scenarios and potentials of ai’s commercial application in ...€¦ · basic layer technology...
TRANSCRIPT
Scenarios and Potentials of AI’s Commercial Application in China
1. China’s AI industry overview 1
2. Industry-specific commercial application 5
3. Regions’ potentials of commercialization 14
Contents
1
Scenarios and Potentials of AI’s Commercial Application in China | China’s AI industry overview
112.4141.9
216.9
339.0
500.0
710.0
0
100
200
300
400
500
600
700
800
2015 2016 2017 2018E 2019E 2020E
CAGR: 44.5%
1,684 1,9712,307
2,700
4,285
6,800
0
1,000
2,000
3,000
4,000
5,000
6,000
7,000
(RMB100 million)
2015 2016 2017 2018E 2019E 2020E
CAGR: 26.2%
1. China’s AI industry overview
Globally, governments are mapping out national strategies on developing artificial intelligence (AI) which becomes the new engine in the future of world economy. The global AI market is estimated to reach RMB680 billion by 2020 with a compound annual growth rate (CAGR) of 26.2%. Meanwhile, China’s AI market is expected to be RMB71 billion by 20201since its growth from 2015, with a CAGR of 44.5% from 2015 to 2020.
Despite the rapid growth, China is still young in developing AI technologies. Currently, the U.S. is way ahead of China in several indicators of key AI fields, such as global chip market share, number of talents and research capabilities.
Figure 1. Global AI market size
Figure 2. China’s AI market size (2015-2020)
Source: chyxx.com, Deloitte Research
1. 2017 China’s AI Industry Data Report, China Academy of Information and Communications Technology
Source: chyxx.com, Deloitte Research
■ China’s market size (RMB100 million)
2
Figure 3. The U.S. is way ahead of China in several indicators
Key fields Indicators China U.S.
Hardware Global market share of semiconductor products (2015)
4% 50%
Financing of FPGA chip manufacturers (2017)
USD34.4 million (7.6% of the global total)
USD192.5 million (42.2% of the global total)
Data Number of mobile subscribers (2016)
1.4 billion (20% of the global total)
420 million (5.5% of the global total)
Research capability and paradigm
Number of AI experts 39,200 (13% of the global total) 78,200 (26% of the global total)
Proportion of speeches delivered at AAAI (2015)
20.5% of the global total 48.4% of the global total
Commercialization Proportion of AI companies (2017)
23% of the global total 42% of the global total
Investments gained by AI companies (2012-2016)
USD2.6 billion (6.6% of the global total)
USD17.2 billion (43.4% of the global total)
Investments from PEs for startups (2017)
48% of the global total 38% of the global total
Source: Public information, Deloitte Research
Scenarios and Potentials of AI’s Commercial Application in China | China’s AI industry overview
3
Though it hasn’t been smooth sailing for AI development, the explosive growth from 2015 has been different from the past as AI in the new era has been commercialized, mainly driven by improved computing power, top-level design, capital support and user demand.
AI industrial chain falls into the following three layers:
Computing power is provided in the basic layer, mainly including AI chip, sensor, big data and cloud computing, in which China is relatively weak. Tech giants represented by BATJ have started to build their own AI related basic discipline labs, increase inputs to R&D and expand investments into start-ups in the basic layer.
Category-specific problems are addressed in the technology layer. China has experienced rapid growth in the technology layer and now focused primarily on computer vision, speech recognition and language processing. Many giants, including Tencent, Baidu and Alibaba, have so far begun to build their own AI platforms with the aim to extend their industry advantages in the age of AI. Start-ups, such as SenseTime and Megvii, are actively setting up their technology platforms and extending their researches into the basic layer, such as algorithm framework.
Practice issues are solved in the application layer, in which AI provides targeted products, services and solutions to industries with commercialization at the core. Currently, this layer makes up the largest part in China’s AI layers by the sizes and numbers of companies. With an advantage of data, giants set about building their open-source platforms targeted on the application layer; among those start-ups, unicorns represented by SenseTime also expands their presence in the application layer while enhancing R&D efforts in the technology layer.
Figure 4. Four drivers for AI market growth
Computing power Investment
User demandPolicy
Continuous technological advances builds a solid foundation for AI growthIn the past five to ten years, AI technology has been commercialized largely on the back of dramatically enhanced computing power thanks to the integration of improved chip processing capability, accessible cloud services and declining hardware price, such as seniors.
AI is the new engine for the future economic growthSeveral policies have been introduced to support AI since 2015, providing large amounts of project funds for the development and implementation of AI technologies and supporting AI talent introduction and corporate innovation.
Capital brings industry boomThe global investment and financing into AI reached USD39.5 billion from 1,208 financing events in 2017, of which the total amount in China reached USD27.71 billion from 369 events. The proceeds raised by Chinese AI companies accounted for 70% of the global total and the number of financing events took up 31%.
AI are increasingly applied in people’s life and business activitiesThe growth of AI technologies satisfies the demands of businesses, customers and governments on improving living standards, business competitiveness and government efficiencies.
Source: Deloitte Research
Scenarios and Potentials of AI’s Commercial Application in China | China’s AI industry overview
4
Figure 5. China’s AI industrial chain map
Source: Public information, Deloitte Research
Basic layer
Hardware
SensorHesai LeiShen Intelligent System RoboSense SLAMTEC
MediaTek HiSilicon Horizon Robotics Cambricon Intellifusion Westwell Canaan JOHAR Technology
DeePhi Tech CloudMinds Liangzijinrong.com
MOMENTA Uisee Technology ZongMu Technology MINIEYE
DJI HerCamera
YuneecZerotech Xiao-I Robot
Yunwen
Sobot
Toutiao.com Mioji Shangzhuangyuan Sing Palette
Tiantiantou SenseTime Dingfudata NBS Data Technology10jqka.com.cn Whalestock.com Licaimofang CreditX
Appier YunnexPrafly Jixianyuan Zhuge.com
yuantiku.com XiaozhiApogee Tech
Yixue EducationKnowbox
UBTECH roobo Rokid Shenhao
Xiaoyu Zaijia Ecovacs CANBOT
iFLYTEK Mobvoi
Unisound Sinovoice
SoundAI AI Speech
Alibaba Tencent Baidu Xiaomi JD 360 Sogou Cheetah Mobile Huawei
Truking TINAVI Nanochap
Radmedical Jianpei iFLYTEK
BangeyishengHuiyihuiying.com QED Technique
12Sigma Wingspan Infervision
Speech recognition Healthcare
Autonomous driving
Drone Customer service
Others
Personalized push
Finance
Marketing
Education
Warehousing& logistics
Industrial cooperation
Intelligent robot
Semantic recognition and analysis
Machine vision
Others
Comprehensive company
Cloud, data and algorithm
Chip
HardwareSoftware Software Software
Technology layer Application layer
Hard
Hard
Easy
Tricorn Toutiao.comMor.AI Turing Robot Mobvoi
ruyi.ai BosonGTCOM
SenseTime Megvii YITU CloudWalk SeetatechSensingtechTuputech Viscovery VIONVISION PERCEPTIN AuthenMetric
TuSimple BooCax Malong Technologies Emotibot Pinguo ReadSense LINKFACE FaceThink Face all Yi+
4ParadigmSpeakIn
IrisKingHongshi Technology
Linx Libiao Robot Water Rock Technology Geek+
Sublue ocean AUBOTouchnet Technology
Scenarios and Potentials of AI’s Commercial Application in China | China’s AI industry overview
Technology
5
2. Industry-specific commercial application
AI technologies have developed rapidly in the past five to ten years and become widely known over time. The Moore’s law has slowed down and the business application of AI has come into focus. As tech giants are deploying vertical industry applications, start-ups need to identify entries and focus on industry-specific solutions to build their strengths. Deloitte expects to see the booming of AI in government administration, financing, healthcare, automotive, manufacturing and retail
based on the following considerations:
• Does the industry generate large volumes of reliable and stable data?
• Tech giants do not have an absolute advantage over start-ups in access to data.
• Is it a hot sector attractive to investors?
• Can the products and applications properly address industry pain points?
Figure 6. China’s digital government market size forecasting
Source: qianzhan.com, Deloitte Research
2,067
2,3432,722
3,140
0
500
1,000
1,500
2,000
2,500
3,000
3,500
2015 2016 2017 2018E
Scenarios and Potentials of AI’s Commercial Application in China | Industry-specific commercial application
(RMB100 million)
6
Digital governmentAs China is advancing digital governance, government administration becomes a major channel through which AI may establish its presence in use scenarios of smart governance and public security. High entry barriers resulting from local governments imposing strict requirements on providers enable strong players to be stronger. As early adopters have built industry barriers, the rest will need to address the consequent data fragmentation to secure long-term progress.
Digital governance is built on top-down policy initiatives and the market is expected to exceed RMB300 billion by 2018 with a CAGR of 15%.
• Smart governance: The most fundamental and rapidly developing field in smart governance. Local governments across China are
promoting the smart governance by establishing one-stop service platforms. For example, Shenzhen Municipal Public Security Bureau has leveraged facial recognition to simplify household registration procedures and built a citywide system for sharing governance information and resources that pools over 3.8 billion pieces of data in 385 categories of information from 29 organizations2.
• Public security: An AI powered security system functioning intelligently in real-time. For instance, using big data and cognitive intelligence technologies, such system can identify a crime before it is executed and detect other potential risks, turning the focus from post-incident investigation to proactive prediction, early warning and prevention.
Figure 7. “AI+” governance industrial chain
Source: Public information, Deloitte Research
Chip
Alibaba Tencent Baidu HIKVISION Percent Huawei
Sensor
Cloud service
Integrated solutions
• Cambricon • Horizon Robotics • HiSilicon
• Cloudwalk • Megvii • SenseTime
• Ultrapower • MiningLamp
• Inspur • HIKVISION • Dahua Technology
• iFLYTEK • SinoVoice
• Cloudwalk • Hytera
• Turing Robot • Ping An Technology • Alibaba Cloud • NetEase
• Alibaba Cloud • Sugon • Tencent Cloud
Basic layer Technology layer Application layer
Machine vision
Speech recognition
Semantic recognition
Smart governance
Public security
Others
Scenarios and Potentials of AI’s Commercial Application in China | Industry-specific commercial application
2. Understanding the fundamental significance of sharing government services resources through Shenzhen’s experience, Xinhuanet
7
FinanceAI technologies are disrupting pivotal areas of the traditional financial industry. As consumer behavior and needs evolve, traditional players are pressed to reshape various areas and parts of their businesses.
Currently in the finance industry, AI is most widely applied to investment advisory, customer service and risk control.
Figure 8. AI technologies are transforming the whole operational process
Source: Public information, Deloitte Research
Fron
t
Services • Online intelligent customer services
• Customer service robots for bank outlets
• Customized and personalized products
• Intelligent investment advisory services
• Internal risk control • Intelligent office
• Targeted marketing
• Credit rating • Risk-based pricing • Dynamic monitoring
• Data analytics • Proactive data security protection
Product
ICBC’s intelligent customer service robot “Gongxiaozhi” provided over 100 million services in 2017
mjzt.com, an intelligent investment advisory service of China Merchants Bank, has over 150,000 users worth more than RMB10 billion
Ping An Group implements intelligent remote management based on data modeling and visualization
Tencent Cloud employs marketing-related big data generated from Tencent ecosystem for accurate user profiling and labelling and advertises by modelling based on self-developed advantageous advertising algorithm
With an intelligent risk control brain based on massive data, Ant Financial is able to bring Alipay’s asset loss rate below 0.001% and make it a world leader
Tencent collaborates with Beijing Municipal Bureau of Financial Work to develop a Beijing-based financial security big data monitoring platform that identifies, monitors and warns against financial risks
Management
Marketing
Riskcontrol
Data
Mid
dle
Bac
k
Businesses Transformation Cases Outcomes
• Smart investment advisory: expands financial services to groups beyond the traditional wealthy. As an online tool, it can automatically evaluate clients’ financial positions and employ big data analytics to provide customized recommendations, manage investment portfolios and invest in quality products.
• Smart customer service: responds to simple questions using technologies such as natural language processing and knowledge graph, and resolve users’ product or service-related issues through human-computer interaction. In finance, smart customer service is mainly applied to subsectors like banking, insurance and Internet financing.
• Reduce labor costs • Improve service efficiency • Enhance customer experience
• Increase ad conversion rate and reduce marketing costs
• Targeted product pricing • Engage “long-tail” customers and expand businesses
• Reduce risk compensations • Reduce risks of bad debts • Quickly identify financial frauds
• Improve management efficiency and costs
• Improve data security level and lower business risks
Scenarios and Potentials of AI’s Commercial Application in China | Industry-specific commercial application
8
• Smart risk control: using tools like big data and knowledge graph, financial organizations can address traditional problems such as transaction fraud, credit risk management and credit default in a more effective manner. Asset loss rate is a key indicator to measure the risk control capabilities of financial organizations. With smart risk control, Alipay is able to achieve a globally competitive asset loss rate of lower than 0.001%.
Businesses engaging in the AI-driven financial services market mainly come from the sectors of AI, Internet and
finance. With technological advantages, AI companies are able to provide hardware devices or software systems for traditional financial institutions, yet still rely integrated solutions and data acquisition on other players; Internet companies, with strengths in data acquisition and engagement in Internet finance, can offer experience and technical support for traditional financial institutions along their journeys of transformation; traditional financial service providers also start to reshape themselves by building their own technological departments to protect data confidentiality and security.
Figure 9. “AI + finance” industrial chain
Source: 36kr.com, public information, Deloitte Research
Chip
JD.com Tencent
Sensor
Cloud service
Integrated solutions
• Cambricon • NVIDIA • NOVUMIND
• Megvii • Cloudwalk • SenseTime
• Inspur • eGOVA • BOSCH
• iFLYTEK • NUANCE
• Xiao-i • Sobot • Yunwen Technology
• Turing Robot • ruyi.ai
• Tongdun • Rong360
• Ant Financial • 360jie.com
• Alibaba Cloud • Tencent Cloud • JD Cloud
Basic level Technology level Application level
Machine vision
Speech recognition
Semantic recognition
Smart investment advisory
Smart customer service
Smart risk control
• Tianhong Asset Management
• ICBC
• 10jqka.com • CMB
Scenarios and Potentials of AI’s Commercial Application in China | Industry-specific commercial application
9
HealthcareDiagnosis and treatment, medical imaging and health management are the three areas that pioneered the use of AI products in healthcare.
• Smart diagnosis and treatment: applies AI technologies in diagnosis and treatment assistance by making the computer “learn” medical knowledge from medical experts and doctors and imitate how doctors think and make diagnosis, so as to provide reliable diagnosis and treatment plans. Several projects have been implemented in China, most notably solutions like IBM Watson Health.
• Smart medical imaging: AI technologies in this area mainly fall into two categories. The first is image recognition. Applied to the process of sensing, it serves primarily to generate meaningful information
by analyzing images. The second is deep learning, which is used in the processes of learning and analyzing, where large volumes of image and diagnostics data are used to train the deep learning capabilities of the neural network for it to acquire diagnostics skills. In China, several AI healthcare platforms including iFLYTEK and Tencent have completed clinical trials of smart medical imaging in collaboration with medical organizations.
• Health management: Currently, it is primarily applied to risk detection, virtual nursing, mental health, online medical consultation, health intervention and health management based on precise medicine. The emphasis on prevention and conditioning and individualized management is driving health management to become popular in preventive medicine.
AI in healthcare is facing regulatory and technical challenges. From the regulatory aspect, as the healthcare industry deals with people’s life safety, patients’ data must be kept with absolute security and confidentiality and safeguarded under rigorous legal regulations. From the technical perspective, smart healthcare requires troves of data and complex training framework, but few companies have both technology capabilities; the algorithms for combined diagnostics of complex disciplines are shackled by technical bottlenecks; and the technologies and products are markedly homogenous.
Source: Public information, Deloitte Research
Figure 10. Smart healthcare industrial chain
Speech
Smart healthcare industrial chain
Natural language processing
Computer vision
Robot
Machine learning
Speech input, speech to text conversion
Structured and classifiable medical records
Influence preprocessing and extract eigenvalues
Food image recognition for balanced diets
Smart-phone robot,robot guide for patients
DNA sequencing Disease prevention analysis
Drug R&D and screening, side effects forecasting and tracking
Power medical data mining and decisions with machine learning
Technical platform to assist medical researches using AI
Electronic medical record
• Unisound • Huiyihuiying.com • iFLYTEK
Hospital management
• Medbanks • Guangzhou Ruida Medical Instrument
Assisted diagnosis and treatment
• TINAVI • Smarobot • SIASUN
Drug discovery • Cipher Gene • 3DMed • Ribo
Hospital management platform
• Infervision • LinkDoc • Synyi AI
Medical imaging • Infervision • YIDUCLOUD • iFLYTEK
Health management
• More Health • Airdoc • iCarbonX
Disease risk assessment
• Berry Genomics • PrecisionMDX
• BGI
Scenarios and Potentials of AI’s Commercial Application in China | Industry-specific commercial application
10
Smart mobilityIn the era of AI, the value of smart mobility ecosystem which affects the automotive industry is being redefined. The three essentials of mobility, namely “human”, “vehicle” and “road are changing dramatically as AI enables humanlike decisions and behaviors, and the entire ecosystem will also change remarkably. Strong computing power and massive data with high values are at the core of a multidimensional, coordinated mobility ecosystem. As the application of AI technologies in transportation is becoming more intelligent, electrified and shared, an intelligent transportation industrial chain centered on autonomous driving will emerge.
Autonomous driving will be applied to driverless delivery and shared driverless cars:
• Driverless delivery: Once commercialized, driverless heavy trucks will set drivers free and achieve companies’ goals of energy consumption and emission reduction, contributing to global warming control. But no technology comes without limitation. Due to high electricity consumption, it is economically unfeasible to extend battery life during long-distance transports for now. The emergence of driverless cars will give rise to car sharing scenarios and transform the automotive ecosystem.
• Shared driverless cars: Traditional carmakers’ model of selling car ownership will be challenged by service-sharing platforms, self-driving software and the Internet of Things (IoT). The emergence of autonomous driving will lead to the continued expansion of shared automotive products. In the future, car’s role as a personal property will be a thing of the past while the concept of shared transportation rise to take its place. Traditional carmakers who are eager for transformation are thus pressed to accelerate their cooperation with AI technology providers and Internet giants that have massive user data.
Source: Public information, Deloitte Research
Figure 11. Autonomous driving industrial chain
Intelligence
BaiduJDSenseTimeMegviiUISEEIdriverplusAlibabaTencent
BAIC MotorSingulatoBaidu autonomous drivingJD.com autonomous drivingGAC GroupSAIC MotorNIOBYD
DeePhiCambriconHorizon RoboticsWestwellIntellifusion Baidu Map
NavinfoBDStar NavigationAmap
BAIC Singulato Zhongyunzhiche
Beijing Perception Technology Beetech Benewake Qfeeltech heyiabc.com HawkEye Technology
Chip Communication
Internet of Vehicles
Braking System
Camera
High-precision map
Steering system
Lidar
Throttle
Millimeter wave
Brake
Ultrasonic radar
Control
Sensing
Vehicle
Autonomous driving industrial chain
Scenarios and Potentials of AI’s Commercial Application in China | Industry-specific commercial application
11
Manufacturing Combining AI and related technologies can optimize the efficiency along the manufacturing processes, collect a variety of production data through industrial IoT and by applying deep learning algorithms, generate recommendations and even achieve autonomous optimization. Yet compared with finance, commerce and healthcare industries, AI’s application potential in manufacturing has been considerably underestimated. SAP’s analysis of China’s biggest 300 AI investment deals over the three years shows that 23.4% investments were poured into commerce and retailing sectors, 18.3% in autonomous driving while less than 1% went into manufacturing-related AI applications. As a manufacturing power, China’s investment in AI is disproportionate with the vast potential of its manufacturing sector, which is the sector with the most promising AI application scenarios. Adoption of AI may lower processing costs of manufacturers by a maximum of 20%, and 70% of the saving will come from
higher labor productivity3. By 2030, AI will increase global GDP by USD15.7 trillion, of which USD7 trillion will come from China; by 2035, AI will improve labor productivity by 27%, driving manufacturing GDP up to USD27 trillion4.
Three key application scenarios of AI in manufacturing:
• Product: Intelligently research, develop and design products, and inject intelligence into products. These may include generative design, a solution that uses algorithms to explore a multitude of possibilities based on given goals and constraints.
• Production and manufacturing: With data management, incorporation of automated and interconnected equipment, robots can achieve precise coordination of product lines, higher prediction accuracy and real time problem detection through machine learning and analysis. Currently, the main applications include product quality inspection, smart automated sorting,
Figure 12. Smart manufacturing industrial chain
Source: Public information, Deloitte Research
Generative design
Prediction of resources demand and sales
Intelligentproducts Autonomous storage
optimization
• AUTODESK • Rhinoceros • JD.com
• Haier • CASICloud
• AInnovation • SIASUN
• WYSEngine
• Huawei • OPPO • Xiaomi • Vivo • Xiao-i
• Geek+ • Quicktron
• RISEYE • Mech-Mind • NeuroBot • COBOT • Alsontech
Product Manufacturing and production Supply chain
Product quality inspection
Sorting
• Aqrose • Govion Technology • Neptune • Raystrong
Production resources allocation
Production process optimization
Predictive production, operation and maintenance
Scenarios and Potentials of AI’s Commercial Application in China | Industry-specific commercial application
3. Soul of the machine: AI in the factory of the future, 36kr.com, 9 May 2018, https://36kr.com/p/5133138.html4. AI-manufacturing fusion will be the prevailing trend, Communication Information News, 29 August 2018, http://www.fjii.com/jx/2018/0829/179887.shtml
12
predictive production, operation and maintenance, production resources allocation and production process optimization.
• Supply chain: Demand/sales prediction and autonomous storage optimization are the main scenarios. Some businesses use machine learning algorithms to identify demand models by integrating data from storage, Enterprise Resources Program (ERP) and customer insights.
Retail As AI has accelerated the integration of new retail omni-channel, traditional retail businesses team up with start-ups to build application scenarios around customer, goods, store and supply chain.
• Consumer: AI enables demand forecasting, customized marketing, shopping experience improvement and intelligent customer support primarily to achieve continuous and effective consumer engagement. During last year’s “Double 11”, Alibaba’s intelligent recommendation system generated 56.7 billion exclusive shelves for users. A study has found that compared with conventional webpages, personalized webpages uplift conversion ratio by 20%.
• Goods: Applications include assisted payment, stocktaking, promotion and pricing through intelligent shelves.
• Store: siting, in-store shopping experience and unmanned stores that primarily serve to maximize the
Figure 13. Application scenarios of AI and AI-powered application in global manufacturing market
Source: Markets and Markets Insights, Deloitte Research
17%12%
5% 14%
9%
21%15%
13%
0.12%
3%
54%
37%
20%
0%
40%
60%
80%
100%
2016 2025E
■ Industrial robot ■ Manufacturing IoT ■ Manufacturing cloud (public)
■ ■ Manufacturing AI ■ Manufacturing big data and business analysis
Intelligent factory application/solution
Scenarios and Potentials of AI’s Commercial Application in China | Industry-specific commercial application
13
Figure 14. Smart retail industry
benefits of investment into stores. AI-powered siting combines a variety of data including sales records, demo-economic data and distance to competitors to bring the data granularity and relevance of siting model to a new high.
• Supply chain: intelligent pricing, delivery and storage for improved efficiency. AI-powered pricing, based on daily price changes
and corresponding sales data of retailers, integrates user profiles and supplementary external data to automatically generate pricing strategies and recommendations for retail goods through data algorithm models, thus providing relevant strategic suggestions around revenue management goals and ultimately forming intelligent pricing plans.
Source: Public information, Deloitte Research
User profile Store sitingIntelligent shelf Demand prediction
Online shopping experience
Intelligent stores
Smart customer service
Unmanned retail
Intelligent logistics
• Percent • Sensors Data • Dt Dream
• BEHE Adtech Solution
• GeoHey
• ImageDT • Cardinal Operations
• Malong Technologies
• Jirui Tech
• Cloudealk • Keruyun • ML
• Abitai
• DeepBlue Technology
• F5 • Tuzi City
• Youhualin • Linx • inData • Lanxin
Attract consumer engagement
Stock management
Redefine stores
Smart supply chain
Scenarios and Potentials of AI’s Commercial Application in China | Industry-specific commercial application
14
AI technologies have been commercialized and applied in various industries since 2015. The prospect of commercialization has been generally recognized by governments, companies and other all parties. In order to promote the industrial upgrading and transform old economic drivers into new ones, governments have introduced guiding opinions on industrial plans related to
AI and offered tax benefits, subsidies, talent introduction and government process optimization to optimize the business environment, attract strong companies and help develop local AI companies and application markets. Driven by policies and capital, a healthy competition in AI industry among regions has taken shape and become a catalyser in the rapid growth of China’s AI commercialization.
3. Regions’ potentials of commercialization
Figure 15. Distribution map of China’s AI companies
Source: iyiou.com, Deloitte Research
0
50
100
150
200
250
300
350
400368
185
131
95
3925 16 12 10 9 6 6 5 5 3 1 1 1 1 1 1 1
Beijing
Guangd
ong
Shan
ghai
Zhejian
g
Jiangs
u
Sichuan
Fujia
nHubei
Tianjin
Anhui
Liaoning
Shan
dong
Henan
Shan
xi
Chongqing
Guizhou
Tibet
Hong Kong
Guangx
i
Hunan
Heilongji
ang
Shan
xi
Scenarios and Potentials of AI’s Commercial Application in China | Regions’ potentials of commercialization
15
PolicyThe number of policies, policy direction and financial subsidy are three ways of measuring regions’ policy efforts. In terms of the number of AI related policies introduced, Shanghai ranks first, followed by Beijing, Shenzhen, Hangzhou, Guangzhou, and Chongqing. With the support in planning and capital, they leverage innovative technology resources and strengths in talent and market to build AI clusters. From the perspective of policy direction, each of these cities focuses on developing the AI industries respectively based on their own characteristics. Guangdong’s policies focus on applications, while
In the city layer, Beijing, Shenzhen, Shanghai and Hangzhou have the largest numbers of AI companies, all having more than 90. They are in the top tier with the combination of AI and urban development far over other cities, thanks to their advantages in policy, capital, technology, talent and application. Based on the location features of AI companies, the report selects Beijing, Shanghai, Hangzhou, Shenzhen, Guangzhou and Chongqing from Beijing-Tianjin-Hebei Region, Pearl River Delta, Yangtze River Delta and Western China and analyzes these six cities from the perspective of policy, capital, technology, talent and application.
Figure 16. Analysis from five dimensions
Source: Deloitte Research
First-level indicator Second-level indicator
Policy Planning and policy
Capital Investments and proceeds raised
Technology Patent
Research universities and institutions
Number of companies
Computing power
Talent Number
Application Convenience offered by smart city for residents’ living
Operating efficiency improved through urban management
Scenarios and Potentials of AI’s Commercial Application in China | Regions’ potentials of commercialization
16
Beijing places stress on the integration of industry, university and research and technological innovation; Shanghai highlights support for talent, innovation ecosystem, industry cluster and investment and financing; Hangzhou emphasizes AI industry structure and application and Chongqing focuses more on promoting the integration of technologies with the traditional manufacturing industry. In the financial subsidy aspect, it mainly includes project fund, subsidized loan and special industry fund for industry development. Hangzhou set up a RMB3 billion special fund for AI Town; Shenzhen offers project funds (capped at RMB45 million), subsidized loans and intellectual property funds for emerging industries; Beijing primarily
offers technology innovation funds while Shanghai set up a RMB200 million special fund for AI innovations. Chongqing provides a fund of up to RMB10 million for qualified projects.
CapitalDynamic capital environment has a positive impact on helping AI start-ups in technology upgrading, user acquisition and market expansion, facilitating upstream and downstream companies in the AI industrial chain to gain the benefits of scale economy. As start-ups act as the pioneer in the R&D and application of new technologies, the proceeds they raise partly indicate the prospect of the region in developing new technologies.
Figure 17. Proceeds raised by AI start-ups (by city)
Source: iyiou.com, Deloitte Research
0
100
200
300
400
500
600550
350
87
25 14 8
Beijing Shanghai Shenzhen Hangzhou Guangzhou Chongqing
Scenarios and Potentials of AI’s Commercial Application in China | Regions’ potentials of commercialization
(RMB100 million)
17
Technology In terms of technology, we analyze the features of China’s regional technology development from patents, research universities and institutions, number of companies and computing power.
Number of patents: With increasing R&D funds and social capital into China’s AI technology reserves, China is a leader in the number of AI patents. According to the statistics of the number of patents held by major
Figure 18. Number of patents held by research institutions, universities and leading companies (by city)
Note: Incomplete statistics; given key statistics of the most outstanding large Chinese research universities and institutions in AI sector, the number of patents of cities may be different from the total.
0
1,000
2,000
3,000
4,000
5,000
6,000
7,000
8,000
9,0008,183
1,3601,013 739 680
Beijing Hangzhou Guangzhou Shenzhen Shanghai Chongqing
established research universities and institutions, Beijing has the most remarkable scientific achievements with more than 8,000 patents.
Features of universities: China’s AI papers published have outnumbered those of the U.S. and other countries since 2014, largely driven by the rapid growth of Chinese AI research universities and institutions, which are also the main force of AI patent applications.
Scenarios and Potentials of AI’s Commercial Application in China | Regions’ potentials of commercialization
18
Features Universities Labs jointly built by government or research institutions and universities
Companies’ labs
Beijing • Strongest R&D technology capabilities
Take up over 50% of the country
• Peking University
• Tsinghua University
• Beihang University
• Institute of Automation, Chinese Academy of Sciences
Over 10 labs:
• National Laboratory of Pattern Recognition
• State Key Laboratory of Intelligent Technology and Systems
• National Engineering Laboratory for Deep Learning Technology and Applications
• 360
• Xiaomi
• Sinovation Ventures
• Baidu
• Meituan
• Toutiao.com
• Lenovo
• JD
Shanghai • Mainly rely on universities; the number of companies’ research institutes/labs is smaller than Beijing but with certain academic foundation
Lots of universities:
• Fudan University
• Shanghai Jiao Tong University
• Tongji University
• SJTU-Versa Computer Science and AI Joint Lab
• SAIC Motor
• Philips
Shenzhen • Mainly rely on companies • Shenzhen University
• Southern University of Science and Technology
Mainly led by governments:
• Shenzhen Academy of Robotics
• Shenzhen Research Institute of Artificial Intelligence and Big Data
• Tencent
• Huawei
• ZTE
Hangzhou • A certain gap with Beijing, Shanghai and Shenzhen
• Zhejiang University • Alibaba
• NetEase
• Geely Auto
Guangzhou • Mainly rely on iFLYTEK Labs jointly built by universities and companies:
• Cooperation between iFLYTEK and universities, including SCUT-iFLYTEK joint lab for brain-machine cooperation intelligence technology and application
• SCNU-iFLYTEK joint lab for the integration and innovation of industry big data application
Chongqing • Weak in AI technology • Chongqing University
• Chongqing University of Posts and Telecommunications
• Chongqing Institute of Green and Intelligent Technology, Chinese Academy of Sciences
• CloudWalk
• Kaize Technology
Figure 19. Features of AI research institutions and universities (by city)
Source: Public information, Deloitte Research
Scenarios and Potentials of AI’s Commercial Application in China | Regions’ potentials of commercialization
19
Number of AI companies: As of the first half of 2018, there are nearly 5,000 AI companies detected around the world, with China only second to the U.S.5. The numbers of AI companies in Beijing, Shanghai, Shenzhen and Hangzhou rank in the global top 20. Beijing has distinct leading advantages with about 400 companies; Shanghai, Shenzhen and Hangzhou grow fast. Chongqing has few AI companies due to technology and talent constraints.
0
50
100
150
200
250
300
350
400
450395
210
119
6346
5
Beijing Shanghai Shenzhen Hangzhou Guangzhou Chongqing
Figure 20. Number of AI companies by city (as of June 2018)
Source: China’s AI Development Report 2018, Tsinghua University, Deloitte Research
Source: 2018 China AI Computing Power Development Report, Inspur, IDC, Deloitte Research
Computing power: From city distribution perspective, leading cities in China are all located in coastal areas with more advanced AI technologies. Hangzhou, Beijing, Shenzhen and Shanghai rank the top five while Chongqing and Guangzhou rank lowest.
Figure 21. Ranking of computing power (by city)
High
Low
4
1
6 3
5
2 Beijing
Shanghai
Shenzhen
Hangzhou
Guangzhou
Chongqing
Scenarios and Potentials of AI’s Commercial Application in China | Regions’ potentials of commercialization
5. 2018 World AI Industry Development Blue Book, China Academy of Information and Communications Technology and Gartner, 2018-09, http://www.caict.ac.cn/ kxyj/qwfb/bps/201809/P020180918696199759142.pdf
20
TalentChina’s AI talents are distributed unevenly but mainly concentrated in Beijing-Tianjin-Hebei Region, Yangtze River Delta and Pearl River Delta with a certain amount of talents in central and western china along the Yangtze River. Beijing is the most dominant city, accounting for 28%, doubling that of Shanghai (12.1%) in the second place. The proportions of Guangzhou, Shenzhen, Hangzhou and Chongqing are all lower than 10%, sitting in the second tier. This is largely because many strong AI companies concentrate in developed areas with support from governments and social capital, as well as higher compensations for AI talents than other areas.
Application of the citiesFrom the application perspective, the deployment of smart city brings substantial benefits for these cities, including improving the convenience and quality of people’s life. With positive support from governments, Hangzhou leverages Alibaba’s leading experience in AI sector to build a smart city centered on services benefiting people. Beijing has stronger research capabilities and leading AI companies, but the smart services benefiting people are not good as Hangzhou due to many considerations of governments. Cooperating with many local AI companies, such as Huawei and ZTE, Shenzhen provides services benefiting people and achieves certain results. Guangzhou and Shanghai make some progresses in smart life. As the first city launching urban services through WeChat in China, Guangzhou has the largest number of active users while Chongqing, the western city, is lagging behind in creating smart life.
Figure 22. Proportion of number of AI talents (by city)
Source: Global AI Talent White Paper, China’s AI Development Report 2018, Tencent, Deloitte Research
Note: The map presents the numbers of AI talents in these cities with the darker the color is, the more AI talents the city has; bar charts represent the proportion of AI talents equals the number of AI talents one city has/China’s total AI talent
0%
5%
10%
15%
20%
25%
30% 27.9%
12.1%8.5%
6.5%3.9%
1.2%
Beijing Shanghai Shenzhen Hangzhou Guangzhou Chongqing
Figure 23. Ranking of smart life (by city)
Source: Super Smart City, Deloitte Research
12 13 14 15 16
1
2
3
4
5
6
Beijing
Shanghai
Shenzhen
Hangzhou
Guangzhou
Chongqing
Scenarios and Potentials of AI’s Commercial Application in China | Regions’ potentials of commercialization
21
AI is one technology sector where China has the opportunity to set rules. China has a huge application market with a rapid growth of AI industry driven by commercialization. Moreover, research institutions and companies are accelerating their AI related researches and innovations. From regional development perspective, China has built several AI company
clusters in Beijing-Tianjin-Hebei Region, Pearl River Delta, Yangtze River Delta and Sichuan and Chongqing in Western China and built regional industrial parks with policy support. Going forward, China needs to increase attention and investments into long-term basic researches and further improve technology capabilities and industrial chains.
Figure 24. Ranking of smart city administration (by city)
Source: Super Smart City, Deloitte Research
0
2
4
6
8
10
12
14
161
2 3 4
5
6
BeijingShanghaiShenzhen Hangzhou Guangzhou Chongqing
Scenarios and Potentials of AI’s Commercial Application in China | Regions’ potentials of commercialization
Office locationsHarbinRoom 1618, Development Zone Mansion368 Changjiang RoadNangang DistrictHarbin 150090, PRCTel: +86 451 8586 0060Fax: +86 451 8586 0056
HefeiRoom 1201 Tower AHua Bang ICC BuildingNo.190 Qian Shan RoadGovernment and CulturalNew Development DistrictHefei 230601, PRCTel: +86 551 6585 5927Fax: +86 551 6585 5687
Hong Kong35/F One Pacific Place88 QueenswayHong KongTel: +852 2852 1600Fax: +852 2541 1911
JinanUnits 2802-2804, 28/FChina Overseas Plaza OfficeNo. 6636, 2nd Ring South RoadShizhong DistrictJinan 250000, PRCTel: +86 531 8973 5800Fax: +86 531 8973 5811
Macau 19/F The Macau Square Apartment H-N43-53A Av. do Infante D. HenriqueMacauTel: +853 2871 2998Fax: +853 2871 3033
Mongolia15/F, ICC Tower, Jamiyan-Gun Street1st Khoroo, Sukhbaatar District, 14240-0025 Ulaanbaatar, MongoliaTel: +976 7010 0450Fax: +976 7013 0450
Nanjing 6/F Asia Pacific Tower2 Hanzhong RoadXinjiekou SquareNanjing 210005, PRCTel: +86 25 5790 8880Fax: +86 25 8691 8776
Shanghai 30/F Bund Center222 Yan An Road EastShanghai 200002, PRCTel: +86 21 6141 8888Fax: +86 21 6335 0003
ShenyangUnit 3605-3606, Forum 66 Office Tower 1No. 1-1 Qingnian AvenueShenhe DistrictShenyang 110063, PRCTel: +86 24 6785 4068Fax: +86 24 6785 4067
Shenzhen 13/F China Resources Building5001 Shennan Road EastShenzhen 518010, PRCTel: +86 755 8246 3255Fax: +86 755 8246 3186
Suzhou 23/F Building 1Global Wealth Square88 Su Hui Road, Industrial ParkSuzhou 215021, PRCTel: +86 512 6289 1238Fax: +86 512 6762 3338 / 3318
Tianjin 45/F Metropolitan Tower183 Nanjing RoadHeping DistrictTianjin 300051, PRCTel: +86 22 2320 6688Fax: +86 22 8312 6099
Wuhan Unit 1, 49/FNew World International Trade Tower568 Jianshe AvenueWuhan 430000, PRCTel: +86 27 8526 6618Fax: +86 27 8526 7032
Xiamen Unit E, 26/F International Plaza8 Lujiang Road, Siming DistrictXiamen 361001, PRCTel: +86 592 2107 298Fax: +86 592 2107 259
Xi’an Room 5104A, 51F Block AGreenland Center9 Jinye Road, High-tech ZoneXi'an 710065, PRCTel: +86 29 8114 0201Fax: +86 29 8114 0205
Beijing8/F Tower W2The Towers, Beijing Oriental Plaza1 East Chang An AvenueBeijing 100738, PRCTel: +86 10 8520 7788Fax: +86 10 8518 1218
Changsha20/F Tower 3, HC International Plaza No. 109 Furong Road NorthKaifu DistrictChangsha 410008, PRC Tel: +86 731 8522 8790 Fax: +86 731 8522 8230
Chengdu17/F China OverseasInternational Center Block FNo.365 Jiaozi AvenueChengdu 610041, PRCTel: +86 28 6789 8188Fax: +86 28 6317 3500
Chongqing 43/F World Financial Center188 Minzu RoadYuzhong DistrictChongqing 400043, PRCTel: +86 23 8823 1888Fax: +86 23 8857 0978
Dalian15/F Senmao Building147 Zhongshan RoadDalian 116011, PRCTel: +86 411 8371 2888Fax: +86 411 8360 3297
Guangzhou26/F Yuexiu Financial Tower28 Pearl River East RoadGuangzhou 510623, PRCTel: +86 20 8396 9228Fax: +86 20 3888 0121
HangzhouRoom 1206-1210East Building, Central PlazaNo.9 Feiyunjiang RoadShangcheng DistrictHangzhou 310008, PRCTel: +86 571 8972 7688Fax: +86 571 8779 7915 / 8779 7916
23
About Deloitte Global Deloitte refers to one or more of Deloitte Touche Tohmatsu Limited, a UK private company limited by guarantee (“DTTL”), its network of member firms, and their related entities. DTTL and each of its member firms are legally separate and independent entities. DTTL (also referred to as “Deloitte Global”) does not provide services to clients. Please see www.deloitte.com/about to learn more about our global network of member firms.
Deloitte provides audit & assurance, consulting, financial advisory, risk advisory, tax and related services to public and private clients spanning multiple industries. Deloitte serves over 80 percent of the Fortune Global 500® companies through a globally connected network of member firms in more than 150 countries and territories bringing world-class capabilities, insights, and high-quality service to address clients’ most complex business challenges. To learn more about how Deloitte’s approximately 286,000 professionals make an impact that matters, please connect with us on Facebook, LinkedIn, or Twitter.
About Deloitte ChinaThe Deloitte brand first came to China in 1917 when a Deloitte office was opened in Shanghai. Now the Deloitte China network of firms, backed by the global Deloitte network, deliver a full range of audit & assurance, consulting, financial advisory, risk advisory and tax services to local, multinational and growth enterprise clients in China. We have considerable experience in China and have been a significant contributor to the development of China's accounting standards, taxation system and local professional accountants. To learn more about how Deloitte makes an impact that matters in the China marketplace, please connect with our Deloitte China social media platforms via www2.deloitte.com/cn/en/social-media.
This communication contains general information only, and none of Deloitte Touche Tohmatsu Limited, its member firms, or their related entities (collectively the “Deloitte Network”) is by means of this communication, rendering professional advice or services. Before making any decision or taking any action that may affect your finances or your business, you should consult a qualified professional adviser. No entity in the Deloitte Network shall be responsible for any loss whatsoever sustained by any person who relies on this communication. ©2019. For information, contact Deloitte China.CQ-002EN-19