiel Zylberberg participants “Manufacturing activities creation: industry 4.0, global chains, circular economy” seminar held dSEA, Universitá degla Studi di Padova, Padova, Italy, 4, 2017, “ Digital Revolution GVCs” session Society Advancement Socio-Economics 28th Annual Conference Lyon, France, July 1 2017. Financial contribution Government Sweden gratefully acknowledged project. ii TABLE OF CONTENTS . INTRODUCTION ................................................................................................................................................. 1 . WHAT IS THE ‘NEW’ DIGITAL ECONOMY ................................................................................................ 2 ADVANCED MANUFACTURING, ROBOTICS AND FACTORY AUTOMATION ................................................................................................. 3 NEW SOURCES OF DATA AND THE INTERNET OF THINGS ......................................................................................................................... 4 CLOUD COMPUTING ......................................................................................................................................................................................... 5 BIG DATA ANALYTICS ...................................................................................................................................................................................... 6 ARTIFICIAL INTELLIGENCE ............................................................................................................................................................................. 7 . BUSINESS MODELS AND INDUSTRY ORGANIZATION IN THE NDE ................................................... 8 THE IMPORTANCE OF PLATFORMS ................................................................................................................................................................ 8 OPEN INNOVATION, OPEN STANDARDS ...................................................................................................................................................... 12 GLOBALIZATION AND MODULARITY ........................................................................................................................................................... 14 . THE NDE AND DEVELOPMENT .................................................................................................................. 16 MANUFACTURING IN THE NDE, WHAT ARE THE TRADE-OFFS ............................................................................................................ 16 geography advanced manufacturing ............................................................................................................. 19 INNOVATION IN THE NDE ........................................................................................................................................................................... 20 geography innovation NDE ................................................................................................................... 22 LINKAGES BETWEEN MANUFACTURING AND INNOVATION IN THE NDE.............................................................................................. 23 SCENARIOS FOR DEVELOPING COUNTRIES IN THE NDE .......................................................................................................................... 24 . CONCLUDING REMARKS ............................................................................................................................... 26 WHAT’ NEW ABOUT THE NDE ................................................................................................................................................................ 26 WINNERS AND LOSERS, OPPORTUNITIES AND RISKS ............................................................................................................................... 27 DATA PARTITIONING, NEW DIGITAL DIVIDE ........................................................................................................................................ 29 REFERENCES ......................................................................................................................................................... 31 1 . Introduction Change air. public debate focused, increasing frequency urgency, imminent arrival “4th Industrial Revolution” creating “” digital economy (NDE) powered advanced “cyber-physical” systems spanning “advanced” manufacturing, transportation, services, biological systems (Rose, 2016; Schwab, 2015, 2017). technical note adopts term NDE frame set technologies processes prominently include: 1) advanced production equipment, robotics factory automation, 2) sources data mobile ubiquitous Internet connectivity, 3) cloud computing, 4) big data analytics, 5) artificial intelligence. technologies processes based, , advanced information communications technology (ICT). poised dramatically reduce demand routine tasks transform location, organization, content knowledge work. Broad policy questions include: • build systems underpinning NDE • world NDE emerge, • NDE alter demand labour skills • degree NDE alter balance power incomes industries societies • NDE introduce unacceptable privacy concerns cyber-security risk, risks concerns manageable outweighed benefits users infrastructure NDE built deployed, benefits risks distributed geographically societies, implications developing countries, places lagging terms technology adoption (UNCTAD, 2017a). NDE reinforce patterns uneven development provide level playing field developed firms, communities, regions countries emerging platform ecosystems NDE concentrate wealth market power technology clusters core platform owners located, relegating places create complements platforms dependent status, technology platforms provide lagging firms tools needed move added segments global chains (GVCs) develop innovative products early answer questions, balance, technical note places questions context introducing series concrete examples NDE developing, explores scenarios related location production innovation. Technological change globalization driven fragmentation organization location industries time. understand characteristics, geography social impacts, NDE measured. transactions interactions NDE electronic, cross borders easy detection characterization, ability official statistics measure basic economic indicators investment, trade, profits hampered. hand, 'big' economic data data agencies overcome problems. 2 technical note seeks define NDE, explores implications dynamics location innovation manufacturing, discussed implications developing countries. concluding remarks implications NDE industry society. . ‘’ Digital Economy , prudent, place NDE context underway decades, including arrival mass market personal computers mid-1980s, maturing digital design tools computerized manufacturing 1990s, boom outsourcing offshoring 2000s extension manufacturing services, increasing fluidity global economy emerged multinational enterprises (MNEs) finally begun wrestle previously disparate corporate information technology (IT) systems semblance interoperability coordination.2 Internet underpinned, enabled accelerated trends, lies core NDE . words, “3rd industrial revolution,” based digital ICTs, set stage 4th. time main consumer applications augmented virtual reality (VR) technologies video gaming, consumer cloud computing data storage remote data processing, deeply corporate applications industrial systems, emergence NDE advanced. NDE emerging combination technologies, ICT space, pervasive mechanical systems, communications, infrastructure, built environment, playing increasingly important role, social political life, research, manufacturing, services, transportation, agriculture (.. precision farming agricultural robots3) ( .. UNCTAD, 2017b). technologies underpinning NDE, importantly rough order maturity, include: 1) advanced robotics factory automation ( referred advanced manufacturing), 2) sources data mobile ubiquitous Internet connectivity ( referred Internet ), 3) cloud computing, 4) big data analytics 5) artificial intelligence (AI). transformative potential NDE realized elements mature, integrated, interoperable, broadly . simple, , uncontested, rapid process. Social technical factors, data security risks backlash digital divides, slow derail development NDE. eventual shape application technologies unknowable possibly unimaginable. Finally, technologies tend develop uneven unpredictable rates, deployment suffer fragmented competing standards. barriers pitfalls dim expectations ardent optimist, high requirement connectivity, interoperability 2 Integration suppliers digital business systems proceeding slowly, progress front. 3 Precision farming refers part meter--meter monitoring soil plant conditions, data areal imaging drones data-driven application water fertilizers ( https://earthobservatory.nasa.gov/Features/PrecisionFarming/). Agricultural robots, milking machines mechanical harvesters, decades gaining broader capabilities machine vision AI, advanced actuators, grippers, manipulators broader (Amelinckx, 2016). 3 integration organizations societies. , stressed McAfee Brynjolfsson (2017), impact revolutionary technology -estimated short term -estimated long term. Advanced manufacturing, robotics factory automation Industrial robots decades, steadily intelligent, agile, flexible. mechanized mass production revolution early 20th Century brought dedicated production equipment repeated operations (Chandler, 1962). time-consuming expensive change machines , range operations severely limited. 1980s 1990s, computer numerically controlled (CNC) production equipment earned label “robot” programmed -programmed increase product variety perform range operations -dimensional space.4 time flexibility speed industrial robots CNC machinery increased costs .5 , simple statistical process control algorithms relied shut adjust production processes automatically move tolerance. , rise computing power advent -cost sensor technology, collection sharing operational data machinery, factories, “predictive maintenance” , preventing processing errors machine breakdowns wear tear mechanical components predictable problems cross critical thresholds. Industrial robots (artificially) intelligent. robots agile aware surroundings, work safely side side people augment assist workers, replacing . “cobots” eventually perceive human movements automatically intelligently adjust movements routines fly machine learning (Hollinger, 2016). products digital realm, continued advance miniaturization mechanical micro-electronic technologies robots smaller, cheaper, power-hungry, powerful. extreme affordability scale emergent class small robot arms priced $300-$2,000 perform tasks, fabrication solid objects (3D printing), engraving, assembly sorting.6 small robots pre-programmed perform variety operations controlled real time smart phone. initial uptake educators7 “maker” 4 , CNC machines considered “robots,” term largely reserved machines flexible, vaguely humanoid arms. International Standard Organization defines robot “automatically controlled, reprogrammable, multipurpose manipulator, programmable axes, fixed place mobile industrial automation applications.” ( https://ifr.org/standardisation.) 5 Manufacturing technologies driven innovation. advanced experimental manufacturing environments, creation materials, complex pharmaceuticals synthetic biologics, difficult separate products processes (Bonvillian, 2017). 6 Dobot, start- Shenzhen, China began 2015 robotic arm priced $1,000. 7 education market significant explosion robotics laboratories secondary, tertiary, primary education. 4 hobbyists, improved performance (faster, accurate, capable carrying heavier payloads) making -cost robots suitable small, -volume manufacturers.8 Industrial robots recognizable part “intelligent factory,” advances inventory control autonomous industrial vehicles contributing productivity quality increases factories. combining image recognition augmented reality (AR) technologies, human operator guided complex variable assembly steps receive real time input expert location remotely view operator .9 sources data Internet Productivity improvements long based data, Frederick Winslow Taylor’ time motion studies workers early 20th Century, Japan’ “lean production” principles continuous improvement total quality management 1970s 1980s, “ Sigma” movement “ defects” US 1990s. Improvements measuring , time takes worker part bin, dimensions part, number defects coming specific supplier, relationship air filtration levels humidity yields semiconductor plant. Needless , art measurement sophisticated, Taylor’ famous stopwatch, cards kanban board signalling parts, barcode readers follow parts factory traceability supply chain, laser scanners test equipment check tolerance parts nanometre. Today, -cost sensor technologies widen scope measurement factories. Sensors embedded, robots production equipment, operator wearable devices, industrial vehicles, buildings, pipelines. enabled falling cost sensors continuously, periodically automatically transmit data power bandwidth requirements. Wireless transmission10 introduces levels flexibility regard sensors practical, allowing remote devices linked centralized systems. data collected - basis multiple sources multiple points system, vast amounts data accumulated time. examples industrial sector, manufacturing main source innovation data-driven productivity improvement. today, - digitization advent Internet data gushing corner industry society, sensors built production lines, electric meters, security cameras, customer service call logs, mouse “clickstreams” online activity, point--sale registers, Facebook “likes” status updates, voice commands Amazon Echo Google Home. 8 small robot arms robots occupy emerging “prosumer” market space appears opening due ubiquity cost mobile computing, wireless communication, GPS, ICT technologies. GPS stabilized, tablet-integrated flying robotic cameras (drones), produced DJI, Shenzhen start-, 3D mapping, inspection, documentation, search rescue. products lie simple inexpensive toys complicated extremely expensive industrial equipment. 9 Google Glass, failing catch hold consumer realm, revival “Enterprise Edition” type application (Bershidsky, 2017). 10 IoT wireless transmission achieved open source protocols Bluetooth, 3G/4G, WiFi, membership- alliance-based protocols ZigBee -wave, proprietary 3rd party protocols Sigfox (Greenough Camhi, 2016). 5 consumers, -vaunted Internet (IoT), people check contents refrigerators store milk, conversely order pay milk touch screen front refrigerator door, technology search application ( market). , devices, televisions automobiles, connected Internet automatically sending information stored “clouds”. platform owners care profits selling devices collecting detailed information users interested , buy, , gain capability push targeted marketing consumers moment place purchases. limits technologies unknown. day passes revelation data collected connected machines, , mapping potential sharing ( advertisers makers home digital assistants) floor plans room content data models Roomba robotic vacuums (Astor, 2017). Users ( ) perceive technologies invasive, watchdogs decry privacy, data identify specific users carry “metadata”: , , activities transactions, lie sources knowledge, innovation, profits, effectively utilized. Cloud computing Cloud computing signal shift centralization computing architecture. UNCTAD (2013, . 2) describes cloud computing system “enables users, Internet digital network, access scalable elastic pool data storage computing resources, required.” significant difference 1970s mainframe era remote computing storage longer centralized enterprises, distributed Internet, accessible authorization means pay access. largest vendor PC software, Microsoft, earns revenue cloud-based software (Levy, 2017). downloading programs installing PCs, web browsers means manipulating software data reside online. Storage space, applications, platforms rented ( monthly subscription), updated vendor. shift software product, purchased physical form disk download, software---service (SaaS) platform---services (PaaS), means software , suitable Internet connection, --date. storage, shifting PCs private networks cloud. average consumer, transition cloud computing slow, cloud services primarily online storage, backup, synchronization devices. corporate systems, contrast, cloud offers huge advantages uptake faster. , cloud provide computing infrastructure flexible regard scale, location, capabilities. Storage, software, services rented predictable subscription fee, accessed Internet. Investment software development hardware shared user base maintenance, upgrades, desks outsourced. result, large companies computing resources.11 , 11 , involved -year transition, largest user Internet bandwidth United States, Netflix, relies network service providers Comcast, Verizon, AT& store serve streaming content customers, cloud resources Amazon Web Services search, personalization, sensitive billing payments portion system (Brodkin, 2016). 6 externalization aggregation computing resources data cloud justifies modifier “” term NDE. data integrity security obvious risks, cloud offers greater promise place data analysed vast quantities. promise heightened -increasing flows data entering cloud day. question cloud-based data answered, part, field data science, big data analytics. Big data analytics cloud place store data run programs. receptacle huge volumes data flowing autonomously IoT. sensors devices IoT automatically feed data cloud, duly tagged fine-grained meta-data ( source, location, .), “mined” insights enable “data-driven decision making” businesses, government agencies, person organization access data means carry analysis. simple easy, large sample sizes increase robustness analysis, introduce risks. central challenges analysing big data develop methods screening “noise” poor data quality (including incorrect metadata tags) weighting interpreting data sources kinds. industrial side, companies General Electric offering host generic, industry- specific, customized data analytics services manufacturers Predix platform. realms public health, social science, marketing, innovation, possibilities emerging “crowd-sourced” insights, tracking timing location disease outbreaks real time analysis Google search terms (McAfee Brynjolfsson, 2016, 2017). Reliance user reviews central feature range online retail businesses, -commerce sites Amazon Alibaba, travel services TripAdvisor, Hotels., AirBnB, HomaAway, -Trip. data targeted marketing improving operational performance , McAfee Brynjolfsson (2016) identify aspects “big” data: volume, velocity, variety. scalable accessible, cloud allowing businesses accumulate unprecedented volumes data, real time, wide variety forms (written, numerical, audio-visual). Volume variety increase accuracy analysis (.. “wisdom crowd,”) high velocity improves responsiveness relevance. volumes data staggering. McAfee Brynjolfsson (2012), “… estimated Wal-Mart collects 2.5 petabytes data hour customer transactions. petabyte quadrillion bytes, equivalent 20 million filing cabinets’ worth text.” biggest challenges big data making decisions data integrity. decision makers data altered original “Blockchain” powerful encoding data sharing method encrypts data, , time location stamps, values altered fact. data integrity data accuracy piece data means. defective sensor provide faulty data, , 10 degrees Celsius temperature sensor machine process operating range. Data accuracy meaning determined analysis, importance data integrity rises data pools large, data pooled publically organizations. blockchain systems intended facilitate. data unalterable, monitored 7 fully transparent “public ledgers”, underlying “crypto-currencies” Bitcoin Etherium, enables data sharing. Blockchain technology creates potential shared data layer enable automatically executable contracts royalty payment systems, distributed file storage, peer--peer retailing, secure crowdfunding, transparent polling corporate governance (Epstein, 2017). Similar verification obtained comparing data multiple sensors IoT. data owners undeniable power NDE, blockchain technologies transparent data verification methods shift emphasis cloud computing data ownership data analysis. words, access data, competitive performance shifts speed, quality accuracy analysis. Artificial intelligence cloud vast quantities data, analytics lead deeper understanding sources data (human machine) social business dynamics represent — including NDE functioning — AI, machine-learning algorithms, “predictions decisions increasingly automated , large scale” (Brynjolfsson, 2016). AI technologies publically , open sourced free, 2008. , date slow unstable mainstream .12 Advances microelectronics, powerful graphic processing chips (GPUs),13 large pools data analysed mathematically represented graphic matrices, allowing machine learning carried deep domain knowledge objects incorporated model. current excitement ( worry) AI coming gradual move “supervised machine learning,” humans tag images data define “” solution advance ( creates appearance machine intelligence) addition “unsupervised learning,” solution defined priori machines classify unlabeled data fly, allowing system performance improve human intervention (Mar, 2017). highly technical subject, . Facial recognition humans extremely adept , task historically eluded computers. dictionary definition AI “ capability machine imitate intelligent human behaviour,”14 computers recognize identify human faces definition artificially intelligent. Facial recognition, extension computer recognition common objects, involves characterization comparison hundreds, thousands unique polygons. modern GPUs identify millions polygons real time (data velocity) , combined databases hundreds thousands millions (data volume) faces (data variety), applications AI powerful practical . , -driving vehicles require instantaneous object recognition, systems image processing rates surpassing 150 frames developed (Redmon al, 2016). Artificial intelligence long history, part tied competing approaches, rule-based decision-making . machine learning (McAfee Brynjolfsson, 2017). Computers good 12 Key innovations “deep learning” efforts University Toronto Carnegie-Mellon University early 1990s. 13 Ironically, , advanced GPUs driven rich video gaming applications. 14 https://www.merriam-webster./dictionary/ 8 making decisions based logical rules, replicating “neural networks” brain machines “learn” create programming response stimuli, proved extremely challenging mathematical computer hardware perspective. , problem solving tasks escalated, rule-based approach ran limits programmers knowing defining rules ante. time, faster processing big data providing computing power wealth examples needed bring machine learning practical application. Machine learning occurs computers alter programming based data analysis, human learning creates neural pathways, referred “ propagation.” Machine learning improved human learning altering existing code — “ propagation” — time code created, increasing system performance rapid thoroughgoing . result, IBM’ Watson AI programs shifting research demonstration tools realm SaaS, provide 3rd party AI services business á la carte basis. , IBM shifting market focus Watson sole focus cloud-based big data analytics large organizations, easier incorporate platform entrepreneurs product designers.15 . Business models industry organization NDE technological advances NDE set business models industry organization characteristics , , meant dynamically cope growing system complexity. individual organization fully understand control underlying technology NDE specific domains operates. Collaboration fields — computer data science biology, political campaigning banking — , sufficient. Systems designed dynamically cope immense growing complexity breaking. systems NDE : 1) rely 3rd parties complementary products services, 2) draw communally held sources knowledge technology, 3) partitioned -contained, manageable, affordable, interoperable segments. words NDE based platforms, open innovation, modularity. importance platforms architecture NDE , continue , characterized set ( ) interoperable technology product platforms (Parker al, 2016; Kenny Zysman, 2016). complexity technologies embedded products services underlying NDE means single company ( country, region, technology cluster, matter) master, control, system elements. time, information technologies, including electronic control mechanical systems,16 developed set nested modules platforms based de jure de facto standards, stretching discrete functional elements (technology platforms) higher- level tools, hardware systems, software environments17 (core platforms) developers 15 uptake entrepreneurs, company offering “unlimited free” access Watson’ AI tools start-ups crowdsourcing site Indiegogo (Dalton, 2017). 16 Systems marry electronics moving machinery referred “cyber-physical” systems (.. Schwab, 2016). 17 Methods managing complexity ICT systems run deeper suggested , driven improvements cost-performance hardware (.., microelectronics). emerged higher-level approaches software development (.., object-oriented programming) design microprocessor architecture (.., embedding higher-level software “microcode” hardware shifting 9 create variety goods services users (higher-level platforms).18 modular system elements altered upgraded redesigning entire system, obvious limit depth complexity NDE. level nested ecosystems, 3rd party complementors (.., makers smart phones apps) emerged provide products services platforms customized enhanced variety markets. addition providing market opportunities 3rd party vendors, enhances platform. turn attracts users platform, turn attracts 3rd party vendors “network effect.” , modular systems, platforms form basis additional platform layers, stylistically depicted Figure 1 technology, core, higher-level platforms. notion platform layering suggests sequential series platforms underlie modern product-based competition. centrality "-sided" markets drive create network effects NDE strategies historically , sequential added chain appears platform ecosystems ways: 1) platform layer 3rd party complementors connected users (, platform owner acts intermediary suppliers buyers), 2) -added sequence platform layers flow higher level platforms. Van Alstyne al (2016) state, intermediaries ( department stores) long acted platforms connect producers users, creating -sided marketplaces. NDE, difference platforms, Amazon’ online shopping platform, easier set , cheaper maintain (albeit Amazon’ huge fulfilment centres bricks mortar) easily scalable ( scale generating vast pools data). concept added chain linked layered platforms illustrated smart phone. rests platform layers, turn harnessed mobile platform delivery higher-level product platforms, mobile social networking online retail ( Figure 2).19 result ecosystem shifting overlapping platforms consisting multiple platform layers ( owners), myriad platform users casual users, “-sided” markets stage connect 3rd party complementors, platform, platform users , finally, users. , Uber’ platform connects drivers riders, Amazon connects buyers product vendors, Airbnb connects apartment owners renters. deeper plumbing NDE, thousands 3rd party vendors providing complementary products services specific technology platforms, including cloud computing AI platforms. result, hardware-centric complex instruction set computing (CISC) software-centric reduced instruction set computing (RISC)). 18 Strategic questions companies elevate technology, products, services part dominant platform, great interest observers mainstay management literature (.., Parker al, 2016), focus . 19 basic level “upstream” chain, calls smart phone depend set general interconnect standards agreed industry level (2G, 3G, 4G), implemented consortiums alliances specific interconnect standards CDMA GSM ( Figure 3). mobile handsets complex time, semiconductor firms Qualcomm MediaTEK developed chip sets handle complexity system, Google’ Android operating system replaced proprietary operating systems, generally incorporated key software code ARM. Handsets designed Samsung depend embedded technology platforms, providers retail mobile services Amazon Facebook depend users owning powerful smartphones access services. 10 platforms invisible users (.., chipsets 3rd party technology embedded , ARM software mobile telecom chips meant run Android operating system). cases users aware platform (.., Amazon, Uber, Facebook products, services, personal networks link ). platform structure NDE final systems extreme levels embedded complexity broad range capabilities. lowers barriers entry 3rd party technology vendors, sell discrete modules, products services system; opens vast opportunities companies, Facebook, build higher-level platforms top - level platforms (.., PCs mobile phones). put simply, rich ecosystem technology product platforms, web services companies PayPal, Airbnb, Alibaba, create PC, smartphone, Internet, software programming languages build maintain websites. create cloud storage services collect analyse vast stores data. sum , NDE platform-based ecosystem ICT-based products services. rapidly evolving combination ubiquitous continuous measurement data collection. IoT data flowing sensor-laden factory automation business process systems Internet-connected user devices, smart phones including growing lost Internet-connected products, home appliances automobiles. IoT generating “big data” pools , reside “cloud”, mined analysed patterns correlations remain hidden, results fed AI systems machine learning automated decision-making suggest upgrades system elements , speculatively, entire system. Figure 1. Platform layering chain NDE ecosystem Source: . Sturgeon, UNCTAD. users Higher level platforms Core platforms Technology platforms • Provide broadly platforms services • ecosystem 3rd party complementors • Build specific apps, products services core level platforms • ecosystem 3rd party complementors • Personal small business apps platforms • Provide core technologies higher level platforms • ecosystem 3rd party complementors Downstream Upstream 11 Figure 2. Platform layering mobile telecom Source: Thun Sturgeon (2017). Platform owners, Facebook, Google, Amazon, Microsoft, Alibaba, General Electric, SAP, , big data AI center business models, capability analysis broader deeper larger swaths society connected IoT improved AI technologies developed deployed. result cycle data streaming connected factories users, data pooling cloud, big data analysis, AI- driven machine learning results continuous rapid cycles platform upgrading system- level leaps productivity innovation. true decisions AI affect structure operation NDE . cases loop data generation machine learning complete entire ecosystem platforms leap . stylized depiction process shown Figure 3. , factory user data generated set advanced manufacturing user platforms, operates modular system layered technology, core, higher-level, platforms. data flow, IoT, cloud, operates set nested modular platform ecosystems. cloud, tools analytic platforms applied sense big data, results subsequently operated AI tools. Finally, machine learning capabilities AI, theory, introduce potential entire system improve human intervention. question mark title Platform Upgrading box Figure 3, shape box, meant autonomous system- level platform upgrading speculative; idea remains realm science fiction. System -improvement machine learning today context specific software programs,20 larger scale, interconnected systems underpinning NDE (.. networks machines connected Internet). words, scope, medium terms, vast, diverse, poorly linked systems NDE connected coherent system, , engage machine learning. 20 Machine learning advanced high-speed securities trading (Deboeck, 1993). Retail social networking platforms Operating system Handsets Chipsets ( single multi-band) Specific interconnect standards General interconnect standards • Facebook LinkedIn • Alibaba Amazon • WeChat WhatsApp • Netflix, Google Search Biadu • Samsung, LG, HTC, Xaomi • Apple iPhone iPad • Android (Google) • iOS (Apple) • Qualcomm • MediaTEK • HiSilicon • Apple series • GSM (Europe) • CDMA (US) • TD-CDMA (China) • TD-LTE FDD-LTE (Global) • 2G • 3G • 4G • ARM IP cores Downstream Upstream 12 Figure 3. cycle platform upgrading Digital Economy Source: . Sturgeon, UNCTAD. Open innovation, open standards complexity multiplication technology domains NDE led industry players heavy “open innovation” (Chesbrough al, 2006) create resources needed develop ensure interoperability range -system elements, network infrastructure, operating systems, AI test datasets algorithms. Open innovation refers strategy relying external, shared crowdsourced resources integral part company’ innovation process. , annotated databases important developing testing accuracy machine learning algorithms. Large, pre-classified annotated image datasets development testing AI software consortiums companies research institutions. Microsoft’ Common Objects Context dataset includes 328 thousand images 2.5 million labelled objects 91 types. Object labelling images carried part “crowd-workers” hired Amazon’ Mechanical Turk platform, workers earn small amounts money performing simple, repetitive tasks online. training dataset 2016 13 ImageNet Large Scale Visual Recognition Challenge (ILSVRC)21 annotations 1000 categories 1.2 million images. ImageNet large-scale open-sourced image database project sponsored Google, Amazon, Stanford Princeton universities. Open source software design specification multiple applications NDE, including cloud services (.. Openstack22), cloud infrastructure (Open Compute Project), computer operating systems Linux (UNCTAD, 2013). Linux related code freely online websites moderated Linux Foundation. Linux active community software developers improvements, offer ‘distributions’ specific applications, provide informal online technical support engineers Linux software. initial motivation software engineers undermine -monopoly-level bargaining power held Microsoft PC operating systems, Linux’ penetration PCs modest. , Internet servers run Linux, host consumer electronics devices (Finley, 2016). time, open source resources free strings attached. Companies Red Hat, Canonical, SUSE developed -profit business models selling proprietary distributions Linux, tailored specific purposes, providing support large companies Linux products run IT systems. Google’ Android smartphone operating system, proprietary ‘distribution’ Linux, “ ” handset makers drive mobile users Google’ search services exposed advertising (Dallas, al, 2017). Open Compute Project (OCP) formed Facebook opted open -house data centre design specifications 2011. Data centres, critical Facebook’ huge cloud operations, considered core competency (Bort, 2015). company’ purchasing power allowed bring range partners -board, including cloud infrastructure services provider Rackspace23, adopted OCP 2012. 2016, OCP supported large data users Microsoft, Hewlett Packard Enterprise, Panasonic, Sony; makers complementary hardware components Intel, Schneider Electric Emerson Network Power; network carriers AT&, Verizon Deutsche Telekom (Miller, 2016). Google, holdout, submitted rack design 2016 (Novet, 2016). OCP open, data centre operators greater control flexibility regard components features, including storage technologies, central processing units, memory, operating systems. enabled Rackspace move - -shelf servers server rack architectures traditional vendors Hewlett Packard, Dell-EMC, Lenovo ( IBM), purchase cost servers vendors based Taiwan Province China, Quanta Cloud Technology, WiWynn (Wistron), Delta Group, Cloudline, joint venture HP Enterprise FoxConn (Miller, 2016). , shift proprietary open hardware designs opening opportunities companies based emerging economies. , Linux, open innovation means lowering costs breaking monopoly dominant design standard, initially; 21 ILSVRC competition research teams creating “algorithms object localization/detection images/videos scene classification/parsing scale” (: http://www.image- net.org/challenges/LSVRC/2016/index). 22 : https://www.openstack.org/software/. 23 Rackspace’ business model lease data center space server farm owners DuPont Fabros Digital Realty (Miller, 2016), specifications server technologies determined Rackspace. 14 key part - standards battles characterized ICT industry 1960s.24 largest companies, rely significant degree proprietary technologies, standards battles underpinning NDE differ earlier rounds tend embrace open innovation sooner, pit open standard . , IBM’ OpenPower initiative ( Foundation), centered PowerPC CPU technology supported Google, competes OCP data centre space. largest public cloud services ( Amazon AWS, Microsoft Azure), Google Compute Platform (GCP) tend draw multiple standards proprietary technologies hedge betting wrong standard, leverage negotiations, tool learning (Johnson, 2017). examples suggest, open innovation lowers cost information search knowledge creation. generates surprising “collaborations” competitors, setting Ezrachi Stucke (2016) call “frenemy” relationships dynamic landscape “virtual competition.” Open innovation involves internally developed legally protected knowledge IP resources. overlaps, distinguished “free innovation,” developers -rewarded; evaluation, replication, improvement “open source” collaborative; distribution fully functional products free peer peer (von Hippel, 2017). types innovation ecosystems, opportunities barriers smaller firms firms located core technology clusters standard setting tend concentrated collaboration dynamic, Silicon Valley (Sturgeon, 2003). Access resources free, high capability requirements exist full participation, -location helpful. Globalization modularity Computerization work, product design engineering, manufacturing logistics, services, enabled geographic fragmentation industries. computerization typically rationalization work processes explicit rules, facilitates standardization, transfer tasks stage , organizations (outsourcing) borders (offshoring), (offshore outsourcing). . 25 years, advancements ability codify transfer highly complex information stage chain , combined plummeting costs movement goods making voice data connections, enabled shift manufacturing countries China Viet Nam sourcing variety services countries India Philippines. Figure 4 stylized depiction sort chain modularity emerge firm, standardized, spread industry, eventually underpin development GVCs. top arrow depicts flow work integrated firm. Work initially carried teams, departments (ovals), tasks completed coordinated teams basis tacit knowledge exchange (orange scribble ovals meet). set format result team’ work, handing completed task team department. teams meet, -person, enable phase work carried , project product finished. work content tasks , highly complex, based tacit knowledge, slow unproductive. 24 Including de facto opening mainframe computer architecture based IBM’ 360 line introduced 1960s, battles personal computer operating systems CPUs 1980s, . 15 work routinized, codified, companies typically seek standardize output format work (orange hexagon information exchanged) seek external suppliers costs, increase flexibility free internal personnel higher work. Broad external suppliers industry, part driven suppliers seeking standardize interact customers, lead industry standard methods exchanging information chain, methods typically include 3rd party IT systems. companies modular linkages affiliates, type chain modularity effect decreasing contractual frictions, theoretically, , increasing potential outsourcing. Put concisely, chain modularity involves systematic partitioning tacit information, increase efficiency firm (intra-firm modularity), firms combined industry standards exchanging information (inter-firm modular), geographic boundaries, enabling outsourcing offshoring modular type GVC linkages (Sturgeon, 2002; Gereffi al, 2005). Figure 4. Stages chain modularity emergence GVCs Source: . Sturgeon, UNCTAD. processes theorized documented (.. Baldwin Clark, 2000; Dossani Kenny, 2013), main point chain modularity enabled — accelerated — technologies, tools platform ecosystems NDE. interoperability ( increasingly open) standards critical functioning NDE, platform structure generally modular, offering specific functionality linked standardized, defined, input output interfaces. platform designers, complementors, users add, subtract, update specific functions modules disrupting disabling larger systems. serve barriers network 16 entry firms capabilities conform standards, firms distant markets business partners. . NDE development analysis focused main features drivers NDE commonality platforms, open innovation, modularity structure - development. current discussion social consequences competitive dynamics , understandably , focused advanced economies ( United States) process forging NDE. firms heartland digital innovation, question compete NDE. governments workers, questions seize opportunities cope negative impacts national security, jobs, society generally (UNCTAD, 2017a). question asked ( Rehnberg Ponte, 2017) effects developing countries smaller firms digital tools. section focuses manufacturing innovation NDE, linkages innovation production, asks world place. Manufacturing NDE, trade-offs Industrialization long viewed path development traditionally, industrialization meant manufacturing. productivity increases globalization rendered path (Whittaker al, 2010), manufacturing plays important role developed countries, share employment advanced economies falls 10 cent. role manufacturing NDE Discussions topic, generally referred “industry 4.0” Germany “advanced manufacturing” US (Bonvillian, 2012; 2017), tend highlight specific features, including additive manufacturing ( 3D printing), intelligent adaptive robots safely work humans, ubiquitous measurement sensor-laden equipment, high levels traceability supply chain, development material processes. , general, role characteristics manufacturing NDE Advances process quality control manufacturing - industrial revolution, biggest coming 1830s interchangeable parts early 1900s mass production, large-scale dedicated production machinery dramatically lowered cost labour hours unit, cost product variety. tooling ( moulds dies stamping forms shape individual parts) product-specific expensive, full machine automation typically costly high-volume environments. high-mix production, switching product raises challenges shop floor ( met increased labour), engineering, quality control, materials management departments, constantly-shifting set specifications, requirements, material flows accommodated. Smaller scale production persisted high-cost locations form customized manufacturing, including prototyping, labour intensive part local innovation ecosystems, form high-mix production reasons market responsiveness, price sensitivity, requirements -location innovation. trade-offs regard production scale, product variety unit costs/labour hours summarized Figure 5. volume fully customized manufacturing tooling 17 manufacturing engineering costs small number products, easily benefit scale efficiencies inputs logistics, unit production labour costs high ( quadrant Figure 5). Products large-scale demand unit produced high volumes ( product variety), benefit automation (lowering labour hours item) scale efficiencies ( upper left quadrant Figure 5). High mix manufacturing lies extremes. high labour content, historically cost effective replace labour flexible machinery, higher overhead costs related logistics, inventory control, materials management. Figure 5. Shifting trade-offs advanced manufacturing: scale, product variety, unit costs Source: . Sturgeon, UNCTAD. markets grown internationalized, demand fragmented, product quality variety important (Piore Sabel, 1989). result, high-volume manufacturers sought compete increasing product variety unduly raising costs. “mass customization” (Pine Davis, 1999) accomplished means. , 1970s 1980s, Japanese firms ( Toyota) developed work organization supply chain management techniques. “Lean production” allowed greater product variety mass production environments ( benefits), part slightly increasing labour hours unit, humans historically flexible responsive machines (Womack al, 1990). , modularity form shared design elements, including underlying product “platforms” common components, enabled companies increase product variation keeping costs check, distinctiveness derivative products called question (Chesbrough Kusunoki, 2001). , computerization deployment robots rendered factories increasingly flexible, reducing time needed change production lines product . Automation, ubiquitous process measurement parts traceability advanced manufacturing environments motivated efforts increase productivity reach ADVANCED MANUFACTURING HIGHER VOLUME & VARIETY: MASS CUSTOMIZATION 18 extreme quality control goals, cases shifting parts million defect rates parts billion. , limits trade-offs depicted Figure 5 persisted, key cost trade- manufacturing variety scale.25 scale/variety/cost calculus change NDE Computerization digital integration design, manufacturing, capacity planning, supply-chain management long term trends, embedded enterprise resource planning (ERP) software. systems gradually introduced greater flexibility speed entire manufacturing chain, costs fallen ease risen, relevant smaller companies. ERP similar systems continue big part NDE continue improve. garnered attention, , rapid improvements production processes 3D printing dispense tooling . Additive manufacturing equipment 3D printers manufacture complex parts “printing” solid objects undifferentiated powders, gels, liquids, metal powders digital design files. 3D printing lowers cost customized extremely -volume production — prototypes — shown -hand portion Figure 5. Fast, -cost prototyping speed innovation process, support “-demand” manufacturing products occasional demand. , John Hart, head MIT’ Mechanosynthesis Group, states, “ nature, additive methods 3D printing replace high-throughput manufacturing operations” (Paiste, 2014), constant push increase production volumes.26 Companies deploying scalable “swarms” 3D printers increase throughput, systems developed companies Formlabs Stratasys (Kerns, 2017). Early adopters 3D printing, including aerospace companies Boeing, Sikorsky Airbus, produce large numbers parts — Airbus reportedly planning produce 30 tons 3D printed parts month 2018 (Kerns, 2017).27 technologies NDE shift volume curve upward outward Figure 5 enabling breakthroughs production processes heretofore exotic experimental materials process technologies, light-weight metals, advanced composites, integrated photonics, flexible hybrid electronics, advanced textiles, modular chemical processes, biofabrication, regenerative medicine, energy production efficiency, recycling remanufacturing (Bonvillian, 2017). 25 , variables cost, volume, variety determinants manufacturing location. Fragmentation GVCs, & geographically severed production, prevalent industries product categories require edge manufacturing processes, fabrication advanced microprocessors, frequent engineering process alterations needed ongoing basis, advanced capital equipment commercial aircraft. Products large, heavy, bulky, / delicate ( large screen televisions), variable color-matched parts sequenced final assembly ( passenger vehicle seats interior parts), tended produced close markets. location production affected industrial polices, regulatory requirements domestic manufacturing military hardware, local content requirements “offset agreements” trade domestic production market access (Gereffi Sturgeon, 2013). 26 evidenced Mechanosynthesis Group Robofurnace, “ automated system high-throughput synthesis nanomaterials, including carbon nanotubes” (Paiste, 2014). 27 production environment commercial aircraft, tens thousands components produced modest volumes, fall high mix, medium volume category Figure 5. 19 advanced manufacturing technologies mature, flexible machines higher throughput capabilities common, deployed high-mix high volume production environments, enabling higher product variety costs. result, NDE drive degree convergence “advanced manufacturing” profile upper quadrant Figure 5 larger-scale production greater product variety conserving costs labour hours unit, suggested arrows converging mass customization. geography advanced manufacturing advanced manufacturing located mass markets exist products, rapid replenishment required, cost sensitivity high, smaller manufacturing units proven durable growing part manufacturing landscape, higher unit costs (Reynolds, 2017). , Silicon Valley, large companies Apple famously turned -cost locations China mass production Mexico high mix, medium volume production, dozens medium volume contract manufacturers. include local firms (.. Amtech, Bentech, AlphaEMS) branch plants globally operating contract manufacturers (.. Sparqtron, Jabil, Flex, Benchmark, AQS). Boston technology clusters world similar agglomerations small medium-sized contract manufacturers firms engaged manufacturing products small scale. anticipate NDE characterized small-scale advanced manufacturing facilities located close consumers, similar farm--table model food, Sanjay Sarma envisions linked networks “distributed virtual factories” (Markowsky, 2012). vision, jobs migrate closer markets, transportation costs CO2 emissions , inventory requirements shrink, consumers desires met wide variety products produced specifically needed. increased flexibility promised advanced manufacturing, high cost initial deployment, effect increasing product variety large-scale manufacturing, putting downward pressure demand high-mix customized manufacturing located markets. Large scale production units, enduring advantage purchasing power, turn incentivizes high responsiveness -location suppliers, investments highly efficient transportation infrastructure, proximate institutional supports domain-specific education training. mind, imagine advanced manufacturing deployed places mass production place, China. , possibility. shift advanced manufacturing general, secular trend, systems deployed dramatically altering location production. result leap productivity, quality, traceability sorts manufacturing facilities, mass, high mix customized, global decrease demand direct manufacturing labour. possibility, arrives quickly, time labour market adjustment, worries observers NDE McAfee Brynjolfsson (2016) ( UNCTAD, 2017a). 20 Innovation NDE effects NDE innovation entrepreneurs companies elite power brokers core platform owners NDE smaller companies firms developing countries advantage NDE richness digital tools supporting innovation NDE, suggested Figure 6, include ways access finance, labour, inputs production services, customer service, sales channels, marketing. toolkit, potential NDE innovation entrepreneurship profound. today’ digital tools, product design requires engineering hours multiple rounds validation testing. Failed tests lead engineering additional hours redesign retest. point, designs work “ ” accepted sunk time expense threaten excessive. results high costs, long cycle times, - optimal products. tools NDE, big data AI, set change . Digital design simulation long time, moving simpler applications, automated circuit software testing, challenging applications, simulation mechanical systems (.. fluid dynamics, automotive drive trains, avionics). , world appears cusp leaps capabilities based cloud-based crowdsourcing, big data analytics, AI. Autodesk, maker AutoCAD, popular digital design platform multiple industries, including automotive, industrial machinery, construction, architecture. company’ Dreamcatcher design automation suite draws data captured large user base ( sources data), combines cloud -house expertise (data analytics), applies AI tools suggest options design engineers. company describes workflow : Dreamcatcher system designers input specific design objectives, including functional requirements, material type, manufacturing method, performance criteria, cost restrictions. Loaded design requirements, system searches procedurally synthesized design space evaluate vast number generated designs satisfying design requirements. resulting design alternatives presented user, performance data solution, context entire design solution space. Designers evaluate generated solutions real time, returning point problem definition adjust goals constraints generate results fit refined definition success. design space explored satisfaction, designer output design fabrication tools export resulting geometry software tools. Figure 7 shows variations bicycle frame designs automatically generated Dreamcatcher. systems users create test hundreds thousands design iterations matter hours, radically reducing cycle times improving product quality. Evaluating potential designs rich 3D augmented virtual reality (AR VR) common expensive (Barbier, 2017). digital fabrication tools 3D printers, mock-ups prototypes easier generate, enabling easier faster decision-making ultimate designs. 21 Figure 6. Examples tools emerging innovation ecosystem NDE Source: . Sturgeon, UNCTAD. Figure 7. Machine-generated bicycle frame design options Autodesk Dreamcatcher AI suite Source: Autodesk Research, https://autodeskresearch./projects/dreamcatcher. Digital design tools offer capabilities investigate cost supply-chain availability components, export finished designs instruction sets automated production equipment world, existing manufacturing clusters, close consumption, adjacent innovation. capabilities place, work hours needed create products fall sharply, expertise needed design high quality products. heavy engineering requirements satisfied software, designers rely subjective, artistic judgment, (.. focus groups, opinions collected social media), primarily technical skills. design tools emerging NDE present opportunities developing countries technical skills , knowledge local market preferences high. Capital • Crowdfunding, ..: - Allied Crowds, gofundme, Kickstarter, Indiegogo • Financial management tools, ..: - -line banking services - Quickbooks online, Kashoo, Invoicera Labour • Freelance marketplaces, ..: - Upwork, TaskRabbit, peopleperhour, 99 Designs, FlexWork • & executuve staffing, .. - Adecco, Manpower Operations, inputs production • Enterprise IT systems (hardware sotware) • Reference designs modules (.. chip sets) • Technology platforms (.. Microsoft Azure) • Digitial design tools (.. AutoCAD) • Cloud computing containers (ACS, Red Hat, .) • Analytic AI tools services (.. IBM Watson) • -cost robotics, 3DP, production equip (.. Dobot, ) • Contract prototype mfg. serv. (.. Jabil Blue Sky, Shapeways) • Contract & design services (industry-specific) Customers markets • Ecommerce platforms (.. Amazon, Alibaba) • -line mobile payments (.. PayPal, GoogleWallet, WePay) • Contract customer service desks (call center services) • Customer contact support software (.., Zendesk) • Logistics (FedEx) products services 22 geography innovation NDE earlier rounds globalization, counter-trends fragmentation spatial dispersion GVCs, driving continued growth technology clusters Silicon Valley, systems underpin NDE developed, standard battles fought won (Sturgeon, 2003). headquarters locations important players NDE, shown Figure 8, reveals extreme level concentration North America, headquarters 63 135 NDE companies market capitalization $1 billion 2015. closer examination North American companies Figure 8 suggests greater level -national concentration. headquarters , exception, located handful postal codes Silicon Valley, California Seattle, Washington. situation, advent AI-assisted design processes mentioned, impact NDE location innovation -fold. hand, iterative engineering work fall expertise required design products, generate additional demand skilled labour core platform owners, highly skilled engineers workers needed firms systems produce higher- level platforms final products services. expertise required design, test validate product designs embedded software platforms, AI portion. making successful systems requires expertise, job counts spill-overs local innovation clusters reduced. hand, downstream innovation ( innovation core platforms, platforms products created ) lead uptick innovation heartland core NDE innovation, products tailored local markets developed produced easily, quickly inexpensively. Figure 8. Headquarters region digital economy companies US$1B market cap 2015 Source: Van Alstyne (2016), based research Peter Evans, KPMG. remains geographic concentration suggested Figure 8 result youth NDE, companies spawned existing dominant clusters computing, software, networking; clusters draw innovative segments traditional industries adapt NDE. NDE, broadly defined, includes existing technology clusters (.. Pittsburgh, Boston, Toronto, Munich Tokyo), 23 winner-- game platform competition, winner-- geographic analogue, driving spatial inequality domestically internationally. , safe assume opportunities NDE firms places established technology clusters significant. General Motor’ move concentrate AI-based -driving technology development San Francisco Bay Area ( Press, 2017), centripetal spatial dynamic mentioned, examples centrifugal dynamic abound. include Uber’ (increasingly troubled) testing driverless vehicles Philadelphia (Kang, 2017), apparent emergence driverless vehicle cluster Boston led start-ups nuTonomy (Vaccaro, 2017), proliferation fully electric vehicle producers China (Helveston al, 2017). important note big Internet companies China, Baidu (Internet search), Alibaba (-commerce), Tencent (social networking mobile services), referred BAT companies, dominant China’ huge domestic market. Tencent’ WeChat social networking platform 800 million users, China.28 lens shifted core platforms higher-level platforms, platform complementors, platform users, possibilities open significantly. resources listed Figure 6 prove hint vast growing set digital resources start-ups, smaller, globally remote companies innovate, grow, improve operations, connect markets. tools cover gambit, including aids & innovation, capital, labour, operations, inputs, marketing market access. regard capital, Allied Crowds website aggregates hundreds alternative financing providers specific focus developing countries, including equity (angel venture capital) investors, crowd funding, donors assist firms ( allowed local regulations) bypassing local banking systems overcome financing hurdles, offer companies (.. marketing). tools create possibilities, exist powerful technology firms core platform owners concentrated places Silicon Valley Seattle, . Linkages manufacturing innovation NDE crucial question tightly innovation geographically tied production NDE. Historically speaking, industrial, innovation, manufacturing policies based assumption strong spatial linkages. Innovation policies hope spawn industries turn generate large-scale employment, including manufacturing. Investment attraction policies hope create manufacturing employment, instance, embody hopes knowledge-intensive activities eventually follow production spill-. -defined, modular break-point innovation production ( depicted Figure 4), successful innovation require manufacturing -located, products require processes produce . , important industries electronics, software, motor vehicles, effective modular linkages innovation 28 relevant firms developed “Chinese Firewall,” term Chinese government’ extensive attempts block content deems subversive Chinese Internet, including refusal grant operating licenses Google Facebook, knowledge Chinese market important, Tencent/WeChat, developed mobile platform includes innovations appeal specifically Chinese users (.. “red envelope” feature WePay easy gifting money friends family line traditional norms), Alibaba, essentially outcompeted Amazon China based product mix fulfillment (Thun Sturgeon, 2017). 24 production stages chain facilitate separation innovation production geographically separated distinct technology production clusters ( depicted Figure 9). Bonvillian (2012) frames shift, industries high levels chain modularity, “innovate /produce ” structure “innovate /produce ” structure. dynamic industries change terms products, process technologies, business models, regulatory requirements, pressures -location increase industries heretofore evolving lines innovate /produce . question , part industry “move” close perspective advanced economies United States, Western Europe Japan, framed replacement “hollowing ” trend “-shoring”, manufacturing returns heartland innovation, productivity increases production (.. result automation) rendered labour cost differentials unimportant, manufacturing tightly linked innovation technical coordination reasons (referred “decodification” Gereffi al, 2005), set reasons, industrial trade policies shift incentives domestic manufacturing. , emerging economies, China, Brazil, Viet Nam long incentivizing investments & combination manufacturing (Zylberberg, 2017). countries seek leverage requirements -location innovation production ways generate substantial & spill-overs borders. Excitement -shoring waxed waned repeatedly, real impact direction change global economy innovate /produce model underpinning GVCs. industry moved manufacturing offshore, , “ ” (.. Kearney, 2014). evidence suggests variation industries (Sturgeon Memedovic, 2010), firms strategic choices regard business models (Berger al, 2005), reasonable assume maturing NDE lead loosening spatial ties business functions, including innovation ( innovation), production, logistics, marketing, distribution -sales service. emergent features NDE poised extend organizational geographical fragmentation work realms, including indivisible geographically rooted activities reside front GVCs, &, product design, knowledge- intensive innovation-related business functions. digitization work, products services, means industries underpinned ICT, connected cloud, based global platforms. loosens constraints -location, firms, large small, ways pursue “optimization” strategies regard location markets distribution, hand, location sourcing business functions intermediate inputs . , regulatory environment change locational incentives dramatically. Scenarios developing countries NDE scenarios outlined, impacts NDE developing countries features NDE add radical change, line earlier advances computerization, beginning 1980s, accelerating 1990s, mainstream 2000s, allowed organizational geographic separation & design 25 manufacturing, leading creation GVCs. Higher business functions, branding product design tended stay established technology clusters, vast investments manufacturing cost market-proximate locations China, Viet Nam, South Africa, Brazil Mexico, creating large numbers production jobs jobs adjacent categories materials supply chain management, manufacturing engineering, logistics distribution (Berger, 2005; Baldwin, 2016; IBRD/World Bank, 2017). , countries regions deeply connected GVCs, China export-oriented economies East Asia, India, Eastern Europe, North Africa, Latin America (.. Mexico) fall “ added traps.” greater share ( profits) tend accrue “lead” firms GVCs control branding, product conception, retail distribution; suppliers advanced production equipment technology platforms intellectual property owners provide key inputs de facto standards chain (.. Intel Qualcomm CPUs). growing number important MNEs successful suppliers higher level platforms developing world, GVCs dominated firms traditional technology clusters industrialized countries (Blyde, 2014). Firms provide routine assembly tasks simple services GVCs tend profits, pay workers , vulnerable business cycles. idea articulated ( direct experience) Acer CEO Stan Shih ‘Smiling Curve’ added ( Figure 9), demonstrated research Linden al (2009) case Apple iPod. 2011 study OECD (2011), based part work estimated China’ added $600 iPhone 4 ( assembly packaging) US $6.54, 1% retail price. raises series questions. NDE open opportunities developing countries, deepen existing geographic divides automation production migrate closer point consumption decreasing importance large-scale direct labour engine growth developers , shifts drive developers quickly higher activities, , AI entrepreneurs developing countries platforms develop sophisticated products suitable export advanced markets ICT-enabled fragmentation knowledge work lead waves global investment &, deepening broadening formation GVCs manufacturing, broad scenarios developing countries NDE. routine business functions manufacturing, software coding, office services, served backbone rapid development, -shored eliminated advanced manufacturing automation. drive retreat GVCs huge social disruptions developers, export oriented factories employ tens, hundreds thousands workers concentrated manufacturing clusters ( Shenzhen, Hanoi, Guadalajara, ). scenario, tools NDE empower developing country firms move chain, dependent innovation coordination functions lead firms GVCs, produce globally competitive compatible products . Rehnberg Ponte (2017) depict scenario, 3D printing, flattening chain curve “smiling smirking.” scenario innovate /produce geographic division labour suggested Figure 9 remains stable NDE alters products processes existing technology production clusters. balance centripetal centrifugal effects geography 26 industries, alterations complexity fluidity GVCs patterns today, doubtless constitute core research questions scholars GVCs coming decades. Figure 9. “smiling” curve added, $600 iPhone 4 Cost differentials share added routine business functions, shifted developing countries. Functions geographically segmented knowledge clusters, higher -added functions, production clusters, pool -added functions, creating -added traps (China, exporter record, contributes 1.1% $600 iPhone’ ) Source: . Sturgeon, UNCTAD, iPhone drawn OECD, 2011. . Concluding remarks ’ NDE main driver NDE continued exponential improvement cost performance ICTs, microelectronics, Moore’ Law. . digitization design, advanced manufacturing robotics, communications, distributed computer networking (.. Internet) altering processes innovation possibilities relocation work decades. , trends NDE . , sources data, smart phones factory sensors, resulting accumulation vast quantities data “cloud,” creating information pools create insights, products, services — risks society. , business models based technology product platforms — platform innovation, platform ownership, platform complimenting — , range industries product areas, radically altering industry structure terms competition. , quantitative advancement semiconductor technology Moore’ Law , areas, graphics processing, advanced point qualitative begun occur practical applications machine learning- based AI. trends share reliance advanced increasingly ubiquitous ICT. 27 range technologies, techniques, scientific avenues opened , part, tools NDE numerous mention. addition applications referred study, space exploration, cost targeted gene editing techniques (.. CRISPR), ultra-small scale manipulation matter, referred nano-technology, added.29 Winners losers, opportunities risks transformational, NDE create winners losers, opportunities risks. positive, utopian vision NDE centre ubiquity democratization information — hard envision twenty years introduction Google search ten years smart phone era — decoupling economic growth natural resource constraints enabled part shortening supply chains advent - demand manufacturing (.. 3D printing) super-efficient containerized urban agriculture (Chambers Elfrink, 2014). NDE, , usher newly equitable environmentally sustainable growth model based maximization human empowerment wellbeing maximization profits resource extraction utilization (Erkoskun, 2011). Personal robots certainty helpful infirm disabled, flexible integrated everyday life (Rus, 2015). , legitimate worries NDE introduce frightening risks, prosper evolution. workers, large sudden productivity increases enabled NDE shorten employment adjustment period softened impact earlier rounds automation. penetration computerization AI knowledge-intensive services jobs risk disappearing, output productivity rise (McAfee Brynjolfsson, 2016). hand, advanced economies remarkable ability create industries, demand skills, create jobs (Autor ,2015). , disruption automation globalization experienced unevenly, , assumed class super intelligent dexterous robots direct labour factories fall coming decades (McAfee, Brynjolfsson, Spence, 2014), driving “employment polarization level industries, localities, national labour markets” (Autor, 2015, . 12), places highly dependent manufacturing, temporarily. , economic social effects NDE expected broader job loss factory automation. Ride sharing revolutionizing individual mobility autonomous road vehicles, freight trucks, knocking door mainstream deployment (Vincent, 2016). Services desks education training payments banking increasingly delivered automated -desk systems include voice recognition AI features. “gig economy” (De Stafano, 2015) creating precarious class “ demand” workers, “dependent contractors” (Smith Leberstein, 2015), including knowledge workers, 29 Nano-technology includes nano-manufacturing creation materials emergence class nano-machinery. Nano-technology impacting range fields, fabrication semiconductors, creating virtuous technological circle processes innovation tools engage innovation. 28 part broadly emergent “precariat” clear institutional means organizing (Standing, 2016). ride sharing apartment rental platforms tend receive lot press attention, platforms connect home care workers clients (.. Care.), connect clients “ demand” workforce platforms,30 involve larger numbers workers, order millions platform, hundreds thousands working ride sharing platforms (Smith Leberstein, 2015). , jobs suitable -demand work, copyediting, data cleaning, moderation online content, transcription, driving, vulnerable replacement AI computer systems. ‘’ centers offshore IT services, Bangalore, hiring significant layoffs time (Narayan, 2017).31 intellectually stimulating satisfying jobs created, rising inequality, potential worker abuse, downward pressure wages worrying aspects NDE. consumers, risks real potential job losses automation AI. , big data AI enable “ degree” price discrimination, prices adjusted constantly real time based consumer' perceived product service willingness pay. variables estimated prior shopping purchasing histories analysed context millions prior purchases shoppers similar habits, consumer bargaining power harmed (Shiller, 2014). hand, automation, mass customization shorter supply lines prices vastly improve consumer satisfaction (Bhasin Bodla, 2014). actual potential benefits connected IoT devices home proven, smart phone apps, free, easy-- map navigation music streaming services proven worth users. Data flows platform owners underpin, theoretically, - enhancements upgrades digital products services. price services users passively, unknowingly, provide app-makers platform owners fine-grained information whereabouts personal habits. , Facebook collects user data includes city, gender, age IP address, , full record website users link tag content ( logged ). addition, company combines information assumptions users harvested online activity information public sources data brokers assemble dossiers users 100 variables (job title, parents’ birthdays, .) target advertisements precisely (Dewey, 2016). large companies, organizations, governments, vulnerability — hacking, identity theft, espionage, larceny, ransoming, industrial sabotage — connecting private communications, industrial systems, public infrastructure Internet (Hampson Jardine, 2016). result, companies deeply engaged advanced manufacturing connections premises factories fear data breaches, obviates advantages data sharing 30 include Crowdflower, Crowdsource, Clickworker, Mechanical Turk, MTurk, crowdsourcing site operated Amazon Web Services connects researchers individuals scientific experiments tedious data analysis tasks exchange small payments. 31 , late 2016 Larsen & Toubro, India’ biggest engineering firm, reduced workforce 14,000 employees, 11.2%, 2017 layoffs IT services firms including Cognizant, Wipro, Infosys Tech Mahindra. 29 pooling larger organization supply base. Ignoring risks grave consequences, undermine promise era. smaller companies, cost expertise required purchase, operate, continually upgrade advanced manufacturing IT systems drive larger wedge large — multinational — firms scale justify needed investments, smaller, locally-oriented developing country firms. winner-- dynamics platform-based industries (.. Google, Uber, Facebook, WeChat), network effect advantages accrue -movers standard setters (Parker al, 2016), lead accentuated polarization industrial base standard-bearers NDE consolidate gains. concerns, NDE holds promise businesses advantage technology mitigate risks. Large small companies rely tools NDE, rich poor countries alike, organizations efficient, reach serve customers effectively, speed product development, invent products services deep pockets deep system-level expertise. core platform owners access comprehensive data higher-level platform owners users, access world’ relevant data required speed innovation carve market space NDE. intriguing prospect, developing country perspective, small firms entrepreneurial start-ups, , access crowd funding build products based technology, core, higher-level platforms. AI-assisted tools built design software, analytics included feature platform analytical tool refine subsequent design iterations, innovation efficient effective opportunities growth multiply. tools lowering cost entry NDE, easy potential benefit economic development entrepreneurs engineers capabilities . Data partitioning, digital divide promise, access data remain partitioned NDE. stylized depiction data flows NDE, shown figure 10, suggests sweet spot data access resides core platform owners, secondarily higher-level platforms. Entrepreneurs small companies access data, analyse making AI tools platforms, access larger insights larger pools data cost purview platform owners. , “digital divide” increasingly describe difference connected digital world remain disconnected, “digital skills” , widening inequality groups places connected. people places connected NDE, benefit , levers control extraction profits lie hands . advantages NDE average users, greater advantages accrue capability authority accumulate, access analyse big data. 30 Figure 10. Actors, data access, data flows NDE ecosystem Source: . Sturgeon, UNCTAD appears beginning disruptive technological wave. prior technological disruptions, steam engines electric power digital computing, logic efficiency run capacities organizations society large absorb adapt , requiring significant reshaping accommodation order reach mature humane footing (Bodrozic Adler, 2017). full impact NDE jobs, international competition, location production unknown, outcomes crucially depend pace change ability organizations societies — including regulators producers economic statistics — understand , measure , manage . users Higher level platforms Core platforms Technology platforms • capture higher level platform user data • capture customer data • analyze data • capture customer data Downstream Upstream nriched products & services ain data flow 31 References Ali-Yrkkö, ., . Rouvinen, ., Seppää . Ylä-Anttila, . 2010. “ Captures Global Supply Chains Case Nokia N95 Smart Phone.” Journal Industry, Competition Trade, 11(3), Sept., pp. 263–278. Web access: https://link.springer./article/10.1007/s10842-011- 0107-4 Amelinckx, . 2016. “Rise ( Milking) Robots.” Modern Farmer. December 12. Web access: http://modernfarmer./2016/12/rise--milking-robots/ Press. 2017. “GM hire 1,100 workers California bolster -driving car program; $8 million tax break.” Reported Los Angeles Times, April 17. Web access: http://www.latimes./business/la-fi-hy-gm-jobs-20170414-story.html Astor, . 2017. “ Roomba Mapping Home, Collecting Data Shared.” York Times, Technology, July 25. Web access: https://www.nytimes./2017/07/25/technology/roomba-irobot-data-privacy.html .. Kearney. 2014. “ hype reshoring rise, 2014 increased offshoring.” PR Newswire, Dec 15. Web access: http://www.prnewswire./news- releases/2014--kearney-reshoring-index--20-basis-points-year--year--2013- uncovers--manufacturers----300009465.html Autor, . 2015. “ Jobs History Future Workplace Automation.” Journal Economic Perspectives, 29: 3, Summer, pp. 3-30. Baldwin, . Clark, .. 2000. Design Rules; Unleashing Power Modularity. Cambridge, MA: MIT Press. Barbier, . 2017. “ Trends Manufacturing’ Fourth Wave.” Intelligence 3:1, Flex . Web access: https://flex./intelligence/manufacturing/-trends-manufacturing-fourth-wave Bhasin, . Bodla, .. 2014. “Impact 3D printing global supply chains 2020.” Engineering Systems Division Thesis (Advisor: Shardul Phadnis). Massachusetts Institute Technology. Web access: http://hdl.handle.net/1721.1/92106 Berger, . MIT Industrial Performance Center. 2005. Compete. York: Doubleday. Bershidsky . 2017. “Google Glass Time; Silicon Valley recognize innovation masses.” Bloomberg View, Opinion | Tech, July 19. Web access: https://www.bloomberg./view/articles/2017-07-19/google-glass--- ---time Blyde, . 2014. “Synchronised factories: Latin America Caribbean era global chains.” VoxEU. Web access: http://voxeu.org/article/latin-americas-missing-global--chains Bodrozic, . Adler, . 2017. “ Evolution Management Models: Neo-Schumpeterian Theory.” press, Administrative Science Quarterly, pp.1–45 Bonvillian, . 2012. “Reinventing American Manufacturing: Role Innovation.” Innovations, 7:3, pp. 97-125. Web access: http://www.mitpressjournals.org/doi/pdf/10.1162/INOV_a_00142 Bort, . 2015. “ Facebook eating $140 billion hardware market.” Business Insider, Jun. 14. Web access: http://www.businessinsider./facebook-open-compute-project-history-2015-6 32 Bonvillian, . 2017. “Advanced Manufacturing: Policy Challenge”, Annals Science Technology Policy: Vol. 1, . 1, pp. 1–131. Web access: http://www.nowpublishers./article/Details/ASTP-001 Boudreau, . (2010). Open platform strategies innovation: Granting access . devolving control. Management Science, 56(10), 1849-1872. Brodkin, . 2016. “Netflix finishes massive migration Amazon cloud.” ARS Technica, Biz IT, February 11. Web access: https://arstechnica./information- technology/2016/02/netflix-finishes--massive-migration---amazon-cloud/ Brynjolfsson, . 2016. “ IoT decision making, security public policy.” Blog: Research & Commentary MIT Sloan Business & Management Experts, June 30. http://mitsloanexperts.mit./-iot--decision-making-security--public-policy/ Cadwalladr, . 2016. “Google, democracy truth Internet search.” Guardian, Internet, Observer, Dec. 4, https://www.theguardian./technology/2016/dec/04/google-democracy-truth-Internet- search-facebook Chambers, Elfrink, . 2014. “ Future Cities: Internet Change Live.” Rose, . (ed.) Fourth Industrial Revolution, Foreign Affairs, pp. 129- 138. Web access: https://www.foreignaffairs./articles/2015-12-12/fourth-industrial- revolution Chandler, . 1962. Strategy Structure. Cambridge: MIT Press. Chesbrough, ., Vanhaverbeke, . West, . (eds.). 2006. Open Innovation: Researching Paradigm. Oxford: Oxford University Press. Chesbrough, ., Kenichi . 2001. “ Modularity Trap: Innovation, Technology Phase Shifts, Resulting Limits Virtual Organizations.” Managing Industrial Knowledge, ed. . Nonaka . Teece, 202–30. London: Sage Press. Dallas, ., Ponte, ., Sturgeon, . 2017. “ Typology Power Global Chains.” Working Paper Business Politics . 91, Copenhagen Business School. Web access: http://openarchive.cbs.dk/bitstream/handle/10398/9503/Dallas_Ponte_Sturgeon.pdfsequence =3 Dalten, . 2017. “IBM Indiegogo bringing Watson' smarts masses; partnership entrepreneurs unlimited access IBM' AI.” Engadget, Febrary 16, https://www.engadget./2017/02/16/ibm-indiegogo-watson-iot-partnership/ De Backer, . 2011. “Global Chains: Preliminary Evidence Policy Issues” OECD Directorate Science, Technology Industry. Paper presented Committee Industry. Web access: https://unstats..org/unsd/trade/globalforum/publications/gvc/%20- %20OECD%20-%202011%20-%20GVCs%20-%20Preliminary%20Evidence%20- %20Policy%20Issues_March%204.pdf Deboeck, . (ed.) 1993. Trading Edge: Neural, Genetic, Fuzzy Systems Chaotic Financial Markets. York: Wiley. Denning, . 2013. “ Wrong Boeing” Forbes, Jan 21. Web access: https://www.forbes./sites/stevedenning/2013/01/21/--wrong-- boeing/#5bbb0be27b1b 33 De Stefano, . 2015. “ Rise ‘--Time Workforce’: -demand Work, Crowd Work, Labour Protection ‘Gig economy’. Bocconi University Legal Studies Research Paper . 2682602. Web access: https://papers.ssrn./sol3/papers.cfmabstract_id=2682602 Dewey, . “98 personal data points Facebook target ads .” Washington Post, Intersect, August 19. Dossani, . Kenney, . 2003. “Lift Shift; Moving office India.” Information Technologies International Development, 1(2): 21–37. Massachusetts Institute Technology, Cambridge, MA. Economic Times. 2017. “Layoffs scare real, exaggerated, finds ET' Jobs Disruption survey.” Economic Times, June 1. Web access: http://economictimes.indiatimes./jobs/layoffs- scare--real--exaggerated-finds-ets-jobs-disruption-survey/articleshow/58927915.cms Eisenmann, ., Parker, ., & Van Alstyne, . . (2006). Strategies -sided markets. Harvard Business Review, 84(10), 92. Epstein, . 2017. “ blockchain meets big data, payoff huge.” VentureBeat, July 30. Web access: https://venturebeat./2017/07/30/-blockchain-meets-big-data--payoff- --huge Ercoskun, . . (ed.) 2011. Green Ecological Technologies Urban Planning: Creating Smart Cities. IGI-Global. Web access: https://www.igi-global./book/green-ecological- technologies-urban-planning/55271 Ezrchi, . Stucke, . 2016. Virtual Competition: Promise Perils Algorithm-Driven Economy. Cambridge, MA: Harvard University Press. Finley, . 2016. “Linux Web. , ' World.” Wired Magazine, Business, August 25. Web access: https://www.wired./2016/08/linux--web-- -world/ Friedman, . (2014). Workers employers: shadow corporations rise gig economy. Review Keynesian Economics, (2), 171-188. Gawer & . . Cusumano. (2014). Industry platforms ecosystem innovation. Journal Product Innovation Management 31(3), 417-433. Gawer . . Cusumano. (2002). Platform Leadership: Intel, Microsoft, Cisco drive industry innovation. Boston: Harvard Business School Press. Gereffi, .; Humphrey, . Sturgeon, . 2005. “ Governance Global Chains.” Review International Political Economy. 12(1): 78-104. Gary, . Sturgeon, . 2016. “Global Chains Industrial Policy: Role Emerging Economies.” Chapter 14 : Elms, . , . (eds.), Global Chains Changing World. pp. 329-360. WTO Publications. Web access: http://www.wto.org/english/res_e/booksp_e/aid4tradeglobalvalue13_e.pdf. Greenough, . Camhi, . 2016. “' calling Internet Industrial Revolution.” Business Insider, Tech Insider, Feb. 10. http://www.businessinsider./iot-trends--shape----interact-2016-1 34 Hagiu, ., & Wright, . (2015). Multi-sided platforms. International Journal Industrial Organization, 43, 162-174. Hampson FO Jardine (2016). ' Watching, Centre International Governance Innovation, Waterloo, ON Canada. Helveston, JP, Wang, . Karplus, . Fuchs, . 2017. “Innovating , , Sideways: () Institutional Origins Experimentation China' Plug- Electric Vehicle Industry.” Manuscript, February 21. SSRN: https://ssrn./abstract=2817052 http://dx.doi.org/10.2139/ssrn.2817052 Hollinger, . 2016. “Meet cobots: humans robots factory floor.” Financial Times, 5. Web access: https://www.ft./content/6d5d609e-02e2-11e6-af1d- c47326021344 Innovation Entrepreneurship, CIIE meeting, 31 March-1 April 2011. International Bank Reconstruction Development/ World Bank. 2017. Global Chain Development Report 2017; Measuring Analyzing Impact GVCs Economic Development. Web access: https://www.brookings./wp-content/uploads/2017/07/tcgp-17- 01-china-gvcs-complete--web-0707.pdf Johnson, . 2017. “ Public Clouds Amazon, Microsoft Google, Big Data Proverbial Big Deal.” Forbes, Jun 15. Web access: https://www.forbes./sites/johnsonpierr/2017/06/15/--public-clouds--amazon- microsoft--google-big-data---proverbial-big-deal/#2a37a76b2ac3 Kang, . 2017. “Pittsburgh Welcomed Uber’ Driverless Car Experiment. Anymore.” York Times. Technology, 21. Web access: https://www.nytimes./2017/05/21/technology/pittsburgh-ubers-driverless-car- experiment.html Kenney, ., & Zysman, . (2016). “ Rise Platform Economy.” Issues Science Technology, 32(3), 61-69. Kerns, . 2017. “3D Printing Dives Mass Production.” Machine Design, Technology, 3D Printing, July 1. Web access: http://www.machinedesign./3d-printing/3d-printing-dives- mass-production Levin, . 2017. “Model 3 Rollout Prove Tesla Manufacture High Rate Speed” Street. https://www.thestreet./story/14013140/1/model-3-rollout--prove- -tesla--manufacture---high-rate--speed.html Levy, . 2017. “Microsoft' cloud business growing fast Amazon', Google .” CNBC Tech. April 27, http://www.cnbc./2017/04/27/microsoft-azure- growing-faster--aws-google-cloud-.html Linden, .; Kraemer, ., Dedrick, . 2009. “ Captures Global Innovation System case Apple' iPod,” Communications ACM. 52(3): pp.140-144. Mar, . 2017. “Supervised Unsupervised Machine Learning -- ' Difference” Forbes, Tech # BigData, March 16. Web access: https://www.forbes./sites/bernardmarr/2017/03/16/supervised--unsupervised-machine- learning-whats--difference/#6464ce81485d 35 Markowsky, . 2012. “ Virtual Factory: American Manufacturing 21st Century.” ILP Institute Insider, 10. Web access: http://ilp.mit./newsstory.jspid=18006 McAfee, . Brynjolfsson, , Spence, . 2014. “ World Order: Labor, Capital, Ideas Power Law Economy.” : Rose, (ed.) Fourth Industrial Revolution: Davos Reader. Council Foreign Relations. McAfee, , Brynjolfsson, . 2016. "Human Work Robotic Future: Policy Age Automation." Foreign Affairs, July/August. McAfee, . Brynjolfsson, . 2017. Machine, Platform, Crowd: Harnessing Digital Future. York: . . Norton & Company. Miller, . 2016. “ Rackspace Open Compute Cloud.” Data Center Frontier. March 3. Web Access: https://datacenterfrontier./rackspace-open-compute-cloud/ Narayan, . 2017. “Decoding Disruption: Lay-offs Software Industry.” Economic & Political Weekly, LII: 16 22, June 3. Novet, . 2016. “Google joins Open Compute Project 5 years, submit rack design.” Venture Beat, March 9. Web access: https://venturebeat./2016/03/09/google-joins-- open-compute-project--5-years--submit-rack-design/ NZ Herald. 2016. “PwC Herald Talks: Data decision making.” Zealand Herald, Business, July 19. Web access: http://www.nzherald..nz/business/news/article.cfmc_id=3&objectid=11677284 Paiste, . 2014. “Faculty highlight: . John Hart, Mechanical engineering professor explores science technology nano manufacturing.” MIT News, February 5. Web access: http://news.mit./2014/faculty-highlight--john-hart Parker, .; Van Alstyne, . Choudary, . 2016. Platform Revolution: Networked Markets Transforming Economy-- Work . York: . . Norton & Company. Pine, . . Davis, . 1999. Mass Customization, Frontier Business Competition. Cambridge, MA: Harvard Business School Press. Piore, . Sabel, . 1984. Industrial Divide. York: Basic Books. Redmon, ., Divvala, ., Girshick, ., Farhadi, . 2016. “ : Unified, Real- Time Object Detection.” IEEE Conference Computer Vision Pattern Recognition. (CVPR), 2016, pp. 779-788 Rehnberg, . Ponte, . 2017. “ Smiling Smirking 3D Printing, Upgrading Restructuring Global Chains.” Global Networks, Early View, June 20. Reynolds, . 2017. “ Face Manufacturing Jobs.” Boston Globe, Opinion, February 13. Web access: https://www.bostonglobe./opinion/2017/02/13/--face-manufacturing- jobs/jsozWRgDPwiXo5yJwA3z2M/story.html Rose, . (ed.). 2016. Fourth Industrial Revolution: Davos Reader. Council Foreign Relations. 36 Rus, . 2015. “ Robots Coming: Technological Breakthroughs Transform Everyday Life.” : Rose, (ed.) Fourth Industrial Revolution: Davos Reader. Council Foreign Relations. Sarma, . 2015. “Trends Advanced Manufacturing Technology Research.” Chapter 6 : Berger, , (ed.) Making America: Innovation Market. Cambridge, MA: MIT Press. Schor, . (2014). Debating sharing economy. Great transition initiative. Schwab, . 2015. “ Fourth Industrial Revolution Means Respond.” Foreign Affairs, Science & Technology, December 12. https://www.foreignaffairs./articles/2015- 12-12/fourth-industrial-revolution Schwab, . 2016. Fourth Industrial Revolution. York: Crown Business. Scott, . ., & Orlikowski, . . (2012). Reconfiguring relations accountability: Materialization social media travel sector. Accounting, organizations society, 37(1), 26-40. Shiller, . . 2014. “-Degree Price Discrimination Big Data.” April 25, Brandeis University, Department Economics Working Paper 58. http://www.brandeis./departments/economics/RePEc/brd/doc/Brandeis_WP58R.pdf Smith, . Leberstein, . 2015. Rights Demand: Ensuring Workplace Standards Worker Security -Demand Economy. National Employment Law Project. Web access: http://www.nelp.org/content/uploads/Rights--Demand-Report.pdf Standing, . 2016. Precariat: Dangerous Class. London York: Bloomsbury Academic. Standing, ., 2012. precariat: denizens citizens Polity, 44(4), 588–608. Sturgeon, ., 2002. “Modular Production Networks. American Model Industrial Organization,” Industrial Corporate Change. 11(3):451-496. Sturgeon, . 2003. “ Silicon Valley Spatial Clustering Dispersal Modular Production Networks.” Journal Economic Geography. April, .3. Sturgeon, . Memedovic, . 2010. “Mapping Global Chains: Intermediate Goods Trade Structural Change World Economy.” UNIDO Working Paper 05/2010. United National Industrial Development Organization, Vienna. Web access: http://www.unido.org/fileadmin/media/documents/pdf/Publications/110905/WP052010_Eboo .pdf Thun, . Sturgeon, . 2017. “ Global Technology Meets Local Standards: Reassessing China’ Mobile Telecom Policy Age Platform Innovation.” Forthcoming : Impact Industrial Policy Regulation Upgrading Innovation Chinese Industry, . Brandt . Rawski (eds.). MIT Industrial Performance Center Working Paper 17-001. Web access: https://ipc.mit./sites/default/files/documents/17-001.pdf United Nations Conference Trade Development (UNCTAD). 2013. Information Economy Report 2013: Cloud Economy Developing Countries;. UNCTAD Division Technology Logistics, ICT Analysis Section. York Geneva. Web access: http://unctad.org/en/PublicationsLibrary/ier2013_en.pdf 37 UNCTAD. 2017a. Information Economy Report 2017: Digitalization, Trade Development. UNCTAD Division Technology Logistics, ICT Analysis Section. York Geneva. http://unctad.org/ier. UNCTAD. 2017b. Role Science, Technology Innovation Ensuring Food Security 2030. . UNCTAD/DTL/STICT/2017/5. United Nations: York Geneva. Vaccaro, . 2017. “ ready, -driving cars coming Boston roads.” Boston Globe, April 25. Web access: https://www.bostonglobe./business/2017/04/25/-ready-- driving-cars--coming--boston-roads/t9yBFEUJvP3HTAcEHtglTM/story.html. Van Alstyne, .; Parker, .; Choudary, . 2016. “Pipelines, Platforms, Rules Strategy.” Harvard Business Review, April. Web access: https://hbr.org/2016/04/pipelines- platforms----rules--strategy Van Alstyne. . 2016. “Platform Shift: Biz Models Changing Shape Industry.” Presentation ICT Industry Roundtable, MIT Industrial Performance Center Accelerating Innovation Brazil Project, Sao Paulo, Brazil, 10. Vincent, . 2016. “-driving truck convoy completes major journey Europe: EU- backed challenge sets stage automation Europe' roads.” Verge. April 7. Web access: https://www.theverge./2016/4/7/11383392/-driving-truck-platooning- europe Von Hippel, . 2017. Free Innovation. Cambridge, MA: MIT Press. Whittaker, .., Zhu, ., Sturgeon, ., Okita, ., Tsai, .. 2010. “Compressed Development.” Studies Comparative International Development, 45:3, pp.439-67. Whittaker, .., Sturgeon, ., Okita, ., Zhu, . Forthcoming. Compressed Development. Oxford University Press. Womack, ., Jones ., Roos, . 1990. Machine Changed World. York: Rawson Associates. World Bank. 2016. World Development Report 2016: Digital Dividends. Washington, DC: World Bank. doi:10.1596/978-1-4648-0671-1 Zylberberg, . 2017. Industrial policy refraction: corporate strategy shapes development outcomes Brazil. PhD Thesis. Oxford DPhil Programme Management Studies, Business School. Web access: https://ora.ox.ac.uk/objects/uuid:9fa15fe3-5ffd-4f2e-a82c- ba4671430aec UN CTA UNITED NATIONS UNCTAD Technical Notes ICT Development 1. Implications Applying Definition «ICT Goods», 2011 2. Updating Partnership Definition ICT Goods HS 2007 HS 2012, January 2014 3. International Trade ICT Services ICT-enabled Services: Proposed Indicators Partnership Measuring ICT Development, October 2015 4. Global Assessment Sex-disaggregated ICT Employment Statistics: Data Availability Challenges Measurement Compilation, December 2015 5. Trade ICT Goods 2015 Expansion WTO Information Technology Agreement, December 2015 6. Search Cross-border -commerce Trade Data, April 2016 7. UNCTAD B2C -commerce Index 2016, April 2016 8. «» Digital Economy Development, October 2017 9. UNCTAD B2C -commerce Index 2017, October 2017 information UNCTAD› work ICT Development contact: ICT Analysis Section Division Technology Logistics www.unctad.org/ict4d ict4d@unctad.org Cover_TN_ICT4D_08_NDE TN_ICT4D_08_The Digital Economy Development final version Oct3_2017.clean.cleanTSformt cover template TN_ICT4D9Index_i