Security Level:

AUTIN
Intelligent Operations Services

Warsaw, 15th October 2019

 Content

1 Operation Challenges

2 AUTIN Introduction

3 AUTIN @Orange Flexible Engine

4 SaaS Model

5 Orange Mali PoC AUTIN Demo

2
2
 Content

1 Operation Challenges

2 AUTIN Introduction

3 AUTIN @Orange Flexible Engine

4 SaaS Model

5 Orange Mali PoC AUTIN Demo

3
3
Inevitable Challenges in Digital Era Request to Reshape the
Operations

4
4
 Operation Model Transformation

Application Operation

Application

Application

Application

Application
DevOps 2

DevOps 3

DevOps n
DevOps
FO App layer
Maintenance
2C 2B … …

1
FO Platform DevOps Team
RAN

VAS
CN

BO
TX

IP

CT VNF FO

TX/IP
VNF Infrastructure Operation
RAN
BO BO NFVi CT VNF VNF

TX/IP
RAN
BO BO NFVi
FME (CT) FME (CT&IT)
FME (CT&IT)
Automation Intelligent

• Operations by domains • VNF: horizontal skill domain • Infrastructure Operation – improving efficiency
• Low level of automation • Improve automation • Application Operation - serving verticals

Future Network requiring Automation & Intelligence to reduce the operational workload, as well as the Organization Transformation,
reducing the FO and FME, focusing and increasing the BO capabilities 5
5
 People Transformation – Digital Workforce

Workflow Orchestration Automation Orchestration

UI Development Engineer
Engineer Engineer

• New Position Resource Orchestration Data Analysis Algorithm Development

25% Engineer Engineer Engineer
• Transformation
BO Engineer
FO
• Retaining • SLA monitoring
• Responsibility • Automation task
Change
15% BO
• Change review monitoring
75%~80%
• Critical incident handling
FO Engineer FM
• Replaced by • Subcontract mgmt.
machine
40%
Preventive Remote Change
Monitoring Scheduling
Maintenance Fix Execution
Field Engineer • Contracting
20%
Transparent
Cost Saving Sharing Model Remote Fault Handling
Subcontract Mgmt.

Future Network Transformation Starts Today by implementing Automation & Intelligence to reduce the operational workload,
and by Transforming the Skills and the Organization
6
6
 Content

1 Operation Challenges

2 AUTIN Introduction

3 AUTIN @Orange Flexible Engine

4 SaaS Model

5 Orange Mali PoC AUTIN Demo

7
7
 Realizing Digitized Operations

Org. Reinvention
• Flat Organization
Skillsets Reinvention
• Agile Decision-making
• Incentive Mechanism
Workflow Task
Automation Automation

Operation AUTomation INtelligence

Engineer Data Algorithm
Modeling Modeling Efficiency & Consistency Proactive & Predictive

1. Task Automation 1. Automatic Alarm Behavior Discovery

2. Workflow Automation 2. On-demand Preventive Maintenance
3. Orchestration through open DevOps 3. Service Impact Prediction
API & 4. Abnormal Fault Detection
Ecosystem

All Online 1. Operation Process Online 4. Anytime, Anywhere with

2. Managed Objects Online Mobility
3. Delivery Resource Online
Measurable & Optimizing

OWS AI Big Data Cloud

Operation Web Services

Unified Digitized Operation Platform (OWS) + 3-STEP Blueprint + Digitized Staff, Organization & Governance
Enable the Journey to Digitized Operations 8
8
All Online
- Massive Data Collection & Modeling from Single Data Platform

Data used by Human Data used by

Machine

Management Operation Mobile Automated Intelligent

Dashboard Monitoring APP Operations Operations

TT CR WO
Data API
FME Car … Data Configur Reporti Impact
Topology Alarm
Modeling ation ng Analysis
Service Service
Events Alarms Log Engineer Service Service

Perf. Capacity …

Data Acquisition Data Storage & Data Processing

Modeling
• Operation Processes Online
• Managed Objects Online
• Delivery Resource Online Unified Data
• Anytime , Anywhere with Mobility
Platform
9
9
 AUTomation
- Effective Orchestration through Open API
Detect Confirm Operate Validate
• Detect to Correct
Workflow
Workflow Automation
• Requirement to Deploy
Automation Engineer • Plan to Prevent
Data Form Automation Data
• Request to Fulfill

Task
Task … … Automation
Automation Engineer Data
Modeling
* Scenario-based API Infrastructure Application
Engineer
Scale-out Scale-out

API
Unified Data Platform

Atomic Topology
Service
Configuration
Service
Alarm
Service
Reporting
Service
Impact
Analysis
Automation
* Atomic API
Module
Reset
Link
Query
Status
Check
Cell
Reset
Parameter
Adjustment
Board
Reset
… Data Data Data
Acquisition Storage & Modeling Processing

MML、MoShell、SSH、Web Service… NBI, SNMP, SFTP, Web Service, …

Mediation

RAN FAN IT SDN NFV Hybrid Cloud Application

10
Legacy Infrastructure Virtualized & Shared Infrastructure
10
 INtelligence
From Automation to Intelligence, from Rule-base to Learning-base

Streaming Learning
Online Batch Learning
Offline Batch Learning • Autonomously provision,
• Root Cause configure and assure the
Identification@AABD network
• Automatic Alarm Correlation • Fault & Risk Prediction
Rules@AABD • Service Impact Analysis
• On-demand Preventive
Maintenance
• Best Workforce Utilization Run
Run

API
Train

Run Train

• Fault Management Automation

Train

• Change Request Automation

• Preventive Maintenance Automation

INtelligence
AUTomation
All online
Rule-base OWS Learning-base
11
11
 Introducing AI to Facilitate Prediction, Reduce Repetition,
and Simplify Operation in the Telecom Industry

Typical Alarm Behavior Auto Root Cause On Demand Intelligent

Risk Prediction
Scenario and Alert Discovery Analysis Proactive scheduling

Model Establishment Offline Training Online Learning Model Application

Intelligent
Patented Algorithms Operation Verified by Expert
Model Model Library
Knowledge Map Real-time Data
Baseline Data Assess Implement Operator Network

Design & Component & failure Production & test Supply chain Network running Trouble Ticket and
Sensor data
engineering data model data data data data repair data
Data
Research & Material & Delivery & Operation & Support
Manufacture Environment
Design Supply Distribution Maintenance Service

12
12
Open Developer Ecosystem

SaaS Ecosystem
Operation APPs for Workflow Task Data Algorithm
different scenarios Automation Automation Modeling Modeling
Engineer Engineer Engineer Engineer

Web APP Mobile APP

(Automation + Intelligence) (Social + Mobility) APP Development
Learning & Training Developer
Support Community

OWS
6,000+ Operation
PaaS Engineers as Developers

APP Development API Development

Platform specifically
designed for ICT Domain
Microservices
Single
Operation Data 20+ Partners

Development
Studio
Environment
Public Platform
Platform
Core
Microservices Data Lake of
Operation Data
DFX
5 Steps for APP Development

No-code & Low-code Drag & Drop Development Auto Deploy

No-code: visual development support WISIWYG “Lego” development Package, online development / update
Low-code: script development

Easy to Integrate Easy to Learn Anytime, Anywhere

Through framework to access of external OWS base develop skill could be With online access, applications
web services and rich API mastered with in a 5-day training are developed anytime, anywhere

Your Your
Idea APP
Services Logic Test/Publish
Analysis
Data Model Page
Services
Design Configuration
Configuration

14
14
 5 Possible Integrated Points between
AUTIN & ONAP
1 Unified Design Time 2 ONAP Events ->AUTIN 3 AUTIN Events >ONAP 4 AUTIN Policy >ONAP 5 AUTIN Controller >ONAP

AUTIN 2、ONAP Events ->AUTIN: Run Time

ONAP
a. When a physical network problem occurs, the events is
transferred from ONAP to the AUTIN to trigger a ticket.
Design Time b. The related information can also be displayed in the portal ONAP Portal
of the AUTIN.
Design Portal SSO AUTIN Portal
1 3、AUTIN Events ->ONAP : ICTOM ICTSM Adaption 2
URL Events APIs（Alarms, Performance ect.） of related physical
Alarm management
network willPolicy
transferred from External API
4、 AUTIN ->ONAP : AUTIN to the
Monitor ONAP for necessary
& Report API
analysis or diagnosis of cloud (Alarm,
network. KPI, Counter...)
Policies of the AUTIN can be uniformly scheduled by ONAP(can WFM
be OPEN API or converted into the policy format in the ONAP ). Performance
Management API
Online Studio 1、Unified
However, usageDesign
theController Time (running
and->ONAP
engine : status) of a specific 3
SDC 5、 AUTIN Analytics
: are logically
policy The
are SDC and
running ononline
the studio
AUTIN. unified to ensure unified DCAE Policy SO A&AI
The interface is opened(SSO,
user experience through the portal).
Unified OPEN API (restful), The
However,
AABD RCI Predictive these functionsService Problem
ONAPareprovides
independent of each other and complementthe
the API Gateway. The AUTIN releases APIs
each other.
Management API
and registers them
The Online with the
Studio API Gateway.
is developed based on processes and SDM
Web 4
page for new APPs. The SDCMCP provides service modeling and
(Controllers) Policy API
Orchestration based on service models. Controllers
5 SDN-C APP-C
Controller API
Mediation

Alarms Performance

Physical Network SDN-NFV Cloud

15
15
 Content

1 Operation Challenges

2 AUTIN Introduction

3 AUTIN @Orange Flexible Engine

4 SaaS Model

5 Orange Mali PoC AUTIN Demo

16
16
 AUTIN Deployed @ Orange Flexible Engine
1 Flexible Engine
AUTIN R&D & Operations
2 Globally Unified Operation
Service Service Service AUTIN Global Unified Operation • Unified Monitoring, Unified
Cluster Cluster Cluster Change, Unified App
AUTIN (OML) (GNOCa) … (GNOCe)
R&D Maintenance Support
upgrade and update
Release Availability …

Orange Group 1
OCEANE SWAN VC4
AUTIN available on Flexible Engine (FE)
• Separate Tenant in the cloud
Independent service Cluster and independent DB
Mediation Local tools Mediation Local tools Mediation Local tools
• Secured and controlled access management (OCD, ISO
NBI NBI NBI
27018)
EMS
EMS EMS … Huawei Ericsson • GDPR Compliancy
Huawei Ericsson Huawei Ericsson
…
• Achieve CSA STAR & ISO 27001
Nokia … Nokia … Nokia

OML (Tenant 1) GNOCa (Tenant 2) GNOCe (Tenant 3)

17
AUTIN Fast & Secure Deployment on Orange Flexible Engine 17
 Fast Deployment on Orange Flexible Engine

Traditional OSS (~4.5 Months)

(~2 Months) (~1 Month) (~2 Weeks) (~2 Weeks) (~2 Weeks)
Hardware Local
Apply Approve Procurement Shipping Installation Testing Trial Online
• At least 4.5 Months
XX
XX
to Deploy & Ready
XX • High CAPEX

-Hardware -Function Test

-Software -Data Configuration
-Network

AUTIN on Flexible Engine (~ Weeks)

(~1 Hour) (~1 Week)
Auto Tenant
Apply Approve Creation Testing Trial Online
XX No Only few weeks &
Hardware No
XX
XX
Procurem Shipping Low/No CAPEX
ent

-Hardware -Data Configuration

-Software
-Network

18
18
 Proactive O&M with AUTIN

Traditional OSS ( Passive Troubleshooting ) Pain Point:

• Passive fault response, "dozens" failures

Root Cause Analysis feedback per customer per year
• R&D manual analysis of problems and
Fault User R&D Reproduce repairing faults at "hourly levels”
Get Logs RCA Resolution Resolve Verify Close
Occur Feedback Analyze Fault • Low-availability solution, service
unavailable when recovery, availability is 99%
Resolve fault with • On-premises O&M, high cost
low--availability architecture

Cloud ( Proactive Risk Detection & Recovery ) Value:

Auto Risk Detection Auto Risk Notification Proactive Risk Recovery • Proactive risk detection & recovery,
“several” failures feedback per customer per
App Call Rapid Rapid Problem year (“hundreds” risk proactive recoveried)
App Server Isolation
Chain • O&M semi-automation analysis of
Dynatrace Replace Expansion
problems and repairing risk at “minutes levels”
APaaS Fault
Error • High-availability solutions, no services
Code (Hidden
risks)
affect, availability is 99.97%
PaaS ELK High Organization • Cloud O&M, low cost
Availability &
Alarms Architecture Mechanism
IaaS（CPU/Disk）
CloudWatch 19
19
 Rapid Upgrade without Perception

Traditional OSS
Pain Point:
Test Environment
“Manually” “Manually” Production Env
“Local”+“Manually” Finish • Yealy version release
Project A Download Verify “Manually” Upgrade
Upgrade • “Manually” + “Local” upgrade for
Version Release
… … … … …
(yearly) every projects
Test Environment
Project X “Manually” “Manually” Production Env • Business affected ~2 hours
“Local”+“Manually” Finish
Download Verify “Manually” Upgrade
Upgrade
Patch Release
(Weekly)

Value:
AUTIN Cloud
• Monthly version release
• “Automation”+“Remote” + “Gray”
Automation
Version Release Small-scale Automation + Remote + Gray Automation & Global upgrade for small-scale projects
Production Env Finish
(Monthly) Projects Upgrade Multi Tenant Test Projects
Upgrade • No business affected
• Automation & Multi Tenant Test
• Global project upgrade one time`
Patch Release
(Daily)

20
20
 AUTIN End to End Security Management

Certificated ISO 27001, ISO 27018*, CSA-STAR, SOC2**

Outside Production Operation Inside

Network Protection

Hacker DDoS

AUTIN Follow GDPR, Audit

Huawei Response
Web
Https/SSL Application SOC -
Web
Firewall Security Monitoring
App Insider
AUTIN security
Cloud security

Multi-tenant Operation
Mgmt OWS-IAM Center Operational
Firewall DB Anti-Virus

Network Cloud-IAM Developing

Encryption
Attack Access Mgmt
IPS / IDS Code Audit
by OBS Data Right Mgmt
Penetration testing
Resource
VPN Mgmt
by OBS Weakness scanning
Jump server

*Planning get certificated 2018

**Planning get certificated 2019 Recovery – Backup, Restore, HA, DR, Multi-Zone, Multi–Data center 21
21
 Deploy Multi-edge Security to Prevent Malicious Attacks

Design Develop Compliance Protect

Application Attack
Defense

Various compliance checks 24-hour monitoring to protect

Risk and threat assessment Code standards and perform security before go-live: AUTIN against attacks.
 Service Requirements  Code Audit  APP Scan  WAF, NGFW, PCS
 Huawei Security Redline  Coding Specifications  xDefend scanning  Forcible audit log
 Best Practices of Security
 SecureCAT  Alarm monitoring
Network Attack Defense

Firewall IPS WAF

AUTIN
Normal connection Firewall: Only allow opened service, or connection
to reduce the attack
Unauthorized access
Intrusion Prevention System (IPS) : By monitoring
Viruses, Trojan horses, and worms and analyzing the behavior and characteristics of
data transmission.
Penetration attack

XSS/CSRF attack Web Application Firewall (WAF) : Provides input

verification, exception detection, status management,
SQL injection attack and audit functions to prevent web-based attacks.

22
22
 Multiple Methods and Level Data Protection

Data transmission security plain text when used

Transmission over HTTPS

Credential (cipher):
Credential (plain): **********
Data Encryption

Username: staff1 **********

Credential (plain):
Password: Pwd_123 Username: staff1
Protection

Password: Pwd_123

Firewall Proxy server Firewall Application server Firewall

Encryption Algorithm Selection Data storage encryption Store over AES/SHA

The AUTIN uses the strong encryption algorithm.

Credential (cipher):
**********
AES128/256 SHA256 RSA2048 ***********
Data is stored only in the database..

FW IPS WAF Authentication Audit Encryption Backup Based on ACL/SG, only necessary services are
Firewall open.
DDoS, virus, worm, Trojan, sniffing, port scanning

In-depth data
IPS protection
Data Life cycle

protection
Management

WAF XSS/CSRF, SQL injection, input validity check

Authentication and access control based on

Authen RBAC model to prevent unauthorized access
tication
Audit Traceable audit logs

Encryption AES/SHA reversible/irreversible encryption

Backup Multi-site disaster recovery data center

Comprehensive data protection: Multiple methods and multi-level data lifecycle 23

23
 AUTIN Ready for Integration with Legacy OSS
 Clarification done with Affiliates during ENO2 due diligence

Conclusion 1 – AUTIN Readiness Conclusion 2 – No Hidden Cost

AUTIN Solution is feasible for Integration with Orange OSS No Hidden cost will remain, with acceptable additional effort from Orange

AUTIN will be ready with No hidden cost. Fast benefits bringing Efficiency & Quality.
Transparency and Visibility, through online & real-time dashboards 24
24
 Content

1 Operation Challenges

2 AUTIN Introduction

3 AUTIN @Orange Flexible Engine

4 SaaS Model

5 Orange Mali PoC AUTIN Demo

25
25
 Selected AUTIN Use Cases for ENO 2
Use Case: Benefits for: Quantified Best Cases impact:

UC1 : AABD – Alarm Correlation FO, BO 94% Alarm Compression

Efficiency & Automation Intelligence
Consistency UC3: Auto UC2: Auto TT Creation FO 98% Auto TT Creation
UC2: Auto TT
WO UC 1 : AABD
UC5: Auto UC3: Auto WO Creation FO, FLM 90% Auto WO Creation
UC6: Auto PM
Diagnose UC 4: RCI
UC12: Auto Proactive & UC4: Root Cause Identification FO, BO, FLM 85% RCI Coverage
UC7: Auto CR
Reporting UC 10 : Smart Predictive
Preventive UC5: Auto Diagnose FO, BO 80% Auto Diagnose
DevOps &
Measurable & UC6: Auto Proactive Maintenance FO, BO 90% Auto PM Coverage
Optimizing All Online Ecosystem
UC7: Auto Change Request Mgmt FO, BO 40% Auto Change Request

UC11: Mgmt. UC8: Auto scheduling FO, FLM 100% Scheduling Coverage
UC9: Alarm query UC8: Auto scheduling
hierarchy & analytics
UC9: Alarm query FO, FLM 100% Transparency

UC 10 : Smart Preventive FLM 10% Incident Prevention

UC11: Mgmt. hierarchy & analytics FO, BO, FLM 100% Visibility

OWS Cloud Big Data AI UC12: Auto Reporting FO, BO, FLM 70% Auto Report

Proven and Mature Use Cases adapted to ENO2 context. Faster roll-out and Better Quality Improvement. Additional Use Cases
26
through the 5 years with Orange specific use cases to be developed by the DevOps (together) 26
 AUTIN SaaS Model Suggestion (based on ENO2 Proposal)

AUTIN SaaS Scope AUTIN Support Scope

Training & Operational Support

Efficiency & Automation Intelligence
Consistency
UC2: Auto TT UC3: Auto WO Classroom Training
UC 1 : AABD
UC5: Auto AUTIN Function Introduction
UC6: Auto PM
Diagnose Use case Training
UC 4: RCI
UC12: Auto Proactive & AUTIN Developer Training
UC7: Auto CR Predictive
Reporting
UC 10 : Smart Preventive

DevOps & Operational Support

Measurable &
Optimizing All Online Ecosystem
First 2 months after Implementation

UC11: Mgmt. hierarchy &

UC9: Alarm query UC8: Auto scheduling Technical Support
analytics

24/7 First Line Helpdesk Support

Remote CSR handling

OI@OWS SDM@OWS WFM@OWS FM@OWS AABD@OWS Query @OWS Platform Release Upgrade 27
27
 AUTIN SaaS – Training & Operational Support

Classroom Training
Grade / Training Duration Number of Max. Number of
Program and Course
Level * Location (Working Days) Classes Trainees per Class

AUTIN Function Introduction L1-L2 Customer premises 2 2 12

Use case Training L1-L2 Customer premises 3 2 12
AUTIN Developer Training L2-L3 Customer premises 4 1 12

Operational Support
Duration
Operational Support Engineer Grade / Level *
(Working Days)

AUTIN engineer L3 40
AUTIN engineer L2 80

* Grade / Level:
• L1: Junior Front or Back Office Operator, Junior OSS Engineer, up to 3 years of Experience
• L2: Front or Back Office Operator, OSS Engineer, 3 to 5 years of Experience
• L3: Back Office Expert, OSS Expert, more than 5 years of Experience 28
28
 AUTIN SaaS –Technical Support Service
Category Description
Support Availability 24/7
Support Channels iCare hotline, iCare self-service website
Basic Monitoring Basic Infrastructure and Application Monitoring
Service Communities Service forums, Service documentation, Whitepapers
Service Scope SaaS Service Information, Service Request, remote problem Solving, Platform Release Upgrade
Priority SLA
All key functionalities are unavailable to the users and the incident need to be Response Time 30 minutes
P1 handled immediately. This condition is generally characterized by complete
Restoration Time 4 hours
system failure and requires immediate correction
Response Time One or several key functionalities are severely affected to the users and the Response Time 60 minutes
P2
& Restoration problem need to be handle ASAP Restoration Time 12 hours
time Response Time 2 hours
P3 Non-critical functionalities are affected
Restoration Time 5 days
Response Time 2 days
P4 Consultation inquiries submitted by the user during the service
Restoration Time 8 days
Availability 99.96%
29
SLA >90%
29
 AUTIN SaaS Responsibility Matrix
Matrix
No. Module Item
Huawei Orange
1 Confirm scope of AUTIN and Kick off R R
2 UAT Testing and Approval R A
3 OWS test bed preparation for software upgrade R
4 SaaS Joint test cooperation for integrated 3rd party OSS/EMS software upgrade R S
5 Provide link between collector server and OWS, collector and EMS S R
6 Provide IaaS （including hardware ,OS ,etc.）for local collector server S R
7 NBI license availability for integrated EMS S R
8 Provide 7 x 24 hours technical remote support services R
Respond to query/problem requests submitted by customer within the specified time;
9 R A
provide Restore & Restore Solution
Technical Support If emergency problem, trigger Escalation Process, create War Room, involve related
10 R S
support teams to analysis root cause and restoration problem
11 OWS software update/upgrade R
12 Coordination of 3rd party OSS/EMS support for trouble shooting S R
13 Provide classroom and IT premises S R
Training &
14 Identify trainee and confirm training plan R A
Operational
15 Prepare training materials and deliver training R S
Support
16 Provide office seat for Huawei Operational Support OWS engineer S R 30
R = Responsibility A = Approval S = Support 30
 Content

1 Operation Challenges

2 AUTIN Introduction

3 AUTIN @Orange Flexible Engine

4 SaaS Model

5 Orange Mali PoC AUTIN Demo

31
31
Our Automation Journey with Orange 1/2

2018.02
AUTIN Launched
during MWC 2018 2018.04
Romania GNOC
Visit and AUTIN
Workshop

2018.06

MS Summit 2018
Paris 2018.09
3GIS Visit and 5G
Operation Workshop

Mali PoC kick off

2019.03

China and India GNOC

Visit
32
32
Our Automation Journey with Orange 2/2

2019.04
Mali AUTIN PoC Approved
2019.05
OFR AUTIN PoC Initiation
AUTOMATION
Workshop with Orange
Tunis

2019.06
2019.07 MS Summit 2019
Casablanca
AUTIN @ xingine
VO 5G Operation
AUTIN GNOC Workshop with OFR
Implementation Workshop

OFR AUTIN PoC Kickoff

We are Here
33
Open a new page! 33
 Orange Mali Operation Challenges
Traditional Operation Digitized Operation

FAN
IT
RAN

VS.
NFV

SDN

Cloud Application

Technology Complexity

Χ Manual operation management is terrible with complex network  All online management system, everything is visible & controllable
Χ Too many alarms appear at same time in different systems,  SLA & OLA real time overview make the network quality
difficult to identify in a short time the right Hub site during measurable, optimizing & help to control the TTR
incident  Good level of digital visibility including detailed Sites outage view
Χ Many Screens to monitor with too many alarms by technologies, Sectors Outages,
 Sites down, WO (PM, PLM & CM ) task overview anywhere, anytime.
Χ Too much information to fill in the TT during big outage
 Transparent Network with real time monitoring overview
Χ The network quality depends on people
 Incident Prediction, auto diagnosis and recovering
Χ Manual report take too much time, can generate mistakes
 Field management efficiency
Χ No visible Process
34
34
 Implemented Use Cases
We started Here Usage started POC Evaluation

We started Here
APR’18 SEP’18 APR’19

AUTIN Delivery UC 4
Solution prepared, Integration of Integration of Integration Integration of UC 1, 2 on UC 3,4,5 on UC 6,7,8 on
submitted & WFM OI ICTOM live live live Enhancement Outage Heat
approved
of SDM
Map

Alarms Outage Heat Map for Operational Management 0  90%

RAN
AABD FM FM SDM WFM WFM
Transmission
… Alarm
Core Auto Auto TT Auto WO Intelligent WO
IT Correlation Diagnosis Creation Creation Scheduling Dispatching
Success
~30%  92% 0%  98% Rate 0%  25% 0%  100% 0%  100% Manual
~190,000/day 133k  15k 0  ~850 0  30/~120 0  30/30 (60) 0  60/60

UC 1 UC 3 UC 2 UC 5 UC 5 UC 5
Alarm Correlation Auto Diagnosis Auto TT Intelligent Scheduling Intelligent Scheduling Intelligent Scheduling

Consumable Management 0%  100%

Preventive Maintenance 0  54 UC 6
Consumable
Management
RPA (Robotic Process Automation) for Automated Reporting 0  2 UC 8
Auto Preventive
UC 7 Maintenance
35
Auto Reporting 35
 Orange Mali AUTIN PoC Architecture

OI Digital Report Orange

AUTIN Mali
ICTOM SDM WFM
Fault Mgmt Process Mgmt Workforce Mgmt
OCEANE

SWAN
Alarms

Local Collector Servers with Multivendor Probes

EMS U2000_RAN U2000_TX NetAct U2000_IP NSP (IP & TX) AMS
In Progress

Done
VENDOR HUAWEI HUAWEI NOKIA HUAWEI NOKIA NOKIA
Ongoing 36
36
 Overall Benefits : AUTIN Impacted positively
MALI RAN NUR
NUR VALUE BEFORE AND AFTER AUTIN Benefits (August’18 compared to April’19)
TT Creation
Technology August'18 NUR April'19 NUR % Decrease
TT Creation time reduced min Before After
2G 920 490 47% by >90% P1 28.74 9.06

3G 565 519 10% P2 25.13 11.15

P3 24.29 10.76
4G 64 55 15%
P4 297.12 10.20

02G NUR analysis improvement analysis MTTR

hrs Before After
 98% MW link normalized3%
 SETOGA Traffic IP Swap and MSTP deactivation which fixed TN issue4% Reduction MTTR >50% P1 3.45 1.89

 Field team sensibilization and field repartition optimization2% P2 5.36 1.65

 Solar sites Hybridation and GE replacement in the north area6% P3 8.55 2.13
 Batteries faulty replacement for some EDM sites3% P4 11.00 3.44
 Follow the GE SNAGS clearance. 4%
WO Life Cycle
 Batteries Boost for solar sites with Mobile rectifier/GE8%
hrs Before After
 AUTIN usage : Operation Efficiency Improvement17%
WO Life Cycle > 80% P1 22:57 06:20
P2 26:01 8:49
We can consider that AUTIN decreased approximatively around P3 27:05 19:27
17% the NUR value from September’2018 to April’2019 P4 46:58 11:03
37
37
Orange Mali AUTIN PoC Conclusions

38
38
Thank You.
Copyright©2016 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.
 Implemented Use Case Review – Alarm Correlation

Outage Heat Map for Operational Management

Alarms
RAN
AABD FM FM SDM WFM WFM
Transmission
… Alarm
Core Auto Auto TT Auto WO Intelligent WO
IT Correlation Diagnosis Creation Creation Scheduling Dispatching
~30%  92%
~190,000/day 133k  15k

• Starting Value: ~30% => Current Value: 92% (for covered Network)
UC 1
Alarm Correlation • Target Value: All domains > 90%
TT Creation
min Before After
• Value :
• Faster Demarcation through Automation P1 28.74
Consumable 6.24
Management
• Reduce the Front Office Workload
P2 25.13 6.34

• Achievements: Preventive Maintenance P3 24.29 6.28

• 680 Rules already Online P4 297.12 10.30
RPA (Robotic Process •Automation)
TT Creationfor
time reduced byReporting
Automated >90%
• From 3 Monitoring & TT screens => 1 monitoring & TT screen
 Implemented Use Case Review – Auto Diagnosis

Outage Heat Map for Operational Management

Alarms
RAN
AABD FM FM SDM WFM WFM
Transmission
… Alarm
Core Auto Auto TT Auto WO Intelligent WO
IT Correlation Diagnosis Creation Creation Scheduling Dispatching
Success
~30%  92% 0%  98% Rate

~190,000/day 133k  15k 0  ~850

UC 1 UC 3 • Starting Value: 0% => Current Value: 98% Success Rate, 14 rules / 850 Active Alarms
Alarm Correlation Auto Diagnosis • Target Value: 90% of the Active Alarms

• Value :
• Faster Demarcation through
Consumable
SupportManagement
for FO Competence & Action
• Reduce the Front Office Workload
Preventive Maintenance
•
Achievements:
• Support the FO/L1 Fault Management with an online and automated Knowledge Mgmt
• TT
RPA (Robotic Process Automation) for Automated Reporting
content Accuracy & Quality
• Contribute on improving the TT Creation & MTTR
 Implemented Use Case Review – Auto TT

Outage Heat Map for Operational Management

Alarms
RAN
AABD FM FM SDM WFM WFM
Transmission
… Alarm
Core Auto Auto TT Auto WO Intelligent WO
IT Correlation Diagnosis Creation Creation Scheduling Dispatching
Success
~30%  92% 0%  98% Rate 0%  25%
~190,000/day 133k  15k 0  ~850 0  30/~120

UC 1 UC 3 UC 2 • Starting Value: 0% => Current Value: 25% (All Passive is automated),

Alarm Correlation Auto Diagnosis Auto TT 30 Auto TT out of an average 120 TT per day

• Target Value: 50% Auto TT for Active & Passive MTTR

Consumable Management hrs Before After
•
Value :
P1 3.45 2.73
• Faster action through Automation
Preventive Maintenance
• Reduce the Front Office Workload P2 5.36 2.86

P3 8.55 3.57
RPA (Robotic Process Automation) for Automated Reporting• Achievements:
• Reduction in TT Creation / MTTR >40% P4 11.00 7.63
• Improve TT Quality:
Last Two weeks: 0 complaints from Orange Mali about TT quality
• Avoid missing alarms (<1%)
 Implemented Use Case Review – Intelligent Scheduling

Outage Heat Map for Operational Management

Alarms
RAN
AABD FM FM SDM WFM WFM
Transmission
… Alarm
Core Auto Auto TT Auto WO Intelligent WO
IT Correlation Diagnosis Creation Creation Scheduling Dispatching
Success
~30%  92% 0%  98% Rate 0%  25% 0%  100% 0%  100% Manual
~190,000/day 133k  15k 0  ~850 0  30/~120 0  30/30 (60) 0  60/60

UC 1 Value: 100% Auto WO, UC 2>90%

UC 3100% Auto Scheduling, UC 5 UC 5 UC 5
• Starting Value: 0% => Current Intelligent Scheduling
Auto Diagnosis Auto TT Intelligent Scheduling Intelligent Scheduling
Auto Dispatching Alarm Correlation

• Target Value: 100% Auto WO, 100% Auto Scheduling, 100% Auto Dispatching
Consumable Management
WO Life Cycle
• Value :
hrs Before After
• Contribute to MTTR reduction Preventive Maintenance
• Better FME Operational Management P1 22:57 4:47

RPA (Robotic Process Automation)

P2 for Automated
26:01 5:27Reporting
• Achievements:
• Reduction in WO Life Cycle > 80% P3 27:05 7:46
• Enhance FME Efficiency P4 46:58 4:53
 Implemented Use Case Review – Auto Reporting

Outage Heat Map for Operational Management

Alarms
RAN
AABD FM FM SDM WFM WFM
Transmission
… Alarm
Core Auto Auto TT Auto WO Intelligent WO
IT Correlation Diagnosis Creation Creation Scheduling Dispatching
Success
~30%  92% 0%  98% Rate 0%  25% 0%  100% 0%  100% Manual
0 => 2 NPM
• Starting Value:133k
~190,000/day 0  ~850
15k auto-reporting 0  30/~120 0  30/30 (60) 0  60/60
• Target Value: Auto Reporting for NPM = 6, FO = 5 and BO = 5

• Value : UC 1 UC 3 UC 2 UC 5 UC 5 UC 5
• Reduce substantial work load
Alarm Correlation Auto Diagnosis Auto TT Intelligent Scheduling Intelligent Scheduling Intelligent Scheduling
• Ensure correctness of reports.
• Ensures timely report generations (24*365)
Consumable Management 0%  100%
• Achievements:
• 2 Auto Reporting for NPM as a Pilot Preventive Maintenance 0  54 UC 6
Consumable
Management
RPA (Robotic Process Automation) for Automated Reporting 0  2 UC 8
Auto Preventive
UC 7 Maintenance
Auto Reporting
 Implemented Use Case Review – Auto Preventive Maintenance

Outage Heat Map for Operational Management

Alarms
RAN
AABD FM FM SDM WFM WFM
Transmission
… Alarm
Core Auto Auto TT Auto WO Intelligent WO
• Starting Value:
Correlation0% => 54 Rules (all Huawei)
Diagnosis Creation Creation Scheduling Dispatching
IT
Success
• ~30%
Target  Auto
Value: 92%PM Rate >0%  98%
90% Rate 0%  25% 0%  100% 0%  100% Manual
~190,000/day 133k  15k 0  ~850 0  30/~120 0  30/30 (60) 0  60/60
• Value :
• Reduce substantial work load in BO domain
• No missing schedules of routine PM
UC 1 UC 3 UC 2 UC 5 UC 5 UC 5
Alarm Correlation Auto Diagnosis Auto TT Intelligent Scheduling Intelligent Scheduling Intelligent Scheduling
• Achievements:
• Daily Auto PM for Huawei covered Network
• Daily report Generated, with Actions in BO Consumable Management 0%  100%

Preventive Maintenance 0  54 UC 6
Consumable
Management
RPA (Robotic Process Automation) for Automated Reporting UC 8
Auto Preventive
Maintenance
 Implemented Use Case Review – Consumable Management

Outage Heat Map for Operational Management

Alarms
RAN
AABD FM FM SDM WFM WFM
Transmission
…
Core
Alarm • Starting Value: 0% => Current Value: 100% of Consumable
Auto Auto TT Auto WO Intelligent WO
Correlation Diagnosis
IT • Target Value: 100% of Consumable Creation Creation Scheduling Dispatching
Success
~30%  92% 0%  98% Rate 0%  25% 0%  100% 0%  100% Manual
• Value :
~190,000/day 133k  15k • Controls0 Stock
 ~850  30/~120
0 each
of Consumable in 0  30/30 (60)
GIS Based Warehouse 0  60/60

• Achievements:
UC 1 UC 3 On-line UC 2 UC 5 UC 5 UC 5
• All consumables Intelligent Scheduling
Alarm Correlation • Auto Diagnosis Auto TT Intelligent Scheduling Intelligent Scheduling
Zero Delay in the Outbound process through online validation

Consumable Management 0%  100%

Preventive Maintenance UC 6
Consumable
Management
RPA (Robotic Process Automation) for Automated Reporting
 Implemented Use Case Review – Outage Heat Map UC 4
Outage Heat Map

Outage Heat Map for Operational Management 0  90%

Alarms
RAN
AABD • StartingFM FM GIS Online, 90% Data
Value: 0% => Current Value: SDMAccuracy WFM WFM
Transmission
… Alarm
Core Auto Auto TT Auto WO Intelligent WO
Correlation •Diagnosis
Target Value: 100% Data Accuracy
Creation Creation Scheduling Dispatching
IT
Success
~30%  92% 0%• Value
98% :Rate 0%  25% 0%  100% 0%  100% Manual
• Floor & Management Visibility
~190,000/day 133k  15k 0  ~850 0  30/~120 0  30/30 (60) 0  60/60
• MTTR Reduction

• Achievements:
UC 1 UC 3 UC 2 UC 5 UC 5 UC 5
• Visible
Auto Diagnosis
Automation Follow
Auto TTup Intelligent Scheduling Intelligent Scheduling Intelligent Scheduling
Alarm Correlation

Consumable Management 0%  100%

Preventive Maintenance 0  54 UC 6
Consumable
Management
RPA (Robotic Process Automation) for Automated Reporting 0  2 UC 8
Auto Preventive
UC 7 Maintenance
Auto Reporting