Innovative HUAWEI CLOUD Services

Foreword

 Internet of Things (IoT), big data, and artificial intelligence (AI) are
integrated to create an automation ecosystem. IoT devices collect millions
of standard data records and then process them in the cloud to train and
improve AI algorithms.

1 Huawei Confidential
 Objectives

 Understand the basic concepts of IoT, big data, and AI.

 Be familiar with IoT, big data, and AI cloud service products on HUAWEI
CLOUD.
 Understand the industry trends of IoT, big data, and AI.

2 Huawei Confidential
 Contents

1. IoT Cloud Services and Industry Applications

2. Big Data Cloud Services and Industry Applications

3. AI Cloud Services and Industry Applications

3 Huawei Confidential
What Is IoT?
 The Internet of Things (IoT) is an extended network where billions of physical objects
around the world are connected to the internet. It combines various information sensing
devices with the Internet to form a huge network, implementing interconnection and
interworking between people, machines, and things at any time and any place.
 The International Telecommunication Union (ITU) defines IoT in its ITU Internet Report
as follows:
 A type of network connecting any object with Internet to exchange information and
communicate using predetermined protocols through information sensing devices such as
RFID, infrared sensor, GPS (global positioning system) and laser scanners. The network
implements intelligent identification, positioning, tracing, monitoring and administration.

4 Huawei Confidential
IoT Is Leading Industry Transformation

60% enterprises 24% IT budgets 63% enterprises

are using or are have invested in have obtained

about to deploy IoT IoT services. significant return on
services. investment from IoT
deployment.

Bicycle Smart agriculture Intelligent

sharing Smart homes transportation
Smart locking
Smart appliances Intelligent devices Vehicle management
Better service experience
Value-added Refined production Fleet management
Improved operational
services Smart farming V2X
efficiency

5 Huawei Confidential
Three Challenges Facing the Industry
Let "things" speak in Let "things" say
Let "things" talk one "language" something valuable

Sensing the physical world and Bringing data online and providing Handling massive data
generating digital signals standardized data processing and analysis

Everything perceptible: temperature, space, Wide connection and multi-protocol access Massive data analysis and computing in
and the five senses Data standardization and a unified material real-time
Everything digitalized: sensors and object OS model Industry digitalization

Major challenges Major challenges Major challenges

• Large numbers of existing devices • Lack of standards and specifications for • Compute and storage issues caused by
not digitalized millions of data elements massive amounts of data
• Diversified devices and inconsistent • Diversified data formats • Complex real-time flow, time sequence,
industry protocols and • Siloed development of different offline, and multidimensional analysis
communication networks vendors, resulting in numerous data • Complex industry application scenarios
• Controls on consumption and cost silos and difficult data communication • Business closure, value monetization,
• Security issues and distribution issues

6 Huawei Confidential
HUAWEI CLOUD IoT Strategy: Enabling Industry
Digitalization
Using Intelligence to build full-stack, all-scenario IoT cloud services with enhanced user
experience in the industry ecosystem to drive customers' business success

280 million 7+ million

Industry leader (IDC/IHS)
connections vehicle connections, ranked
No.1 in world IoT ranking
No.1 in the world

All By intelligence For industry

connections

• Hundreds of millions of connections • AI native: Ascend and Kunpeng chips, • Technology enablement: one-stop

+ +
and millions of concurrent connections Atlas edge, and cloud-based EI online development, testing, release,
• Device access in one hour and • Digital twin: digital profiling of and application management
installation and deployment in minutes vehicles, roads, and people • Ecosystem enablement: 50+ industries,
• Solid reliability and security • Edge-cloud synergy: inference on the 3,000+ partners, 400+ certified devices,
• Low power consumption and cost cloud and execution at the edge 1,000+ industry applications, 20+
selected products

7 Huawei Confidential
HUAWEI CLOUD IoT Full-Stack Cloud Services Power
Thousands of Industries
Smart city Smart campus IoV
Smart logistics
Better governance, living, and Energy-saving, efficient, and safe Safe and comfortable travel
Digital warehousing
economy digital campus experience

Fully connected devices and

Urban neural network Enablement of connectivity, Automatic identification of
converged data
Industry transformation and data, ecosystem, and inbound and outbound goods,
All-scenario linkage and automatic stocktaking, and
upgrade evolution
intelligent cloud and edge accurate search

HUAWEI CLOUD Full-Stack Services

Device Edge Cloud

Connection services Data analysis

Industry enablement services

industry alliance
services

IoT security

Ecosystem

AIoT
LiteOS IoT edge Global SIM Link IoT data IoV V2XS IoTC
analysis

Device access Smart logistics IoT Studio

8 Huawei Confidential
LiteOS: E2E Full-Stack Technology

HUAWEI
Cloud CLOUD
IoT 30+ chips 10+ modules 10+ industries

Device

Chip & Smart Smart water Smart Smart More

Module firefighting metering logistics campus

API 2G 3G 4G 5G NB-IoT WiFi

IDE
LiteOS
SDK

OS LiteOS inside

9 Huawei Confidential
Global SIM Link: Making Connections Simpler and
Smarter
Core competitive advantages
Target customers Wide coverage
• New retail 80+ carriers in 100+ countries and regions
• Manufacturing
• Sharing economy Low cost
• Logistics & Devices connected to local networks to use
transportation local data plans; one-stop SIM card
• Public utilities management

Quick access
Pre-integration with mainstream modules of
Chinese vendors to reduce the difficulty of
integrating SIM cards and devices

Intelligent network selection

Remote eSIM/vSIM card provisioning enables on-
demand selection of carrier networks and
automatic number switching

Collaborative ecosystem

10 Huawei Confidential
Device Access Service: Providing All-Scenario,
Pan-Network, and Generic-Protocol Access
IoT applications
Public utilities Smart city Smart campus IoV Smart+ Logistics and warehousing More

Device Access Service HUAWEI CLOUD

products
Device management Message communication Rule engine
Lifecycle mgmt. Device modeling Data parsing Message Data
routing circulation DIS
Device Subscription and
Device shadow Device labeling Custom topics
grouping push Device
Real-time Configuration DMS
Batch operations OTA upgrading File upload collaboration
monitoring and sending
OBS
Device access
Access using native protocols Industry protocol access Generic-protocol access ROMA
(CoAP/MQTT/HTTP) Protocol plug-in Access framework & SDK
Device + edge IoT data analysis

IoT SDK Edge CIG access More

11 Huawei Confidential
IoT Studio - An Application Development Hosting
Platform
Target customers
• Small- and medium-sized enterprises that
Visualization Web App market
components framework
want to quickly build IoT applications
• Enterprises that require report and data
Framework example Smart
Panels generator logistics analysis for their IoT applications but face
high R&D costs
Elevator
Graphs Elevator monitoring IoT application platform
monitoring Quick release of
… … … applications to the App App App
cloud Core competitive
Custom URL,
account allocation HUAWEI advantages
IoT Studio CLOUD Quick construction

Custom reports
Device Device Rule Diversified release modes
access models engines
Device Device Log Multi-level management
groups labels mgmt.
…
IoT capability modules Cost effectiveness

12 Huawei Confidential
IoT Security Solution

Cloud
Platform security and
Pipe privacy protection
Malicious access detection
and interception
Device
Device security capabilities Data lifecycle security
assurance
Resource Industrial Key
sufficiency control services

Host
Cloud security Anit-
intrusion WAF SOC SOC
protection DDoS
detection

Security O&M + security operations + compliance certification

13 Huawei Confidential
IoT Applications in the Industry
Huawei actively participates in and leads the formulation and promotion of IoT standards, and aims
to build a mutually beneficial industry chain and ecosystem with industry partners.
280+
million
devices
connected

3,000+
partners Water Gas Street lamps Logistics Parking Campus

50+
industries

30+
core positions
in standards
Environmental
organizations Firefighting Agriculture Public utilities
protection
IoV Transportation

14 Huawei Confidential
IoT Industry Cloud Helps Mawei (Fuzhou) Lead the IoT
Industry
Mawei (Fuzhou) IoT Industrial Cloud
Huawei has established an IoT cloud computing
innovation center in Mawei District, Fuzhou to attract
IoT enterprises and talent to develop the local IoT
industry. On June 28, 2019, Huawei signed a contract
with the Mawei district government.

Talent cultivation Investment and promotion

Promoting IoT enterprise Introducing IoT Government

products at Huawei development in Mawei assessment and study
Ecosystem Conference at HUAWEI CONNECT on Mawei IoT projects

15 Huawei Confidential
Zhaobangji Works with Huawei to Build Smart Buildings

Customer requirements
• Campus benchmark: Build an
innovative, super-high-rise modern
complex that integrates ecological
HUAWEI CLOUD IoT design and cutting-edge
One-stop visualized
operations technologies. The Zhaobangji
building is located at the gateway
of Qianhai Free Trade Zone in
Shenzhen, China.
• Intelligent: Improve emergency
response and build an
environment-friendly, comfortable,
safe, and innovative campus
through digital and intelligent
management of building
infrastructure.

16 Huawei Confidential
 Contents

1. IoT Cloud Services and Industry Applications

2. Big Data Cloud Services and Industry Applications

3. AI Cloud Services and Industry Applications

17 Huawei Confidential
What Is Big Data?

McKinsey: Big data refers to datasets whose

size is beyond the ability of typical database Volume
software tools to capture, store, manage, and
analyze.

Big Data
Value Variety
4V

National Institute of Standards and

Technology (NIST), USA: Big Data is a term
used to describe the large amount of data in
the networked, digitized, sensor-laden, Velocity
information-driven world.

18 Huawei Confidential
The Digital Era Is Generating Massive Volumes of Data

100+ million Office records on

messages sent on mobile phones
Weibo and WeChat
every day Information and
decision-making Innovative services

Less than 5% data ... ...

value, but requiring Credit
Online
query of
Finance Targeted
investigation: context recommendation
several days to collect 2 to 5 seconds
historical
records
and analyze

Social Data from tens Machine

25+ TB log data
data every day of thousands of data
smart devices

Increased revenue due to rapid information

Diverse data and systems Huge gap acquisition and decision-making

A solution is required to handle distributed storage, parallel computing, and

real-time stream processing of big data.

19 Huawei Confidential
Evolution of the Enterprise-Ready Data Analytics Platform
Types of data Hadoop platform Converged big data
Complex computing SQL, Java, and Python platform
models Various computing engines
(such as MapReduce, graph
Flexible programming APIs computing, and more) Data mining
Conventional data algorithms/models/tools
HDFS, SN
warehouses No support for
complex computing
• I/O SQL, UPF Advanced programming APIs
performance In-memory computing
bottleneck SQL execution engine Rapid, interactive, and
SQL, UPF

management
• Poor real-time query of Batch In-memory Stream
structured data computing computing computing

System
scalability Row-based storage, MicroEngine, concurrency
disks, single node/SE
• Poor real-time cluster Column-based storage,
performance
Unstructured memory ③ ③ ③
data storage

Stream computing
Unified resource
②
management ②
Heterogeneous CQL, rules
streaming data
Real-time processing Mesh operator execution
① Unified data storage
Memory

Single architecture Split architecture Converged architecture

20 Huawei Confidential
HUAWEI CLOUD EI Development Milestones and
Achievements  Core project PMC and committer  Quick acquisition and elastic scaling
 Hadoop Core/HBase: 7  Enhanced reliability and high
 Spark + CarbonData: 8 performance
 CarbonData: top Apache project  Enriched ecosystem

Cloud EI services
Enterprise big data platform
Telecom big (FusionInsight)
data solution Cloud EI services
(Telco industry)
Big data technology Reliable and secure
research self-management

Traditional BI Performance-oriented
and equipment-based
( Telco industry )
Hadoop kernel
optimization and
ETL & analytics community
technology contributions

AI practice: Focusing on internal

AI technology Huawei development and supporting
research intelligent upgrade.
2002 2007 2011 2013 2015 2017
21 Huawei Confidential
Huawei One-Stop Big Data Solution

DAYU: Data Operations Platform Enterprise

applications
Data modeling Data integration Script development Job scheduling Data governance

MRS: One-Stop Big Data Platform Reports,

IoT Data
dashboards
access services
Data Stream query Interactive query Data mining Batch
processing
access Streaming
SparkSQL Presto MLlib Hive DLV OLAP analysis
CQL
Flume
Stream computing Batch processing
DWS
SparkStreaming Flink Storm Spark MapReduce Tez Track mining
Loader
CSS
Parquet TXT ORC CarbonData HBase
Third- Smart assistant
party Kafka
GES
tools HDFS OBS Vision service
...

22 Huawei Confidential
MRS - One-Stop Big Data Platform
Pain points in building a private cloud
big data platform MapReduce Service (MRS) provides enterprise-
class big data clusters on the cloud. Tenants
High
construction
can fully control the clusters and run big data
costs components such as Hadoop, Spark, HBase,
Difficulties in Kafka, and Storm in the clusters.
Unscalable maintenance
resources
HUAWEI CLOUD MRS: an enterprise-class
big data service on the cloud
Insecure;
No DR
capability

Easy Enterprise-
O&M class
Advantages

High Low
No convenient one- Slow service security costs
stop application rollout

23 Huawei Confidential
DAYU - One-Stop Data Operations Platform
• Easy-to-search data • Verifiable data quality
catalogs • Controllable and • Quick buildup of themed
Easy-to- • Clear definitions of data Easy-to- traceable data use Easy- •
databases
standards • Interconnection with 20+ Rapid development of data
find • Full-link visualization control heterogeneous data to-use •
applications
and development sources 10+ industry templates

DAYU
Drag-and-drop GUI

Global data Full-link data Intelligent quality

catalogs development check Data
Data Data consumption
sources integration Data
openness
Real-time
ingestion Users
Intelligent data lake
Offline
data Applications
migration
Data
Third- services
party
Analysis
tools

24 Huawei Confidential
DLV - One-Stop Visualization Development Platform
Data Lake Visualization (DLV) is a one-stop data visualization platform that adapts to various data
sources in the cloud or on-premises. By dragging and dropping 2D and 3D visual components on
DLV, you can quickly customize and create a data screen of your own.

Application
scenarios

25 Huawei Confidential
DWS - Data Analysis Assistant
Data
Data
import to Data analysis Display
sources
databases

DIS
OLAP
analysis
Databases
MRS
Reports

Files CDM Ranked No. 1 cloud data

warehouse in China by the Gartner Ad hoc
Magic Quadrant query
OBS

Internet data Superior performance | Fully Data

parallel architecture, vectorized mining
engine, and intelligent optimizer,
Third-party improving performance by 2 to 6
ETL tools times compared with similar open- Full-text
IoT data source software search

26 Huawei Confidential
Big Data Application - CDM Helps Migrate Data to
HUAWEI CLOUD
Data storage, processing, and
Data sources Data access Service applications
analysis

ETL implemented with

the SparkStreaming
application Industry popularity
analysis
Ad event tracking

Report
Periodic analysis
Crawler data
server load
Kafka collection SQL
Spark/Hive
User interest
Click logs analysis
Data dump/query
One-off
migration of
CDM historical data
Member data

AWS S3
OBS Channel placement
effects analysis

27 Huawei Confidential
Big Data Application - Building a Unified Data Center
Cloud Platform
Internal users External users

Data Comprehensive Financial operations Material operations Financial operations

Precision recommendation More
applications reporting and management and management and management

Intelligent data operations

Data visualization Data governance
DLV DLG
Big data O&M management

Intelligent Data Lake

Big data security

Data analysis Multi-dimensional Spatio-temporal
Search and analysis Memory analysis Graph analysis
analysis analysis
layer MPPDB
Elasticsearch
HBase Redis GES

SQL
Data computing
layer Stream computing (Flink) Batch computing (Spark) AI computing (TensorFlow)

Data storage
Unified storage, compatible with HDFS/OBS APIs
layer

Real-time Data
Structured Unstructured Batch collection Structured Unstructured
collection collection

Data
Enterprise mgmt. Spatial and geographic data Power grid topology Metering External More
sources

28 Huawei Confidential
 Contents

1. IoT Cloud Services and Industry Applications

2. Big Data Cloud Services and Industry Applications

3. AI Cloud Services and Industry Applications

29 Huawei Confidential
What Is AI?
 Artificial Intelligence (AI) is a technical science that studies and develops
theories, methods, technologies, and applications for simulating and extending
human intelligence. This term was first coined by John McCarthy in 1956.
McCarthy defined the subject as the "science and engineering of making
intelligent machines, especially intelligent computer programs". The purpose of
AI is to enable machines to think like people and to make machines intelligent.
Today, AI has become an interdisciplinary domain that involves various fields.

30 Huawei Confidential
10 Changes Bringing an Intelligent Future

Training in days or months Training in minutes or seconds

Scarce & costly computing power Abundant & affordable computing power

Mainly in cloud, some at the edge Pervasive AI for all scenarios. respects and protects user privacy

Algorithms invented before the 1980s Data and energy-efficient, secure, and explainable algorithms

No labor, no intelligence Automated/Semi-automated data labeling

Less practical Industrial-grade AI, more practical

No real-time updates Real-time, closed-loop system

Inadequate integration with other technologies Synergy between AI and cloud, IoT, edge computing, blockchain,
big data, databases...

Applicable only to highly-skilled experts Basic skills supported by one-stop platforms

As Is To Be
Scarcity of data scientists Data scientists + Domain experts + Data science engineers

31 Huawei Confidential
Full-Stack, All-Scenario, and Chip-Powered HUAWEI
CLOUD AI

32 Huawei Confidential
ModelArts Training Flowchart
Online learning

Data Data processing Model training Model training Deployment

Data collection
 Common AI frameworks Cloud real-time
  Model repository 
AI
Data filtering
 Built-in algorithms service

 Model traceback app 1
Data labeling
 Auto hyperparameter

 Precision  Cloud batch service
Version optimization

 Distributed cluster tracking  Device-edge-cloud
management synergy
 Model visualization  Model
 Public datasets
 Auto learning (ExeML) evaluation  Auto hard example
 Multi-person
labeling
 Ascend AI chips  Model mining
 Intelligent data
 Few-shot/semi-supervised conversion
filtering learning
Intelligent data
 Data


augmentation/hyperparam AI market
labeling
 Intelligent data eter/NAS multisearch
analysis
 Graph neural network  Model AI
learning
 Labeling, scoring, transactions app 2
and evaluation
 Reinforcement learning  API transactions
 Federated learning  Secondary fine-
tuning of built-
in models

Notebook + PC SDKs

33 Huawei Confidential
Innovative Data Processing of ModelArts

Intelligent data filtering

Data pre-screening based on unsupervised
learning

Hybrid intelligent labeling

Hybrid intelligent labeling based on semi-
supervised learning/AET/active learning

Data feature analysis

Automatic extraction of distribution
statistics and data diagnosis based on
data features

Labor saved

50% to 80%

34 Huawei Confidential
Huawei HiLens: AI Development and Application
Platform with Device-Cloud Synergy
Huawei HiLens provides extensive device-cloud synergy capabilities such as large-scale device management,
dynamic application deployment, online training, and device model optimization.

HiLens Console HiLens Devices

Skill inference and execution
Device-Cloud Synergy Chip adaptation
Skill Device
development management
Skill Data
market management

1,000x 10% 10x 10+

Higher cost- Higher model Higher Faster business
effectiveness precision performance development
Analyzes data locally, greatly Utilizes the unique device Directly optimizes the device Provides an integrated full-process
reducing the data traffic on the environment data to train models (by splitting, development platform that
cloud. personalized models online. quantizing, and pruning) to seamlessly interconnects with
improve performance while customer data across different
protecting privacy. scenarios.

35 Huawei Confidential
General APIs for Extensive Cognitive Services
Content
Moderation
General OCR Facial Detection
Recognition &
Card OCR Human Analysis Comparison
Service
Visual Retrieval
Receipt OCR OCR Image
Cognition
Domain OCR Recognition
VCC
ImageSearch
Custom OCR VCT
Video Analysis
VCR
VGS

AI
Question Answering
Bot
General APIs
NLP
Task-oriented
Conversational Bot Speech and ASR
CBS
Semantics
Speech Analytics TTS

CBS Customization MTS

36 Huawei Confidential
OCR Digitalizes Enterprise Reports, Bills, and Documents
• Adapts to different
tables
Sets no limits on table size and
analyzes various layouts and
tables for quick electronic
documentation.

• Extracts structured
information
Extracts tables and seals based on text
recognition and location, and constructs the
spatial and semantic relationships of fields
based on semantic understanding.

• Supports cross-row and cross-

column data
Provides an adaptive layout analysis algorithm for
tables with different row and column structures,
including cross-row and cross-column data, and
with/without table lines. Works seamlessly even
when images have multiple elements covering the
text area (such as lines, stamps, etc).

37 Huawei Confidential
Image Intelligence - Freeing Human Resources from
Massive Repetitive Work
Image Recognition ImageSearch Content Moderation

Image Tagging, Celebrity

Copyrighted image search and Pornography, violence, and politics
Recognition, Recapture Detection,
Dark Enhance, and Defog product search

Image tagging: Miqi cakes

Celebrity and content recognition: Copyrighted image search: TukuChina ZBJ.COM
INKE videos Product search: Miqi cakes MeeLive
Dark Enhance: housing transaction
websites

38 Huawei Confidential
Diverse Video Analysis Algorithms Enable Smart Cities
IOC

High-density crowds Fights Protest demonstrations Illegal roadside stalls Smoke detection
conservation
Water

Manually adding a ruler to

identify the water level through
videos

Reservoir water area and level Flood control and emergency

Floating debris detection Water intrusion detection Water gauge identification
alarms response
Transportation

Intersection congestion Detection of motor vehicle parking Detection of non-motor vehicles in

Intersection traffic statistics Heavy-duty vehicle detection
violations motor vehicle lanes
capabilities
General

Detection of lingering Area headcounts and Abandoned object On-duty personnel Work gear and helmet
Facial recognition Vehicle recognition
and intrusions heat maps detection detection detection

39 Huawei Confidential
Customer-defined CBS Makes Smarter Machines
Frontend

CBS Response

Controller To Human

Response
KGQA
LU
IRQA TaskBot
KG
Searcher DM

Searcher
Slot
LG
model
LTR
Answer reRank Answer

Answer

40 Huawei Confidential
HUAWEI CLOUD - Make It Happen!

10+ industries
500+ AI projects

City Manufacturing Healthcare Home-care

Traffic flow optimization Production quality monitoring Pathology analysis Home security protection
Transportation safety/Traffic Sales forecast and production Gene analysis Elderly and children care
trend analysis planning

Campus Logistics Automotive Internet

Passenger flow density Inventory and route Fleet management and Content moderation
analysis management intelligent operation Intelligent video
Abnormal event monitoring Document identification and Driver behavior recommendation
intelligent customs clearance qualifications

41 Huawei Confidential
Railway + AI - Automatic Fault Detection and
Identification
Classifies and resolves faults based on different service requirements

Secure Efficient Precise

Real-time identification of major Manual detection of a large number of minor faults Some faults are difficult to detect
faults that will result in forced stops is time-consuming and labor-intensive. manually.
reduces accidents. AI-based automatic identification improves detection AI-based automatic identification
efficiency. reduces false negatives.

 High identification efficiency: Millions of trains need to be inspected every day, which is labor-intensive. However, AI can significantly improve
identification efficiency.
 Strong model generalization capability: There are thousands of fault types. Manual operations require professional experience. AI can adapt to all
fault types.

42 Huawei Confidential
Healthcare + AI - Giving More Patients Access to
Pathologists

Shortage of pathologists due to

Pathological examination of cervical cancer
difficult training
10,000 pathologists for 1.3 billion people Sensitivity: 99% Specificity: 80%
At least 90,000 additional pathologists needed

43 Huawei Confidential
Empowering Homes with AI
Intelligent care for Intruder Intelligent analysis and Household appliance
the elderly and detection and retrieval of family & food identification
children alarms members from
surveillance videos

Human figure Human figure & Video face tagging

& attribute Item
detection & crying abnormal sound
recognition recognition
recognition detection

Advantages Advantages Advantages Advantages

• Over 95% accuracy; • Over 95% accuracy fast • Device-cloud synergy, • Sustainable training in
fast response in response in device-based face/human a short period, and
milliseconds milliseconds figure recognition, and continuous
• Customizable • Custom optimization cloud-based facial analysis improvement of
vocabulary training algorithm available on • High-performance accuracy
devices of lowest concurrent analysis for
version offline videos, on-demand
GPU resource scheduling,
and cost reduction

44 Huawei Confidential
Protecting Intellectual Property Rights with AI
Adding
watermarks and
changing aspect
ratios
Copyrighted
image Changing
backgrounds,
changing colors,
and cutting

Blurring and
adding
characters

50 million copyrighted images vs. 100+

million images from the Internet
Precise comparison leaves no room for image piracy

45 Huawei Confidential
 Quiz

1. What Are the Types of General Cognitive Services Provided by HUAWEI CLOUD
AI?

A. OCR

B. Sight

C. CBS

D. Speech and semantics

46 Huawei Confidential
 Summary

 This course describes the concepts of IoT, big data, and AI, and helps you
understand the cloud services and industry applications of HUAWEI
CLOUD.

47 Huawei Confidential
 Recommendations

 Huawei iLearning
 https://e.huawei.com/en/talent
 HUAWEI CLOUD Website
 https://www.huaweicloud.com/intl/en-us/
 HUAWEI CLOUD Help Center
 https://support.huaweicloud.com/intl/en-us/index.html

48 Huawei Confidential
Thank you. 把数字世界带入每个人、每个家庭、
每个组织，构建万物互联的智能世界。
Bring digital to every person, home, and
organization for a fully connected,
intelligent world.

Copyright© 2020 Huawei Technologies Co., Ltd.

All Rights Reserved.

The information in this document may contain predictive

statements including, without limitation, statements regarding
the future financial and operating results, future product
portfolio, new technology, etc. There are a number of factors that
could cause actual results and developments to differ materially
from those expressed or implied in the predictive statements.
Therefore, such information is provided for reference purpose
only and constitutes neither an offer nor an acceptance. Huawei
may change the information at any time without notice.