Analysis on Research Level of the Five Major Platform Technologies Related to the Fourth Industrial Revolution

Transcription

1 NRF ISSN ISSUE REPORT 2017_No.4 Analysis on Research Level of the Five Major Platform Technologies Related to the Fourth Industrial Revolution I. Understanding the Fourth Industrial Revolution II. Analyzed Technology Derivation and Technology Summaries III. Research Paper Performance Analysis Method IV. Research Paper Performance Analysis Result V. Status of NRF Research Project Support VI. Implications

2 NRF ISSN ISSUE REPORT 2017_No.4 Analysis on Research Level of the Five Major Platform Technologies Related to the Fourth Industrial Revolution I. Understanding the Fourth Industrial Revolution II. Analyzed Technology Derivation and Technology Summaries III. Research Paper Performance Analysis Method IV. Research Paper Performance Analysis Result V. Status of NRF Research Project Support VI. Implications

3 Contents I. Understanding the Fourth Industrial Revolution... 3 II. Analyzed Technology Derivation and Technology Summaries... 5 III. Research Paper Performance Analysis Method...17 IV. Research Paper Performance Analysis Result...20 V. Status of NRF Research Project Support...48 VI. Implications...51 <Reference>...53

4 Ι. Understanding the Fourth Industrial Revolution As human civilization has advanced, there have been several turning points in its history so called the Agricultural Revolution, Scientific Revolution and Industrial Revolution, in general. The Scientific Revolution, started in the 15th century and led to the First Industrial Revolution at the end of the 18th century. It changed human kinds economic and social structure significantly, as three industrial revolutions have been carried out so far. The Industrial Revolution fundamentally changed the form of human civilization, starting with the change of power and innovating the industrial and economic system, generally categorized as follows. The First Industrial Revolution was the change of power using steam (people, animals machinery). The Second Industrial Revolution brought the mass production system using electricity and fossil fuels. The Third Industrial Revolution was the emergence of the digital age based on electronic engineering and ICT. The Fourth Industrial Revolution was mentioned at the Davos Forum in 2016 for the first time. It is expected to be a cyber-physical system (CPS) revolution based on the bio industry, 3D printers, robots, and artificial intelligence, but still there are many arguments around it. Davos Forum The Davos Forum was established in 1971, under the name of the 'European Management Forum'. It expanded the theme from Europe to the entire world in 1973, and the meeting place was fixed in Davos, Switzerland in It was renamed the World Economic Forum in 1987, but as held in Davos, it is commonly called the Davos Forum. Comparison of the Four Revolutions The First Revolution The Second Revolution The Third Revolution The Fourth Revolution Period 18 th century From the 19 th to the early 20 th century Later 20 th century After 2015 Motto Mechanization revolution based on the steam engine Mass production revolution based on electric energy Computer and Internet-based knowledge information revolution Intelligent revolution based on artificial intelligence Contents British textiles became a huge industry by using the steam engine Mass production using conveyor belts with power supplied to factories Emergence of global IT companies through the Internet and smart revolution Innovation of industrial structure social system by hyper-connection and hyperintellectualization of people, objects, and space Ⅰ. Understanding on the Fourth Industrial Revolution 3

5 The ultimate goal of the Fourth Industrial Revolution is automation, meaning that all work that used to be done by people will be done without people. Technologies enabling the automation of existing work to be done without the help of people have emerged as core technologies of the Fourth Industrial Revolution. Several industries with potential to bring the Fourth Revolution were addressed at the Davos Forum, e.g., bio industries, 3D printing, robots, artificial intelligence, and IoT (the Internet of things), sharing one thing in common: automation. It is still arguable whether the Fourth Industrial Revolution is a global innovation to create an entirely new world or just an extension of the information and communication revolution, but it is noteworthy that it changes the state of the industry and promotes an economic and social paradigm. This study aims to understand the global research trends and the level of research in Korea by selecting the core platform technologies that provide the driving force of the Fourth Industrial Revolution. History of Industrial Revolution from the Fourth Industrial Revolution Viewpoint 4 Analysis on Research Level of the Five Major Platform Technologies Related to the Fourth Industrial Revolution

6 П. Analyzed Technology Derivation and Technology Summaries 1 Deriving the Technology to Analyze As the platform technology constituting the Fourth Industrial Revolution is defined differently by institution or organization, in this research, five technologies - artificial intelligence, big data, IoT, cloud computing, and 3D printing - are selected as the main platform technologies and analyzed. Platform technology is a base module technology to be used commonly and repeatedly to provide various systems and services. (Kim, 2017) In this study, five platform technologies are selected based on the OECD report "The Next Production Revolution" * The OECD presented six major technologies, but simulation was excluded as it is practically difficult to analyze with keywords in papers; thus, the five aforementioned technologies were analyzed. Comparison of Core Technologies Presented for the Fourth Industrial Revolution Institutions WEF (World Economic Forum) Industry 4.0 OECD "Next Production Revolution" Presented technologies Unmanned transportation 3D printing Advanced robotics New materials IoT / remote monitoring Block chain / Bit coin Shared economy Genetic engineering Synthetic biology Bio-printing Big data and analysis Autonomous robots Simulation IoT Cyber security Cloud computing 3D printing Virtual reality Big data Cloud computing IoT AI 3D printing Simulation Source: Kim (2016) Ⅱ. Analyzed Technology Derivation and Technology Summaries 5

7 Convergence Development Flowchart of Digital Technology Related to the Fourth Industrial Revolution (OECD) The ultimate goal of the Fourth Industrial Revolution is research and development of automation technology that replaces humans To this end, technologies can be broadly divided into two categories: one is body technology and the other is soul technology. It can be said that robotics, genetic engineering, biology, and biotechnology are technologies for realizing the body, and artificial intelligence, big data, IoT, and cloud computing are technologies related to the soul, and 3D printing technology is for effective verification before each implementation. This study selected and analyzed artificial intelligence, big data, IoT, cloud computing, and 3D printing technology because these are most likely to be realized in the near future, and they already show some visible progress. 6 Analysis on Research Level of the Five Major Platform Technologies Related to the Fourth Industrial Revolution

8 2 3D Printing 3D printing technology, first introduced by 3D Systems in 1988 following Charles Hull's invention in 1984, first adopted SLA (stereolithography apparatus) technology but later developed a technique to create the molding through various methods with an input model. 3D printing technology varies by the material used and how it is processed. Plastic is mainly used as a material, but metals and paper can also be used. 3D printing dramatically improves manufacturing productivity through technology and procedures creating physical 3D objects by the use of digital technology. The common processes of 3D printing are thermoplastic extrusion, inkjet, photopolymer, and laser sintering, and these expanded the production range and promoted popularization. The processing stages of 3D printing are shown below, categorized depending on how it is sintered. Details Processing Sintering 1 Melt the material or prepare the powder 2 Create a cross section of an object through bonding or hardening process 3 The target object is completed by stacking the cross sections 3DP (3 dimension printing): Solidifying of powder material by spraying adhesive FDM (fused deposition modeling): Solidifying melted material deposited layer by layer Polyjet: Curing with UV light after liquid photopolymer resin is sprayed DLP (digital light processing): Curing by light to container with liquid photopolymer resin SLA (stereo-lithography apparatus): Solidifying by laser to container with liquid photopolymer resin SLS (selective laser sintering): Using a laser as the power source to sinter powder material Prototype development costs have been drastically reduced with the emergence of 3D printing technology. 3D printing technology is a base technology transforming an existing structure where the Pareto principle (reducing production cost through mass production in small variety) is dominated into the long-tail structure (producing small quantities in various kinds). The technology is currently recognized as a means to reduce the cost of prototype development. However, as the related technology advances, it is expected to bring a revolution in logistics by producing a product directly at home using a design plan sent to the customer online rather than delivering the product from the factory. Still, there are many challenges in making such an expectation reality in the near future. Thus, it is foreseeable that the technology will be used for producing small quantities of various items and developing prototypes. 3D printing has been used mainly for prototypes. Because of limitations in quality, cost, size, and time, the application area is now expanding into the development of 3D printer technology fields. In the short term, 3D printing is expected to spread quickly to industries with high cost / small quantities such as automobiles, aerospace, medical equipment, etc., it will take some time to penetrate low cost / middle or mass production. Ⅱ. Analyzed Technology Derivation and Technology Summaries 7

9 As the range of 3D print materials becomes wider (plastic, metal, ceramics, paper, cell, etc.), the ripple effect of the technology is also expected to increase. Major Applications of 3D Printers Dental medical device Human implantable medical device Customized device for treatments Smart mold Personalized personal items 3D electronic parts Transport Equipment parts Power generation parts 3D printing design service 3D printing content distribution service Source: Roadmap on 3D Printing Strategy Technology 8 Analysis on Research Level of the Five Major Platform Technologies Related to the Fourth Industrial Revolution

10 3 Internet of Things (IoT) IoT technology started from the concept of "ubiquitous computing," as suggested by Mark Weiser in 1988 Ubiquitous computing is a set of technologies proposed to eliminate the restricted access to the network only through the terminal by inserting the network and computing functions into all things so that users can access any network anytime and anywhere without restriction of space and time. In 1999, Kevin Ashton newly introduced the concept of IoT in which RFID and sensors were attached to existing objects and incorporated into the network. Both ubiquitous computing and IoT aim to build a continuous network access environment without restriction of time and space, but ubiquitous computing chooses a top-down method while IoT takes a bottom-up one. As the result of the bottom-up method, IoT is now more visible, and the concept has spread more widely. Components of IoT can be defined as things with intelligence, connectivity and communication, and new value offering. The technical elements are sensing, network infrastructure, interfaces, and security. Components Things with intelligence Connectivity and communication New value offerings Details Things with intelligence must have intelligence to act on their own Things must communicate with other objects connected to the network New value must be provided through generated information Composition and technological elements of the Internet of Things Sensing Acquisition and real-time delivery on information of circumstances Network Infrastructure Internet connection on objects Security Prevention of hacking and data leakage Service Interface Processing, handling and merging of collected information Source: Internet of Things (Connecting Lab) Ⅱ. Analyzed Technology Derivation and Technology Summaries 9

11 The main theme of IoT research is what kinds of things can work together to solve a given problem. To do this, it is necessary to identify what objects are presently accessible, judge how to work with them, classify and assign tasks to each things, and put together tasks performed by each entity. Currently, it is just a simple step to connect objects, but this is expected to advance later with research on how to link connected objects. The existing AI field was studying the collaboration between entities - so called the Multi-Agent System, and this part is expected to be a junction between AI and IoT. IoT is expected to be highly applicable through the technology of connecting devices and entities to the Internet network by incorporating sensors and communication functions in the objects. With the utilization of IoT, it is foreseeable that the production process to will be integrated, production costs will be reduced, and all other services and future prediction technologies will be carried out in a transparent and prompt way. 10 Analysis on Research Level of the Five Major Platform Technologies Related to the Fourth Industrial Revolution

12 4 Big Data The keywords of the Fourth Industrial Revolution are connectivity, data-based platform, speed, and automation, and big data* technology is the core of a hyper-connected and intelligent society in the future. * A massive volume of data produced from various sources (people, objects, sensors, etc.). Fourth Industrial Revolution and Big Data Technology The four key elements of the Fourth Industrial Revolution are 1 Internet of things, 2 Big data 3 Cloud computing 4 Artificial intelligence Big data acquired through IoT are stored in the cloud and analyzed and utilized through artificial intelligence. The Third Industrial Revolution Automation Informatization AI Big Data Cloud IoT Big Data is material for intelligence-oriented era. The Fourth Industrial Revolution Intelligence-oriented [Source: Kim Chang-gyeong (2017)] Source: National Science and Technology Advisory Committee (2017) Big data technology is divided into element technology for management, analysis, and offering of a data set, and comprehensive platform technology. Classification Big data analysis technology Big data processing technology Big data visualization technology Detailed technology Data mining, natural language processing, text mining, emotion analysis, SNS analysis, cluster analysis, etc. Hadoop, NoSQL, in-memory computing etc. Word cloud, infographic, etc. If the core of artificial intelligence is determining how to create and use knowledge, big data can be viewed as the underlying technology for collecting, generating, processing, and managing the source data sets to create the knowledge. The characteristics of big data are 3V (volume, velocity, and variety), as in recent years, and sometimes 4V (where veracity or value added are included). Ⅱ. Analyzed Technology Derivation and Technology Summaries 11

13 3V of Big Data Source: Big data is expected to realize data-driven innovation (DDI) by value extraction and result analysis of the large quantity of formal/informal data. DDI by big data will be widely seen in production processes, organizations, and markets, and companies are paying attention to the effects of increasing labor productivity. America's private enterprises like IBM, Microsoft, Google, Facebook, etc., are leading the global market, while the U.S. government is also promoting a R & D strategy. In Europe and Japan, where private data are not widely used, promoting public big data policies are led by the government, while China emphasizes the linkage between the government and industry. Academic research supporting the development of big data technology is actively carried out globally. 12 Analysis on Research Level of the Five Major Platform Technologies Related to the Fourth Industrial Revolution

14 5 Artificial Intelligence The history of artificial intelligence is largely divided into three stages. Recently, the discussion of machine learning through deep learning has attracted attention. Detailed technology advances 1950's 1980's After 2000 Recent American scholars held an AI debate at Dartmouth College in 1965 and later the AI research accelerated. * In the debate, scientists dreamed of a world in which machines (1) use language, (2) form concepts, (3) solve humankind's various problems, and (4) progress by themselves. Focus on the development of rule-based expert systems in well-defined domains. * It was well applied to narrowly defined problems, but it was impossible to learn/process uncertainties Sudden emergence of machine learning through deep learning. * The milestones where AI showed ability beyond human performance: chess (1977), quiz game (2011), Strong (2013), video/speech recognition (2015), the game of go (2016). Research of strong AI technology is started. * Current artificial intelligence technology is weak (narrow) AI, carrying out a certain given task. And strong (general) AI technology that can be applied to various recognition areas is still in its infancy. Artificial intelligence can be largely classified into strong AI and weak AI, differentiated in accordance with the degree of similarity between human intelligence and artificial intelligence. Strong AI: Strong AI not only behaves like a person, but it should be like a person on the inside. It is also considered impossible in reality, as the way the intelligence works is basically unknown. Weak AI: Examples of weak AI are Deep Blue, who beat the chess champion, Watson, who won Jeopardy, Alpha Go, NPCs in MMORPG games, and other AI in the game. Weak AI has enough to mimic as if it has intelligence, and imitating intelligence for limited functions in limited domains is also possible. There are two ways to create knowledge for artificial intelligence: the symbolic approach and the statistical approach. Symbolic approach: The symbolic approach is a way to use symbols to represent the knowledge of domain experts. Basically, this follows the learning process of people with new knowledge. *Advantages: Knowledge acquisition and behaviors based on the idea that knowledge is clearly observed. *Disadvantages: Experts often fail to express their knowledge in a clear way. Statistical approach: The statistical approach is a method to solve the problems of the symbolic approach, gaining knowledge by extracting the statistical meaning in the data set after aggregating sufficient data. As the performance of the machine increases, knowledge generation through machine learning is attracting attention because more data can be collected easier, and statistical results can be calculated faster from the collected data. Ⅱ. Analyzed Technology Derivation and Technology Summaries 13

15 Around AI, big data, cloud computing, IoT, block chain and fintech, VR, and AR are combined to create new platform and a value. > The role of AI and humans is optimized and evolving through robots and the Internet as a medium in CPS. (Evolution direction: simple knowledge expert knowledge judgment knowledge) Cyber and physical space Cyber : AI Learn various situations relating to human activities Optimized instruction and suggestion proposed Big data accumulation / analysis new goals and hypotheses provided Interface: Robot and Internet Medium that connects cyber space and physical space Tasks requested by AI and humans are carried out and feedback is given Physical: Human New role as a user, manager, and developer of AI is expanded Fusion of technology related to Fourth Industrial Revolution around AI Coexistence of AI and Human Source: Industrial strength improvement through co-evolution by AI, robot and human, Industrial Competitiveness Council (2016). AI is expected to increase the level of economy and industry, improve educational opportunities and quality of life, and be utilized in the administration and security sectors. Vision Economy and industry Improve educational opportunities and quality of life Governmental administration and security Applied field manufacturing, logistics, finance, transportation, agriculture, marketing, communication, and science and technology education, medical, legal, and customized services security and law enforcement, safety, and forecasting 14 Analysis on Research Level of the Five Major Platform Technologies Related to the Fourth Industrial Revolution

16 6 Cloud Computing Cloud computing is a concept combining grid computing with server-based computing technology. Server-based computing is a type of computing environment in which one initial server computer and several terminals are combined. In the early days of computers, as the computer speed was faster than the actual computer operation, it was more efficient for multiple users to use a single computer than each one using one computer. This is a structure designed to make full use of the CPU during wait time because the CPU wait rate is high, as I/O use time is longer than CPU usage time when using computers. Advantages and Disadvantages of Server-Based Computing Systems Note Advantages Disadvantages Contents - Basically, the terminal used by each user has a function that transmits only input to the server computer and receives the result value, so the cost of the terminal is low - Users can create their own work environment if they can communicate with the server without space restrictions. - User must be online to do the task. A commumication cost is added to every task, and communication traffic is increased accordingly. - As the content of the task and user information are stored on the server, it is vulnerable to server security breaches. Grid computing is the collection of relatively inexpensive computer resources from multiple locations to reach the result of a supercomputer as a process of bundling several computers, splitting the content to be executed, assigning content to each computer, and collecting and combining the results. As the advancement of grid computing complemented the limitations of the existing single server, the benefits of a server-based computing system became available. Today, cloud computing technology uses client devices working as a terminal to connect to the cloud server, allowing the work environment to be used without constraints of time and space, and, with the server functions enhanced by grid computing, it is used as underlying technology to perform complex tasks at a low cost. Ⅱ. Analyzed Technology Derivation and Technology Summaries 15

17 Cloud computing services have software, infrastructure, and platform as services. Cloud computing service model Classification SaaS (Software as a Service) IaaS (Infrastructure as a Service) PaaS (Platform as a Service) Content A software package as a service to the user. Data access support through resource provision such as servers Offering an environment to develop software Based on the Internet, cloud computing is a technology contributing to business growth and expansion by connecting computer processing resources or data to computers or other devices and sharing them. The use of cloud computing enables the tasks of organizations to be consolidated in common data storage and allows rapid processing and interoperability. Startups and small and medium-sized companies can lower ICT investment costs by providing large amounts of computing resources Applicable Areas of Cloud Computing Source: 16 Analysis on Research Level of the Five Major Platform Technologies Related to the Fourth Industrial Revolution

18 Ш. Research Paper Analysis Methods Analysis DB and tool Analysis DB: Scopus DB (Elsevier) Analysis Tool: Elsevier SciVal Analytics Engine Scopus-based research performance analysis solution - Can be used to establish strategic decision making and research strategy developed through research and comparative analysis of more than 7,000 organizations and 220 countries Consists of an analytical menu such as Overview, Benchmarking, Collaboration, and Trends - Overview: Analysis of research results of more than 7,000 organizations and 220 countries - Benchmarking: Various analyses using 15 metrics - Collaboration: Analysis of existing research cooperation results and searching for future research cooperation partners - Trends: Analysis on research trend in the world around research keywords Analysis period: (5 years) (as of July 2017) Ⅲ. Research Papers Performance Analysis Method 17

19 Evaluation metrics Among the most commonly used indicators used in the analysis of research results, the analysis of each applied indicator was conducted based on the following evaluation metrics. Evaluation Metrics for Major Research Performance Evaluation Metrics Scholarly Output Citation Count Field Weighted Citation Impact(FWCI) Citations per Publication Absolute number of Top 1 Percentiles The total count of papers published by specific countries, institutions and researchers, to represent productivity The total number of citations of the paper published by specific countries, institutions, and researchers The ratio of the total citations actually received by the denominator s output, and the total citations that would be expected based on the average of the subject field. Average FWCI index =1, ex) When FWCI is 1.23, it means the research was cited 23% more compared to world average Average number of citations per paper Ratio of cited papers that are in the top 1 % in the world Absolute number of Top 10 Percentiles International Collaboration rate The total count of cited papers that are in the top 10% in the world The ratio of papers published together by authors in different countries 18 Analysis on Research Level of the Five Major Platform Technologies Related to the Fourth Industrial Revolution

20 Extracted Keywords Related to Five Technologies in the Papers Technology Sub technology Extracted keywords 3D printing IoT Big data AI Cloud computing Material technology for 3D Printing Spraying and processing for 3D Printing IoT Security IoT Platform IoT Network IoT Device Big Data Storage, processing and management technology Big Data collecting technology Big Data analysis and prediction technology Learning and reasoning technology Other AI technologies Situation Understanding technology Visual Understanding Technology Language Understanding Technology Neurocomputing Technology Other Clouding technology Cloud linking technology Cloud platform technology ((3D or 3-D or 3-dimension or "Three dimensional") and (print* or manufacture or product or "integrated circuits")) or (("rapid prototype" or additive or "layered manufacture") and (photolithography or Photopolymerization or stereolithography or "laser sinter")) IoT or IOT or internet of things or MTC or "machine-tomachine" or "machine communication" or "smart object" or "smart thing" or "web of things" or WoT or "internet of everything" or IoE or "machine orient" or "smart object" or "ubiquitous sensor" or "Wireless sensor networks" or "network protocols" or "embedded systems" or "cyber physical systems" ((Big or large or volume) and (data or "data handling" or "data mining" or "data processing" or "data visualization" or "database systems" or "distributed computer systems" or "Learning systems" or "Predictive analytics" or "Competitive intelligence")) ("artificial intelligence" or AI or "Adaptive systems" or "Ambient intelligence" or "Autonomous agents" or Backpropagation or Biocybernetics or "Brain models" or "Case based reasoning" or "cognitive systems") or (((machine or unsupervised or supervised or Reinforcement) and learning) or (neuromorphic or "neuro feedback" or "brain feedback" or "brain wave" or EEG or Electroencephalogram or artificial neural network or BCI or "brain computer interface")) cloud or virtualization or "Virtual machine" or "Computer operating systems" or "Cloud chambers" or "Cloud computing" or "Cloud seeding" or "Cluster computing" or "Data as a service" or "DaaS" or "Grid computing" or "Infrastructure as a service" or IaaS Ⅲ. Research Papers Performance Analysis Method 19

21 IV. Research Paper Performance Analysis 1 Summary on Result Analysis Among the major countries, the U.S. and China are at the top of the field in terms of the number of papers, while Germany, England, and India also show strength. In quantitative terms (the number of papers) Korea placed on eighth in 3D printing and fifth in IoT, but in Big data (14th), AI(13th), Cloud computing(12th), they were relatively weak. Country Ranking on Five Technologies (Based on Scholarly Output) Ranking 3D printing IoT Big data AI Cloud computing 1 U.S. China U.S. U.S. U.S. 2 China U.S. China China China 3 Germany India U.K. U.K. Germany 4 U.K. Germany Germany Germany U.K. 5 Japan Korea(5th) France India India Korea(8th) - Korea(14th) Korea(13th) Korea(12th) In terms of quality including FWCI*, Korea's standing is lower than that of the U.S. and major European countries and higher than other Asian countries including China and Japan. * FWCI is the ratio of the total citations actually received by the denominator s output, and the total citations that would be expected based on the average of the subject field. Average FWCI index =1, E.g. A FWCI of 1.23 means the research was cited 23% more compared to the world average FWCI of Major Countries on Five Platform Technologies 20 Analysis on Research Level of the Five Major Platform Technologies Related to the Fourth Industrial Revolution

22 Korea accounts for 2.7% of the world's papers, slightly below the world average in terms of quality. Compared to the U.S., a lead country in the field, the quality gap is large. Comparison of Research Papers : World vs. Korea vs. U.S. Category 3D printing IoT Big data AI Cloud computing Scholarly Output (5 years) 78,237 79, , , ,270 World FWCI Citation per Publication Scholarly Output (5 years) 3,201 3,885 9,107 6,246 6,812 Korea FWCI Citation per Publication Scholarly Output (5 years) 17,833 11, ,535 71,919 53,470 U.S. FWCI Citation per Publication FWCI on Five Platform Technology IV. Research Papers Performance Analysis Result 21

23 Institutions in China, and France showed strength in terms of scholarly output, while in terms of quality index, the American institutions placed in the upper ranks. One to six domestic institutions are positioned within the top 100 of Scholarly output standards for each category. Ranking of Major Research Institutes (Based on Scholarly Output) Ranking 3D printing IoT Big data AI Cloud computing 1 Chinese Academy of Sciences Chinese Academy of Sciences CNRS Chinese Academy of Sciences CNRS 2 Ministry of Education China Ministry of Education China Chinese Academy of Sciences CNRS Chinese Academy of Sciences 3 CNRS CNRS Harvard Harvard Universite Paris Saclay 4 Harvard Beijing of Posts and elecommunications Universite Paris Saclay Ministry of Education China Ministry of Education China 5 Universite Paris Saclay Universite Paris Saclay of Toronto College London Harvard Seoul National (25th) KAIST(86th) ETRI(25th) Sungkyunkwan (33th) Seoul National (100th) Seoul National (74th) Seoul National (70th) Korea (90th) Sungkyunkwan (95th) KAIST(42th) Korea (77th) Seoul National (90th) Kyungpook National (99th) 22 Analysis on Research Level of the Five Major Platform Technologies Related to the Fourth Industrial Revolution

24 2 Analysis of 3D Printing Summary of major indicators In the past five years, more than 78,000 papers have been searched in the world. The major indicators used in the field are as follows. Performance Indicators of the World's Publication Performance Indicator Total/Average Scholarly Output 13,117 14,209 15,854 16,747 18,310 78,237 Citation Count 139, ,494 93,751 63,020 25, ,169 FWCI Citation per Publication In the past five years, more than 3,000 papers have been searched in Korea. The major indicators used in the field are as follows. Performance Indicators of Korea's Publication Performance Indicator Total/Average Scholarly Output ,201 Citation Count 6,232 5,754 4,927 3,177 1,214 21,304 FWCI Citation per Publication The FWCI and citations per publication trends in the last five for the world versus Korea are as follows. FWCI : World vs. Korea Citation per Publication : World vs. Korea IV. Research Papers Performance Analysis Result 23

25 Analysis of keywords The major keyword in the world's publications in the field is shown below. Major Analysis on Keywords(World) * Text size represents frequency; red represents increase and blue represents decrease. The major keywords in Korea's publications are the field is shown below. Major Analysis on Keywords(World) * Text size represents frequency; red represents increase and blue represents decrease. Comparison of the Top Ten Keywords in the World and Korea Ranking World's publications Korea's publications 1 Printing Printing 2 Manufacture Three dimensional 3 Three dimensional Design 4 Design Integrated circuits 5 Three dimensional computer graphics Scaffolds 6 Production Three dimensional computer graphics 7 Technology Electronics packaging 8 Rapid prototyping Production 9 Models Printing presses 10 Antennas Fabrication 24 Analysis on Research Level of the Five Major Platform Technologies Related to the Fourth Industrial Revolution

26 Analysis of countries In comparing the number of publications by country, the U.S., China, and Germany ranked in the top three and Korea ranked eighth. Publication Status Comparing the major indicators of the top ten countries, the U.S., China, and U.K. represented the top three in the absolute number of Top 1 percentiles, and FWCI which shows quality standards, the U.K., U.S., and Italy followed. Ranking (Scholarly Output) Country Scholarly Output Absolute number of Top 1 Percentiles FWCI Citation per Publication 1 U.S. 17, China 15, Germany 5, U.K. 4, Japan 3, France 3, India 3, Korea 3, Italy 2, Canada 2, IV. Research Papers Performance Analysis Result 25

27 Analysis of each institution Among the major institutions, the Chinese and French institutions such as China Science Academy (CAS) and CNRS, showed strength in scholarly output. Korea's Seoul National and KAIST ranked 25th and 86th, respectively. Based on FWCI criteria for qualitative levels, Harvard and other institutions in the U.S. including MIT and Cornell, were on top. Major Institutions Based on Scholarly Output(World) Ranking Institution Scholarly Output Absolute number of Top 1 Percentiles FWCI Citation per Publication 1 Chinese Academy of Sciences 1, Ministry of Education China 1, CNRS Harvard Universite Paris Saclay Tsinghua Zhejiang CSIC CEA U.S. Department of Agriculture Massachusetts Institute of Technology Universidade de Sao Paulo Peking RAS of Tokyo Shanghai Jiaotong of Texas at Austin of Michigan Georgia Institute of Technology Texas A and M Seoul National KAIST Korea Sungkyunkwan Analysis on Research Level of the Five Major Platform Technologies Related to the Fourth Industrial Revolution

28 Korea s top 20 institutions are given below. Seoul National, KAIST, and Korea were in the upper ranks based on scholarly output. Major Institutions Based on Scholarly Output(Korea) Ranking Institution Scholarly Output Absolute number of Top 1 Percentiles FWCI Citation per Publication 1 Seoul National KAIST Korea Sungkyunkwan Yonsei Hanyang POSTECH Pusan National Kyung Hee Kyungpook National Chungnam National Konkuk Chonnam National Samsung Inha Ajou Chonbuk National ETRI Kangwon National KIST IV. Research Papers Performance Analysis Result 27

29 3 Analysis of IoT Summary of major indicators In the past five years, more than 80,000 papers have been searched in the world. The major indicators used in the field are as follows. Performance Indicators of the World's Publication Performance Indicator Total/Average Scholarly Output 13,927 14,628 15,286 16,338 19,492 79,671 Citation Count 78,075 68,050 55,535 35,364 12, ,464 FWCI Citation per Publication In the past five years, more than 4,000 papers have been searched in Korea. The major indicators used in the field are as follows. Performance Indicators of Korea s Publications Performance Indicator Total/ Average Scholarly Output ,060 3,885 Citation Count 4,614 2,744 2,423 1, ,197 FWCI Citation per Publication The FWCI and citations per publication trends in the last five years for the world versus Korea are as follows. FWCI : World vs. Korea Citation per Publication : World vs. Korea 28 Analysis on Research Level of the Five Major Platform Technologies Related to the Fourth Industrial Revolution

30 Analysis of keywords The major keywords in the world's publications in the field are shown below. Major Analysis on Keywords(World) * Text size represents frequency; red represents increase and blue represents decrease. The major keyword in Korea's publications in the field are shown below. Major Analysis on Keywords(Korea) * Text size represents frequency; red represents increase and blue represents decrease. Comparison of the Top Ten Keywords of the World and Korea Ranking World's publications Korea's publications 1 Wireless sensor networks Wireless sensor networks 2 Sensor nodes Sensor nodes 3 Embedded systems Internet 4 Internet Embedded systems 5 Sensor networks Sensor networks 6 Energy utilization Network protocols 7 Network protocols Energy efficiency 8 Algorithms Energy utilization 9 Sensors Sensors 10 Energy efficiency Communication IV. Research Papers Performance Analysis Result 29

31 Analysis of countries Comparing the number of publications by country, China, the U.S., and India ranked as the top three and Korea ranked fifth. Publication Status Comparing the major indicators of the top ten countries, the U.S., China, and Italy were the top three in the Absolute number of Top 1 Percentiles, and FWCI which shows quality standards, Italy, the U.S., and the U.K. followed. Ranking (Scholarly Output) Country Scholarly Output Absolute number of Top 1 Percentiles FWCI Citation per Publication 1 China 19, U.S. 11, India 7, Germany 4, Korea 3, Italy 3, France 3, U.K. 3, Spain 2, Japan 2, Analysis on Research Level of the Five Major Platform Technologies Related to the Fourth Industrial Revolution

32 Analysis of each institution Among the major institutions, Chinese and French institutions such as China Science Academy (CAS) and CNRS showed strength in scholarly output. Korea's ETRI and Sungkyunkwan ranked 25th and 33rd, respectively. Based on the FWCI criteria for qualitative levels, institutions in Europe and the U.S. including of Bologna, Swiss Federal Institute of Technology Zurich, of Michigan, of California, Berkeley, and Intel ranked at the top. Ranking Major institutions based on Scholarly Output(World) Institution Scholarly Output Absolute number of Top 1 Percentiles FWCI Citation per Publication 1 Chinese Academy of Sciences Ministry of Education China CNRS Beijing of Posts and Telecommunications Universite Paris Saclay Tsinghua INRIA Institut National de Recherche en Informatique et en Automatique 8 Zhejiang Shanghai Jiaotong Anna Northeastern China Nanjing of Posts and Telecommunications Southeast Harbin Institute of Technology Huazhong of Science and Technology Beihang Nanyang Technological Xidian Islamic Azad Beijing Jiaotong ETRI Sungkyunkwan KAIST Korea Seoul National Kyungpook National IV. Research Papers Performance Analysis Result 31

33 Korea s top 20 institutions are shown below. Based on scholarly output, ETRI, Sungkyunkwan, and KAIST were in the upper ranks. Major institutions based on Scholarly Output(Korea) Ranking Institution Scholarly Output Absolute number of Top 1 Percentiles FWCI Citation per Publication 1 ETRI Sungkyunkwan KAIST Korea Seoul National Kyungpook National Yonsei Hanyang Ajou Kyung Hee Samsung Soongsil Chungnam National Dongguk Konkuk Chung Ang Seoul National of Science and Technology of Ulsan Chonbuk National Pusan National Analysis on Research Level of the Five Major Platform Technologies Related to the Fourth Industrial Revolution

34 4 Analysis of Big Data Summary of major indicators In the past five years, more than 390,000 papers have been searched in the world. The major indicators used in the field are as follows. Performance Indicators of the World s Publication Performance Indicator Total/Average Scholarly Output 66,809 72,967 79,915 83,798 88, ,948 Citation Count 905, , , , ,724 2,585,491 FWCI Citation per Publication In the past five years, more than 9,000 papers have been searched in Korea. The major indicators used in the field are as follows. Performance Indicators of Korea's Publication Performance Indicator Total/Average Scholarly Output 1,450 1,533 1,859 2,115 2,150 9,107 Citation Count 23,950 13,853 14,647 8,748 2,910 64,108 FWCI Citation per Publication The FWCI and citations per publication trends in the last five years for the world versus Korea are as follows. FWCI : World vs. Korea Citation per Publication : World vs. Korea IV. Research Papers Performance Analysis Result 33

35 Analysis of keywords The major keyword in the world's publications in the field are shown below Major Analysis on Keywords(World) * Text size represents frequency; red represents increase and blue represents decrease. The major keyword in Korea's publications in the field are shown below Major Analysis on Keywords(Korea) * Text size represents frequency; red represents increase and blue represents decrease. Comparison of the Top Ten Keywords of the World and Korea Ranking World's publications Korea's publications 1 Models Models 2 Algorithms Algorithms 3 Data mining Korea 4 Patients Data handling 5 Digital storage Data mining 6 Data handling Republic of Korea 7 Cloud computing Patients 8 Research Digital storage 9 Applications Cloud computing 10 Database systems Industry 34 Analysis on Research Level of the Five Major Platform Technologies Related to the Fourth Industrial Revolution

36 Analysis of countries In comparing the number of publications by country, the U.S., China, and U.K. ranked as the top three and Korea ranked eighth. Publication Status Comparing the major indicators of the top ten countries, the U.S., U.K., and Germany took the top three places in the Absolute number of Top 1 Percentiles, and FWCI which shows quality standards, the U.K., Canada, and Australia followed. Ranking (Scholarly Output) Country Scholarly Output Absolute number of Top 1 Percentiles FWCI Citation per Publication 1 U.S. 122,535 3, China 61, U.K. 35,978 1, Germany 33,159 1, France 21, Italy 20, Canada 19, Australia 18, India 16, Japan 14, Korea 9, IV. Research Papers Performance Analysis Result 35

37 Analysis of each institution Among the major institutions, Chinese, French, and U.S. institutions such as China Science Academy (CAS), CNRS and Harvard showed strength in scholarly output. Korea's Seoul National ranked 100th. Based on the FWCI criteria for qualitative levels, institutions in the U.S. and U.K including of Chicago, UC Berkeley, Oxford, and Carnegie Mellon were on top. Major institutions based on Scholarly Output(World) Ranking institution Scholarly Output Absolute number of Top 1 Percentiles FWCI Citation per Publication 1 CNRS 7, Chinese Academy of Sciences 7, Harvard 6, Universite Paris Saclay 3, of Toronto 3, Stanford 3, College London 3, Ministry of Education China 3, of Oxford 3, of Washington 3, of Michigan 3, of Cambridge 2, Johns Hopkins 2, Imperial College London 2, CSIC 2, Tsinghua 2, of California at Los Angeles 2, Columbia 2, Massachusetts Institute of Technology 2, of British Columbia 2, Seoul National 1, Analysis on Research Level of the Five Major Platform Technologies Related to the Fourth Industrial Revolution

38 Korea s top 20 institutions are shown below. Based on scholarly output, Seoul National, Yonsei, and Korea were in the upper ranks. Major institutions based on Scholarly Output(Korea) Ranking Institution Scholarly Output Absolute number of Top 1 Percentiles FWCI Citation per Publication 1 Seoul National 1, Yonsei Korea Sungkyunkwan KAIST Kyung Hee Kyungpook National Hanyang of Ulsan Pusan National Chonnam National Konkuk Ewha Womans The Catholic of Korea Kangwon National Chonbuk National Inha POSTECH Ajou ETRI IV. Research Papers Performance Analysis Result 37

39 5 Analysis of AI Summary of major indicators In the past five years, more than 280,000 papers have been searched in the world. The major indicators used in the field are as follows. Performance Indicators of the World's Publication Performance Indicator Total/Average Scholarly Output 50,537 53,338 56,115 57,282 63, ,707 Citation Count 522, , , ,470 63,893 1,530,356 FWCI Citation per Publication In the past five years, more than 6,000 papers have been searched in Korea. The major indicators used in the field are as follows Performance Indicators of Korea's Publications Performance Indicator Total/Average Scholarly Output 1,077 1,151 1,164 1,311 1,543 6,246 Citation Count 12,032 9,748 6,472 4,663 1,624 34,539 FWCI Citation per Publication The FWCI and citations per publication trends in the last five years for the world versus Korea are as follows. FWCI : World vs. Korea Citation per Publication : World vs. Korea 38 Analysis on Research Level of the Five Major Platform Technologies Related to the Fourth Industrial Revolution

40 Analysis of keywords The major keyword in the world's publications in the field are shown below. Major Analysis of Keywords(World) * Text size represents frequency; red represents increase and blue represents decrease. The major keyword in Korea's publications in the field are shown below. Major Analysis of Keywords(Korea) * Text size represents frequency; red represents increase and blue represents decrease. Comparison of Top Ten Keywords of the World and Korea Ranking World's publication Korea's publication 1 Neural networks Neural networks 2 Learning systems Learning systems 3 Artificial intelligence Artificial intelligence 4 Models Neural Stem Cells 5 Learning algorithms Learning algorithms 6 Classification (of information) Support vector machines 7 Algorithms Models 8 Support vector machines Classification (of information) 9 Forecasting Forecasting 10 Brain Brain IV. Research Papers Performance Analysis Result 39

41 Analysis of countries Comparing the number of publications by country, the U.S., China, and U.K. ranked as the top three and Korea ranked 13th. Publication Status Comparing the major indicators of the top ten countries, the U.S., China, and U.K. took the top three places in the Absolute number of Top 1 Percentiles, and FWCI which shows quality standards, the U.S., U.K., and Canada followed. Ranking (Scholarly Output) Country Scholarly Output Absolute number of Top 1 Percentiles FWCI Citation per Publication 1 U.S. 71,919 2, China 54, U.K. 19, Germany 16, India 15, Japan 12, Canada 11, Italy 10, France 10, Australia 9, Korea 6, Analysis on Research Level of the Five Major Platform Technologies Related to the Fourth Industrial Revolution

42 Analysis of each institution Among the major institutions, the Chinese, French, and U.S. institutions such as China Science Academy (CAS), CNRS, and Harvard showed strength in scholarly output. Korea's Seoul National ranked in 74th. Based on FWCI criteria for qualitative levels, institutions in the U.S. and Canada including Microsoft, UC Berkeley, Howard Hughes Institute, Stanford, and of Montreal, of Toronto were on top. Major institutions based on Scholarly Output(World) Ranking Institution Scholarly Output Absolute number of Top 1 Percentiles FWCI Citation per Publication 1 Chinese Academy of Sciences 3, CNRS 3, Harvard 3, Ministry of Education China 3, College London 2, of Toronto 2, Stanford 1, Islamic Azad 1, INSERM 1, Massachusetts Institute of Technology 1, of California at Los Angeles 1, Tsinghua 1, Johns Hopkins 1, Carnegie Mellon 1, of California at San Diego 1, of Michigan 1, of Oxford 1, National Institutes of Health 1, Shanghai Jiaotong 1, Zhejiang 1, Seoul National IV. Research Papers Performance Analysis Result 41

43 Korea s top 20 institutions are shown below. Based on scholarly output, Seoul National, KAIST, and Korea were in the upper ranks. Major institutions based on Scholarly Output(Korea) Ranking Institution Scholarly Output Absolute number of Top 1 Percentiles FWCI Citation per Publication 1 Seoul National Korea KAIST Yonsei Sungkyunkwan Hanyang Kyung Hee Yeungnam Kyungpook National of Ulsan Chung Ang Pusan National Chonnam National Inha Ajou POSTECH Chonbuk National KIST ETRI Konkuk Analysis on Research Level of the Five Major Platform Technologies Related to the Fourth Industrial Revolution

44 6 Analysis of Cloud Computing Summary of major indicators In the past five years, more than 230,000 papers have been searched in the world. The major indicators used in the field are as follows. Performance Indicators of the World's Publication Performance Indicator Total/ Average Scholarly Output 41,334 44,282 46,755 47,182 48, ,270 Citation Count 333, , , ,650 39, ,754 FWCI Citation per Publication In the past five years, more than 7,000 papers have been searched in Korea. The major indicators used in the field are as follows. Performance Indicators of Korea s Publications Performance Indicator Total/ Average Scholarly Output 1,141 1,248 1,402 1,493 1,528 6,812 Citation Count 8,033 7,721 6,230 3,313 1,316 26,613 FWCI Citation per Publication The FWCI and citations per publication trends in the last five years for the world versus Korea are as follows. FWCI : World vs. Korea Citation per Publication : World vs. Korea IV. Research Papers Performance Analysis Result 43

45 Analysis of keywords The major keyword in the world's publications in the field are shown below. Major Analysis of Keywords(World) * Text size represents frequency; red represents increase and blue represents decrease. The major keyword in Korea's publications in the field are shown below. Major Analysis of Keywords(Korea) * Text size represents frequency; red represents increase and blue represents decrease. Comparison of the Top Ten Keywords of the World and Korea Ranking World's publications Korea's publications 1 Clouds Cloud computing 2 Cloud computing Clouds 3 Virtual reality Virtual reality 4 Models Mobile devices 5 Applications Applications 6 Algorithms Augmented reality 7 Technology Models 8 Research Servers 9 Industry Technology 10 Design Management 44 Analysis on Research Level of the Five Major Platform Technologies Related to the Fourth Industrial Revolution

46 Analysis of countries When comparing the number of publications by country, the U.S., China, and Germany ranked in the top three and Korea ranked 12th. Publication Status When comparing the major indicators of the top ten countries, the U.S., Germany, and U.K. were the top three in the Absolute number of Top 1 Percentiles, and FWCI which shows quality standards, Italy, the U.S., and Canada followed. Ranking (Scholarly Output) Country Scholarly Output Absolute number of Top 1 Percentiles FWCI Citation per Publication 1 U.S. 53,470 1, China 41, Germany 17, U.K. 15, India 11, France 10, Italy 10, Japan 9, Canada 8, Spain 8, Korea 6, IV. Research Papers Performance Analysis Result 45

47 Analysis of each institution Among the major institutions, the Chinese and French institutions such as China Science Academy (CAS), CNRS, and Universite Paris Saclay showed strength in scholarly output. Korea's Seoul National ranked 70th. Based on the FWCI criteria for qualitative levels, institutions in the U.S. and Australia including of Melbourne, UC Berkeley, Carnegie Mellon, Washington, and Stanford were on top. Major institutions based on Scholarly Output(World) Ranking Institution Scholarly Output Absolute number of Top 1 Percentiles FWCI Citation per Publication 1 CNRS 3, Chinese Academy of Sciences 3, Universite Paris Saclay 2, Ministry of Education China 1, Harvard 1, Tsinghua 1, of Tokyo 1, Universite Pierre et Marie Curie 1, Beijing of Posts and Telecommunications 1, PSL Research 1, of Toronto 1, Beihang 1, Shanghai Jiaotong 1, RAS 1, California Institute of Technology 1, ETH Zurich 1, Wuhan 1, College London 1, of Wisconsin 1, Zhejiang 1, Seoul National Analysis on Research Level of the Five Major Platform Technologies Related to the Fourth Industrial Revolution

48 Korea s top 20 institutions are shown below. Based on scholarly output, Seoul National, KAIST, and Electronics and Telecommunications Research Institute were in the upper ranks. Major institutions based on Scholarly Output(Korea) Ranking institution Scholarly Output Absolute number of Top 1 Percentiles FWCI Citation per Publication 1 Seoul National KAIST ETRI Korea Yonsei Kyung Hee Sungkyunkwan Hanyang Samsung Dongguk Kyungpook National POSTECH KISTI Konkuk Pusan National Chonnam National KIST GIST Chung Ang Soongsil IV. Research Papers Performance Analysis Result 47

49 V. Status of NRF Research Project Support Analysis summary Analysis Purpose: To analyze the status of research projects supported by the National Research Foundation of Korea by linking the performance analysis of research Analyzed DB and items: Project title and research keywords of National Research Foundation of Korea for the last five years (from 2012 to 2016) Projects subject to analysis are limited to basic research projects and national policy research projects in the science and engineering fields. Analyzed keywords Technology 3D printing IoT Big Data AI Cloud Computing Major keywords 3D printing, cell printing, 3D lithography, 3D patterning, 3D micro manufacturing, additive manufacture IoT(internet of everything), Internet of Things, M2M(machine to machine) Big Data AI, artificial intelligence, machine learning, deep learning Cloud, virtualization Extract projects via major and similar keywords in accordance with research project keywords DB, excluding projects that are not relevant Status of support Support for five technologies has risen in recent years. It is analyzed that approximately 700 projects and KRW 73 billion were supported as of Support Status ( ) Technology D printing Number of projects Won (hundred million) IoT Number of projects Won (hundred million) Big Data Number of projects Won (hundred million) AI Number of projects Won (hundred million) Cloud Computing Number of projects Won (hundred million) Total Number of projects Won (hundred million) Some projects overlap in two to three categories. 48 Analysis on Research Level of the Five Major Platform Technologies Related to the Fourth Industrial Revolution

50 In particular, resources for big data after 2013 and AI after 2016 has seen dramatic increases. Trend in the Number of Projects Supported for Each Technology Trend in Volume of Funds for Each Technology Support status by project (Basic research projects) consist of 86% of the supported projects and 62% of the overall funding volume. (National research projects) consist of only 14% of the supported projects but take 38% of the overall funding volume. Project Basic research project National research projects Number of projects Won (hundred million) Number of projects Won (hundred million) V. Current Status of NRF Research Project Support 49