The ATHENS 2004 Olympic project

1

• The ATHENS 2004 Olympic project
• Palaiseau Mars 16, 2005
• Hubert Tardieu
• Executive Vice President CSI

2
ATHENS 2004 Games

• A massive and complex IT
• environment
• A dedicated Olympic power infrastructure
• Very sophisticated data network
• Servers 900, Computers 10,500
• Results systems terminals 4,000
• Printers 4,000
• Very high degree of redundancy
• Innovative IT Security features
• Un- parallel testing program

3
An 3,400-people Operation Team Adventure

• AO team 397
• Major Events 252
• AO Greece 37
• Volunteers 79
• Interns 29
• 44 nationalities
• Partners 700
• Volunteers 2,400

4
R
5
Olympics Our Mission

• Produce competitions results and deliver them
• to all participants, spectators, media and press
• Bring knowledge and design all key systems
• 5 to 3 years before Games start
• Provide and deploy all key core planning
  applications
• 3 to 1 year before Games start
• Integrate and test all Results Systems
• 2 years before Games start
• At Games time, operate/monitor all the Technology
  systems

6
Results Info Diffusion Systems
Venue
Main Press Center International Broadcasting
Center
Primary Data Center
7
ATHENS 2004 A flawless performance

• 177,244 accreditations issued and activated
• 301 events - 4,500 of live competition
• Live commentator services delivered for 19 sports
• About 16 million of INFO2004 pages viewed
• Peak of 915,000 pages accessed on day 7
• More than 3GB of live results provided in 800,000
  messages to the ATHENS2004 web site, broadcasters
  and Sport Federations
• 11,270 reports printed and 60 million pages
  distributed
• About 73 million of visits to the ATHOC Web site
  

8
Main problems

• Venue readiness, last moment planning
• Machines deployment and SW distribution
• Security Access Control lists deployment
• Volunteers program
• Broadcasters and Olympic Village
• installation
• Performance of the Info system
• Connectivity in Tennis and
• Water Polo

9
The Secret of the Success
Experience
Testing
No Risk
Discipline
Partnership
10
What is in it for our customers?

• Testing program
• Fast track Deployment
• Operations model and tools
• Integrated Information Security
• Partnership Management/
• Engagement model
• Knowledge management

11
What is in it for our customers?

• Testing program
• - Integrated program, tools, process,
  change control
• Fast track Deployment
• - Policies, asset management,
  Automated system configuration,
• configuration management
• Operations model and tools
• - Unique model with advanced
  monitoring tools
• - Change Management and Training

12
What is in it for our customers?

• Information Security
• - Integrated approach definition of
  normal behavior
• - Complex architecture and stringent
  machines profile strategy
• - Posture monitoring / Alarms filtering
  and correlation
• Partnership Management (Influence but no
  authority)
• - Strategy and process
• Knowledge management
• - Size and structure of the whole program

13
Executive Summary

• Scale
• Indisputable reference if we can do it for the
  game we can do it for any project
• Unique expertise applicable to other projects
  processes, technology and engagement model
• Communication - Business Relationships

14

• Les Tendances de lintegration de systemes
• Palaiseau 16 Mars, 2005
• Hubert Tardieu
• Executive Vice President CSI

15
Les differents metiers de service en IT

• Le conseil
• Lintegration de systeme
• Les Managed Service

16
Les grands acteurs du service

• Global
• IBM
• Accenture
• EDS
• Cap Gemini
• CSC
• Atos Origin
• Consulting
• Bearing Point
• ErnstYoung
• Deloitte
• SI
• Sopra
• Steria
• Unilog

17
A Technology Partner in the 90s

• TWO SIGNIFICANT PROJECTS WHERE ATOS ORIGIN(SEMA)
  WAS THE TECHNOLOGY PARTNER
• EDF N4
• EGG/ ZEBANK BANQUE AGF INTERNET BANKS
• ATOS ORIGINS CHARTER AS THE TECHNOLOGY PARTNER

18
A NEW GENERATION OF COMMAND CONTROL CENTER FOR
A NEW GENERATION OF NUCLEAR PLANT
3
19
N4 SYSTEM
A unique fully computer-based nuclear plant
command and control system
contract in excess of 250 millions
a team of 110 staff
more than 1.5 million linesof Ada source code
control of 35,000 devices and parameters
4
20

N4 COMMAND CONTROL COMMAND
Objectives

• An innovative project integrating lessons
  learned from operating
• EDF nuclear plants (15 years)
• An entirely computer aided command control
• An ergonomy fitting security constraints
• A high availability system
• An open system

5
21
N4
COMMAND CONTROL CENTER Performances

• Sizing
• - 35 000 sensors - 19 000 activators
• - 500 ms response time
• - 3 500 pages operating procedures
• - 5 000 alarms
• Availability
• - reliability and IT system redundancy restrict
  manual control
• (never more than 8 hours)

6
22
N4 COMMAND
CONTROL CENTER Technical Solutions

• Technical options
• - double Ethernet
• - computer redundancy
• - ADA language
• Industrial choices
• - Hartmann Braun
• - Digital Equipment

Management System
Level 3
Communication Interfaces
Command Center
Operator workstations
Computers
Level 2redundant
Communication Interfaces
Level 1
Automate
Level 0
Activators, sensors
7
23

N4 COMMAND CONTROL CENTER The
System Integration reference

• Managing 10 sub contractors
• Using 6 development and integration platforms
• Implementing all software components
• Help EDF to implement command and control
• Training operators through simulators
• Limit of responsibilities
• - nuclear risk outside of contracts
• - control process defined by EDF

8
24
N4 CONTROL
COMMANDTechnology Challenges

• First use of ADA language in large civil
  contract
• In June 90, after 3 years of prototyping
• - EDF decides to terminate its contract with
  Level 1 Vendor.
• - SEMA realizes that Gould Unix 16 bits machine
  cannot deliver.
• ? SEMA takes the initiative of
• - suggesting Level 1 delivery by Hartmann
  Braun.
• - investigating DIGITAL VAX (32 bits) and VMS to
  replace Level 2
• hardware.
• ? EDF gives SEMA the responsibility of Level 1
  and Level 2.
• In March 92, Sema and EDF discover that Level 1
  CAD fitting previous
• Level 1 Vendor cannot match HB spec.
• ? Sema helps EDF to adapt its Level 1 CAD.

9
25
N4
CONTROL COMMAND Lessons Learned

• A contract worth more than 250 millions
• One of the largest industrial IT system
• - more than 4 000 m/m
• - a team peaking at 110 staff
• - 1.5 millions of ADA lines
• All planning steps have been met
• - Feb 86 initial order (Level 2)
• - June 90 delivery of V1
• - Dec 90 terminating previous Level 1 contract
• - Mid 91 signing the contract with EDF (Level 1
  Level 2)
• - May 92 site delivery of Level 1 subsets
• - July 93 site delivery of first version Level 2
  (KIC)
• - July 93 - June 95 stage deliveries
• - June 95 final delivery
• Performance in line with specifications.

10
26
SEMA AND
THE SECOND GENERATION OF INTERNET BANK
A NEW GENERATION OF MULTI LAYERED IT SYSTEM TO
SUPPORT MULTI CHANNEL, MULTI TRADE INTERNET BANKS
11
27
EGG/Zebank, BANQUE AGF
A unique multi-channel, multi-application Internet
banking system

• 2 contracts in excess of
• 20 million Zebank
• 10 million Banque AGF
• a team of 60 60 staff
• manage several 100 000 customers
• under service level agreement
• integrate best of breeds third party
• packages around Sema architecture

Sample of ZEBANK screen
12
28
EGG/Zebank,
BANQUE AGF Objectives

• An innovative projects integrating lessons
  learned by SEMA in banking
• since 15 years.
• A seamless functional architecture to manage
  banking, brokerage, loans,
• insurance,...
• A technical architecture allowing Internet,
  mobile and phone banking .
• A highly reliable system where all transactions
  are executed in real time.
• An open system giving access to finance third
  party offering.

13
29
EGG/Zebank,
BANQUE AGF Performances

• Sizing
• - 1 200 visitors, 600 customers and 100 CSRs
  can access
• concurrently
• - response time within seconds for all major
  banking and
• brokerage transactions
• Availability
• - 24 x 7 with less than 2 hours interrupt /
  week
• - fully monitored system
• ? network
• ? system
• ? application
• ? security

14
30
EGG/Zebank,
BANQUE AGF Technical Solutions

• Technical options
• ? 3 OS - UNIX SYSTEM
• - AS 400 BANKING
• - VMS BROKERAGE
• ? FULL ACTIVE REDUNDANCY
• ? XML, JAVA SCRIPT
• ? FORTE FOR FOCUS-2-ONE
• Industrial choices (best of breed)
• ? SUN FOR UNIX SYSTEMS
• ? AS 400 / IBM BANKING
• ? COMPAQ BROKERAGE

Internet
CSRs
BROADVISION
FOCUS 2-ONE
Brokerage
Banking
Insurance
15
31
EGG/Zebank,
BANQUE AGF A reference in System Integration

• Best of Breed approach for components under
  severe availability /
• performance constraints.
• Use of 4 platforms for development,
  qualification, integration
• and operation.
• BOT for Banque AGF (FRANCE TELECOM operates).
• BTO for ZEBANK (revenue per transaction - SEMA
  operates for 4
• years).
• Banque AGF delivered in 6 months.
• ZEBANK delivered in 12 months.

16
32
EGG/Zebank,
BANQUE AGF Technology challenges

• Integrating through real time protocol
  applications on three different
• operating systems.
• Using object oriented techniques - providing
  genericity - in the
• middle ware while delivering performance.
• Specializing architecture layers by function
• - Internet access in Broadvision
• - Customer product data base and work flow
  management in the
• middle ware
• - Trade specific process in the back end.
• Building a system which is reliable in order to
  secure transaction based
• revenue.
• Create shift teams to operate an Internet bank
  with sufficient
• understanding of banking process, security and
  network management.

17
33
EGG/Zebank,
BANQUE AGF Lessons Learned

• 2 contracts worth 30 million
• A business system inheriting from skills
  acquired through technical
• system development
• ? real time from energy defense
• ? system management from telecom
• ? SLA from outsourcing
• Time frame of the projects
• Banque AGF order in April 2000, launch October
  2000
• ZEBANK order in Nov. 1999, launch Nov. 2000
• Performances in line with expectations.
• Banque AGF more than 150 000 customers after 6
  months.

18
34
The Technology Charter Partner

• The charter
• implementing complex solutions intricating
  strategy, business process technology where no
  precedent exist.
• bring innovation by concentrating on global
  architecture while using proven technologies and
  understanding the substance of the process.
• deliver on time despite major technical
  difficulties by being right from the inception
  about the global architecture and anticipating
  when needed changes of components.

35
Les grandes tendances 2004 en Integration de
Systeme

• Process
• Capability Maturity Model
• TMA
• Progiciels
• Global Sourcing

36
Les tendances technologiques 2005-2006

• Open Source
• Securite
• SAP
• ECM
• RFID
• Voice over IP

37
LIntegration de systeme change

• Projet TMA
• Specifique Progiciel
• Artisan Industriel
• Local Global

38
Les fondamentaux restent

• Lintegration de systeme est une discipline
  dingenierie
• La modelisation est indispensable pour pouvoir
  decrire et predire
• Lintegration de systeme peut senseigner et sera
  enseignee dans des pays a bas couts qui viendront
  concurrencer les pays europeens
• La maitrise de la complexite reste le facteur de
  differenciation

39
Les carrieres sont diverses et variees

• Chef de projet
• Architecte Technique
• Consultant
• Expert
• Ingenieur commercial
• Responsible de PL

40
Les Challenges a venir

• Parier sur la globalisation en differenciant les
  metiers a promouvoir en Europe
• Urbaniser plutot que construire du neuf
• Industrialiser
• Prendre la responsabilite sur la duree ( 5 a 10
  ans) en sengageant a reduire le Total Cost of
  Ownership
• Parier sur la capacite des individus a
  sapproprier des outils de plus en plus
  informatises

41
Thank you