Ambient computing

Ambient computing
Ambient computing
Michel Banâtre
Paul Couderc
Frédéric Weis
Content
‹ Two examples (video)
‹ Mobility
‹ Introduction to ambient
– Definition
– Context
– Reasons
‹ Examples
–
–
–
–
–
The precursors
Location based
User assistant
Colaborative backup
Physical web
‹ Implementation
computing
Introduction
‹ Video
– Ubi-Board
– Ubi-Bus
‹ Some
explanations
Content
‹ Two examples (video)
‹ Mobility
‹ Introduction to ambient
– Definition
– Context
– Reasons
‹ Examples
–
–
–
–
–
The precursors
Location based
User assistant
Collaborative backup
Physical web
‹ Implementation
computing
Mobility (1)
‹
Mobile applications (« standard »)
–
–
–
–
–
–
‹
Distributed files systems,
Databases,
Web,
Mail,
Multimedia
…
Mobile support for execution
– Mobile computer,
– Appliances

PDA, mobile phone…
Mobility (2)
‹
Problems
– Limited resources







Processing power
Energy, (battery),
Network, (bandwidth, connectivity,…),
User interface, (period of interest) Environment variable
Energy <-> MIPS
Networks,
Dependability
– Security,
– Reliability, ….
Mobility (3)
‹ Solution
– Adaptative approaches
Transparency
 Application level,
 Hybrid

Mobility (4)
Transparency
‹
‹
Principle: To emulate at the system level missing resources (Middleware
approach)
Examples:
– Cellular network


handover technique to mask user mobility between cellulars (the mobility is
transparency from the user point of view
Disconnected file systems (CODA)
– To emulate the connection to the file server
• Hoarding in connected mode, similar to prefetch technique in architecture.
• caching in the disconnected mode
• Update synchronisation at the reconnection time (hard problem due to multiple
writers).
• One limit: coherency problem at the reconciliation time
Mobility (5)
Transparency
‹ Examples
– Info stations


To provide a continuous network with a high bandwidth onto a
discontinuous network (discontinuous high bandwidth small cellular)
Problems
– To take into account the discontinuity and the distance between two
cellulars.

Solutions
– Implementation of (Hoarding) mechanisms at the info station level, and
caching mechanisms at the terminal level.
– Take into account the distance between info stations and the velocity
and trajectory of user in order to define the size of the cache in the
terminals.
Hybrid solution
‹ Principle
– To combine the two previous approaches


Transparency, if possible,
Provide the adaptation at the application level otherwise..
‹ Example
– ASR CODA

After a disconnections mode, if there is a conflict problem for file
update reconciliation there is a specific application Application
Specific Resolver to solve the conflict.
Analyse
‹
Transparency
– No impact on the applications (in particular existing one) but it is hard to
take into account mobility aspects.
‹
Specific adaptation
– More complex
– But it is possible to have a fine adaptation the behaviour of the
application and integrate mobility aspects in the application design.
Content
‹ Two examples (video)
‹ Mobility
‹ Introduction to ambient
– Definition
– Context
– Reasons
‹ Examples
–
–
–
–
–
The precursors
Location based
User assistant
Colaborative backup
Physical web
‹ Implementation
computing
Ambient Computing (1)
Main principles
‹ Another
way to exploit mobility and embedded
systems.
‹ In the mobility area,
– The user distinguishes his laptops and his real life
environment.
- Real life environment: movement, temperature, noises, …
information to be exploited.
-
-
Neccesity to provide new models to acces and to manage useful
information based on the notion of context.
To imagine new applications.
- Behind this there is …Ubiquitous computing, ambient
computing, pervasive computing….
Ambient Computing (3)
Main principles
‹ Context
can be related to a location problem,
position (x,y).
‹ It is necessary to have a strong integration of
computing equipments and the real world.
– It is based on



Mobile and wireless facilities,
Embedded systems,
Nice user interfaces
‹ First definition given by Mark Weiser
« The computer for the 21st century »
in 1991
Ambient computing (3)
Weiser’s definition
‹
Principle
– « Merging physical and digital world »
Spreading intelligence in everyday objects
‹
Motivation
– To help users when they interact with their real environment.
– The computer has to des appear and the interaction with this computer have to be
implicit (if possible)
– Context awareness (context sensitive))
– Information which characterizes a real life situation
– Information which relates one entity with its current task.
Three steps
– A set of sensors capture information from the environment,
– From this information, we have a situation,

– Services and/or information delivering are provided related to this situation.
Adaptative vs ambient computing
approach
‹ Adaptative
approach
– We try to make applications already running in new
environments related to mobility and wireless
communication facilities.
‹ Ubiquitous
computing
– We assist user in a transparent way when he moves
inside his environment.
Context sensitive
‹ Computer
use is quite different.
– It is not machine centric, but it combines at the same
time information captured from the real world and
information from the information system.

Information and services need to be accessible in an
spontaneous way depending of the context.
– Two problems:
To define and to get a context,
 To exploit a context.

To get a context (1)
‹ To
get a context = to detect and analyse situations.
– Ex. : number of persons in a room, name of these persons,…
‹
Four classes of context:
– Computer resources context:

Access to a network, available bandwidth, available resources,…
– User context :

Identity, profile, physical location, ….
– Environment physical context related to the user :

Noise intensity in a room, temperature, people near a the user,…
– The time parameter:

It can be combine with the other context elements in order to be able to
manage the context history.
To get a context (2)
‹
Simple methods,
– Embedded sensors on the entity; GPS,…
– Embedded sensors in the environment: movement detection, light detection,
…
‹
Combined methods, take into account a set of criteria.
– Context = set of persons in a same room
– Application = to get benefit of these persons in the same room in order to
exchange information based on common interests.
– Context related to the physical proximity of mobile entities (communication
range) and of their mobility.
Context exploitation
‹ Adaptation
 Information presentation (ex: virtual reality)
 Performance increasing (ex: energy saving)
‹ Information access
 To extract information from the information system using the
context (the nearer pizzeria)
 Notification
‹ Contextual annotation
 To generate information in the SI from the context (ex: observations
“on the field”)
Reasons for ambient computing
‹ Why
ambient computing is possible now ?
– Technological reason
 Wireless
communications, Short distance wireless
communications
 Embedded systems
– New users behaviour
 Mobile
phone, web accesses.
Technological reasons (1)
‹ High
hardware integration (small size)
– "Systems on the chips", smart cards …
‹ Design
of new embedded operating systems
– New solution for critical resources management

Energy
– Dynamic update of the processor frequency in order to reduce the
energy consumption, in particular in multi processor architecture)

Real time
– Programming

Mobile code : Java, API MIDP
Technological reasons (2)
‹ New
technologies
– GPS, RFID, …
‹ Wireless
networks short range
– WPANs : Bluetooth, wireless USB …
– WLANs : 802.11 a/b/g, which can be used either in the ad hoc or
infrastructure mode
– Cellular networks : 2G/2.5G (GSM / GPRS), 3G (UMTS)
– Future networks 4G : discontinuous coverage networks based on
very high bandwidth network (info station) and traditional
cellular networks.
New behaviour
‹ New
vision of the computer due to the development
of internet.
‹ People are more reactive to the deployment of
implicit information systems.
– Mobile phone uses anytime, anyway
– New local services are increasing




Calendar, distributed agenda
SMS, MMS
Web access (WAP)
….
Content
‹ Two examples (video)
‹ Mobility
‹ Introduction to ambient
– Definition
– Context
– Reasons
computing
‹ Examples
–
–
–
–
–
The precursors
Location based
User assistant
Collaborative backup
Physical web
‹ Implementation
The first ones: ParcTab (1)
‹ ParcTab
project Xerox-Parc
‹ A campus especially equipped
with
– Each user has a small computer
(the tab) with a tactile screen and
an hand written recognized system.
– The computer is able to
communicate with a server using
infra-red cellular located in the
environment
The first ones: ParcTab (2)
Tab
Pad
board
User
IR antenna
The first ones : ParcTab (3)
General architecture
SHELL
MAIL
TAB
SHELL
TAB
MAIL
CALENDAR
SHELL
MEMOS
ParcTab
Agent
TAB
Tab
VOTE
CALENDAR
Transceiver
IR
Ethernet
Gateway
Applications
LOCATE
The first ones: ParcTab (4)
‹ The
user location is based on the infra-red cellular in
the physical environment.
‹ Applications
– « classical ones »



Information access using the cellular network.
Mails
Location et paging
– Contextual applications



Group management and shared board,
Resource selection (printer, fax,…) taking into account the physical
proximity ( to be closed to an infra red access point)
To activate an application taking into account a specific context.
The first ones: Active Badge (1)
‹ Designed
at the ATT
lab Cambridge
‹ First system provided
an indoor positioning
‹ Goal
– To localized people in a
building
– The first idea was to
help the phone operator
to ring a fix phone in the
nearest office near be
the called person
The first ones : Active Badge (2)
‹
Individual tag with low energy
consumption
– Infra red communication
– A beacon is sent every 15 seconds
toward a (infra red) sensor
‹
Sensor network in a building
– Connected using phone circuit
– Each sensor is able to store upto
20 user uid before processing
‹
QuickTime™ et un
décompresseur TIFF (non compressé)
sont requis pour visionner cette image.
A master station is connected to
this sensor network in order to
deliver user identifier to the phone
operator.
The first ones: Active Badge (3)
Commands provided to the operator
FIND (name)
– To localize the current tag and deliver the five last localizations of this tag.
WITH (name)
– To localize the current tag and deliver information related to proximity tags.
LOOK (location)
– Look for tag near a given location.
NOTIFY (name)
– Alarm toward a specific tag. ( the alarm is delayed until the tag is closed to the
sensor network).
HISTORY (name)
– To deliver the history of a tag location during the last hour.
Location systems
‹ Mainly
oriented on the physical location of users
– ex. : Smart Floor
‹ Major
problems
– Cost for deployment
– Confidentiality.
‹ Application
domains
– WEB Technologies: Cooltown, Mobisaic
– Assistant for visit



Museums : Hippie
Cities : Lancaster’s guide, CyberGuide, GUIDE, Smart Sight
Campus : Metraunot, C-Map
– Assistant for shopping : ShopNavi
Smart Floor
‹ System to identify and to locate an user
– Researches done at Georgia Tech.
– To propose a new solution which differs from the classical
ones based on tag or voice recognition
‹ Based on the measure of biometric parameters
 Biometric parameter: GRF (Ground Reaction Force),
 To define a GRF signature for each user,
 To identify a user from the comparison between the current GRF and
the GRF database where all the GRF are stored.
– Efficiency of the recognition: 90 % (group of twenty users).
‹ Application
– Information delivery to students on interactive board depend
on the identification.
Mobisaic
‹
Location sensitive information system.
– Based on a WEB architecture
– Informations are stored on http servers.
– User accesses to servers are based on a wireless network.
‹
Two basics mechanisms
– Dynamic URLs

Environment variables are part of URLs
– Ex. : http://www/offices/$(location).html

Association: dynamic URL to static URL is done on the client appliance
– Dynamic documents

A web page is updated as soon as the user context is changing, (for example
his location).
– The client is linked explicitly to variables,
– This linking operation is managed by the infrastructure
Cooltown (1)
Principles
‹
‹
Déveloped by the HP Labs
Goals
–
–
‹
‹
To combined the web infrastructure and the basic concepts behind ambient
computing.
Attach a virtual representation to places people, day life objects,…
Each entity is associated to a set of web pages, the access to these
web page is related to the physical proximity.
Web presence
–
–
Either we have a local server associated to an object (printer, video
projector,…)
Either a remote server is in charge of providing information, the access is
initiated from an object, (ex: painter in a museum).
CoolTown (2)
CoolTown in a meeting room
Internet
Place Web Portal
Place
manager
URL
URL
-Reservation
-Printer
-Electronic whiteboard
-People
Print
URL
Chat
URL
CoolTown (3)
web presence on a passive object
Tag resolver
tag
WWW
2. Transformation en URL
1. Get id
3. Get HTTP
Lancaster’s guide (1)
‹
‹
‹
Electronic guide for tourism in the
old Lancaster city.
Experiment with 60 users during four
weeks.
Experimental architecture
–
–
‹
Tablets PC
WiFi 802.11
Physical location based on WiFi
cellular.
Lancaster’s guide (2)
‹
Localisation object
galery
castle
References
between objects
Main characteristics of the
information system:
–
–

coffee
Link to an html page

‹
Geographical information,
Hypertext information
Implementation of data caches to
mask disconnection problems,
(between two WiFi cellulars)
–
Pages
HTML
Dynamic and distributed
Based on
However these data are considered dirty if
the disconnection time is too long, (data
will be out of the location context)
Lancaster’s guide (3)
Applications
‹ Interactive
guide to find information
‹ Navigation into the city using a map providing by
the system.
‹ Creation d’un tour personalized tour.,
‹ Communication with other visitors
‹ Connection with others services (hotels,
restaurants).
C-MAP system
‹ System
for guide tours in museums
– To allow to users to locate points of interest on a map.
– To take into account the profile of the users, (semantic map)

Dynamic management of a personalized visit.
– User location based on active badge (ATT)

Size of a cellular, two meters.
‹ Information
stored on a centralized server, accessed
using a WLAN.
– The UID of the tag of a user are sent using html.
Metraunot system
‹
Guide system on a campus
– Design to give the good direction to users.
– Localisation is based on bar code.

Bar codes are associated to events which are automatiquely integrated in the
user information system.
– Information is stored on three servers connected using WLAN.



(1) A data base for localisations
(2) A data base to store information about user location.
(3) A data base for events
– The appliance is responsible to syntetize a context using (1) and (3) and
find the user direction using (2).
Shopping applications
‹
Applications designed to help the client in a shop.
– Exploits a richer context than those which are based only on localization.
– Take into account the physical proximity of products in the shop.


Can be used to personalize advertisement.
Can be used to build user profiles
– Ex:WEB navigation in a shop,

‹
Ajout de la dimension physique à la navigation WEB
ShopNavi (SONY)
– The client get directly information related to the product he is taking off,
– Help to localize a product

Virtual reality.
User « memory » assistant (1)
‹
‹
This assistant is based on past events
Forget-me-not
– To create a data base of context in order to help
the user to find a past situation.
– Requests on this data base such that «is ju-Julio
present? »
– Similar architecture to Parc Tab (Xerox)

The Tab collects information related to user
activities
– Person encounters,
– Access to an application (agenda, board, …).
User “memory” assistant(2)
‹ Automatic
occurs.
recall of a task to run when a situation
– CyberMinder

Notification of the recall using the communication message system
– Mail
– SMS,…

Each message has a deadline
– The physical context is considered


Notification triggered
Notification process,
Encounters applications
‹
To exploit the physical proximity of mobile users
– To enrich encounters with contextual services,
– Major interest with Bluetooth
‹
Simple interactions …
– Proxy Lady

Notifications triggered when proximate interesting profile are discovered
(mails exchange)
– More complex complex…
– Proem

Automatic exhange of user profile, (professional cards),
– Side Surfer

Automatic exchange of interesting documents based on user profile.
Collaborative backup
Mosaic
Before, data were
produced
on fixed station.
Now, new devices create data during disconnection
period
„
Short-range wireless communications
(WiFi, BlueTooth, etc…)‫‏‬
„
Mobile terminals
(cell phones, PDAs, digital cameras, mobile sensors, mobile
robots, ...)‫‏‬
„
New data
(Pictures, movies, schedules, contact lists, etc…)‫‏‬
Risk of data loss when the device fails
A collaborative backup system can cope
with this problem
48
Collaborative backup
MoSAIC running
‹ Scenario
:
– Alice takes notes on her devices during a meeting
– After the meeting, she takes the bus home
– Once at home, she notices that she has lost her PDA
Loss of the device ¸ Loss of data
– But, thanks to MoSAIC, Alice recovers her data from the
Internet once at home

The data have been transparently backed-up on neighbor
terminals by MoSAIC
49
Collaborative backup
MoSAIC's Basic Idea
MoSAIC uses neighbours interaction to backup data
Reliable storage
on the Internet
Home terminal
Short-range wireless
communications
Very high data
resilience
Low data resilience
Increasing data resilience
50
Collaborative backup
Issues
‹ Handling
data coherency and data dissemination
– Fragmentation, replication, etc...
– Scheduling replica backups
‹ Resource
management
– Network management
– Memory management
‹ Security
– Encryption of data
– Trust between terminals
51
Collaborative backup
Applications
‹ Personal
devices
– PDA
– Cellphones
MoSAIC - http://www.laas.fr/mosaic
‹ Swarm
robots
– Mobile robots realizing collaborative tasks
‹ Mobile
sensors
– Delivery tracking
– Contagious disease tracking (for animals)‫‏‬
52
The Web in the psysical space
‹ Motivations
– Mobile navigators (Wap,…).
– Numerous documents related to physical objects
and/or represented physical objects/ places.
Opportunities to access these informations taking into
account the physical context.
 Opportunities to build web site implicitly from the
physical world.

Context in the Web (1)
‹ Relative
position of a document in the
information space.
– Proximity due to other documents (or pages)

Example :
– Set of pages which referenced a document/page.
Context in the Web (2)
‹
Existing dimensions
– References (<a ref=…>)


Static links specified by the creator/writer of the documents.
Dynamic links
– Ex: collaborative navigation :
• Clients who have bought this book have also bought this other one…
– Temporal location

The history of navigation (past/next button)
– Textual proximity:

A set of world/phrase which can be discovered in a document
Context in the Web (3)
Mobility impact
‹ Numerous
« starting points/roots » are related to
the physical context.
– Examples :
Time schedule for flights in an airport.
 Information on a product in a shop.
 Attendee profiles in a conference.

Context in the Web (4)
The idea
‹ To
consider the spatial navigation as a new dimension
in the Web, like the hypertext, or historical one). :
– In the spatial mode, le navigator displays spontaneously the
page for the current place.

Ex: the good scheduling, if I am closed to trains to the airport
– Obviously classical navigation is always available, the user
need to choose it.


Temporal navigation (next/previous)
Hypertext navigation (given url)
Context in the Web (2)
Looking for information in the physical space.
‹ Association
of information to physical objects.
‹ A set of word in a given geometrical space
(sphere,…) specifies a textual context.
– Example :
« chocolat »,
« sarrazin »
cuisine.com
Crepes, Genoise au
Chocolat, etc.
New applications
‹
Smart space
– New possibilities for interactions,

‹
Help to reduce handicap (cultural and/or physical).
Automatic checking
– Intelligent frig.
‹
‹
‹
Stock management
Game,
Etc.
Content
‹ Two examples (video)
‹ Mobility
‹ Introduction to ambient
‹ Examples
‹ Implementation
computing
Implementation problems
‹
To implement an implicit link between real environment and
processing.
– To get a perception of the state of the environment and of its change
(sensor).
– To process the collected data and determine the impact of the result of
this computation on the environment.
– Distributed execution in a very volatile environment
– Information are « physically » linked to physical objects which may have a
strong mobility
– Physical processing are tightly coupled with physical activities.
Ambient computing application design
‹ Logical
approach
– GIS geographical Information System
‹ Physical
approach
– Spatial Information System
Principles
Classical implementation
‹
Based on geolocalisation
– Physical -> logical representation -> Physical
Service plateform
GPS
Logical representation
model
?
processing
Perception
Closer cab?
Actions
Physical world
– Drawbacks

Complexity, scalability, cost, privacy (people)…
Principles
one observation
‹
It is not necessary to have a global infrastructure:
– Information is linked to objects
– The interactions in the real world can be considered as the processing of this
information.
‹
Ex: Urban transport services
Arrival time :
04h07pm
Line# 16
Destination: university
Next stop: tournebride
blind: STOP 16 requested!
The Spatial approach (1)
‹
Principles of the spatial machine
– The structure of the information system is based on physical objects and
geometrical properties related to these objects

Orientation, relative position of objects, physical coverage of the data associated
with an object
– The physical space is considered as a « big memory » used to structure and
manage all information. An information item belongs to a shape (cube,
sphere, …)
– The addressing mechanism and Read/Write operations are based on these
physical properties
– The application control flow is driven by physical mobility
<d>
<a>
<a>
<e>
<b>
<c>
<f>
Application mobility
<g>
The Spatial approach (2)
‹ Information
system built from the physical space.
– Data are asociated to geometrical space around the considered
physical objects
– The computation is directly expressed from the interaction
between objects from the physical space.
<‘a’>
<12,‘b’>
<‘c’,5>
<‘a’>
<‘a’>
<12,‘b’>
<12,‘b’>
<‘c’,5>
The Spatial approach (3)
‹
Two questions:
– How to support spatial addressing?
– How to reflect process mobility in a such information space?
The Spatial approach (4)
‹ Spatial
coverage of an information
– Geometrical property :

Volume
– Sphere
– cube

‹
To belong (or not) to a volume.
Information structure
– Information is structured using tuples
– A tuple is always associated to a physical object.
<10int, ‘pierre’string, 50.3float>
<‘a’>
<5>
<‘somme’,10>
The Spatial approach (4)
‹ Tuple-space,
derived from Linda
‹ Linda
– Programming language for process synchronisation.
– Shared memory between processes
– Anonymous addressing based on type
motifs
tuples
<string, int>
<10, string>
<‘a’,5>, <‘b’,’b’>, <‘b’,3>
<5,’a’>, <10,5>, <10,‘a’>, <10,’a’,5>
– Linda primitives :




Out(t)
In(pattern)
Rd(pattern)
Eval(t)
The Spatial approach (5)
‹
Instructions
– Out(t)

To publish a tuple
– Rd(m)


To read a tuple among those corresponding to the type m
The process is suspended until one corresponding tuple is found
– Capture(m)


To read all the tuples corresponding to the type m.
The process is suspended until at least one corresponding tuple is found.
– Drop(t)

To retrieve one tuple, the instruction can be run only by the process which
have created the tuple (out operation).
The Spatial approach (6)
‹
Physical objects are able to:
– To publish tuples
– To « view » tuples associated to proximate physical objects.
‹
tuple-space management
– Embedded computer (mobile phone, access point,….)
‹
Tuple spatial coverage
– Short distance Wireless communication. Infrarouge




IEEE 802.11
Bluetooth
RFID
Mixed of Bluetooth and RFID (see technical details)
The Spatial approach (7)
‹ Application
design
– Building and managing information system implicitly :


Spontaneous construction of the information system simply by
disposing physical objects in the space. The relations between
the information are derived from the spatial arrangement of
the objects.
Spontaneous evolution of the information system when objects
move.
The Spatial approach (8)
Differences with classical solutions
Logical representation
model
Perception
Logical representation
processing
Actions
Physical World
Physical world
‹
‹
Processing
The global infrastructure is not necessary
Simplicity (architecture et programmes).
– It is not necessary to maintain an intermediate representation of the physical world (Geographical
database)

No centralisation of confidential informayion related to people.
– Easy to deploy (low cost).
‹
Independent from operators.
– Free services.
The Spatial approach (9)
Results
‹
Support for a spatial information systems
– Well defined to support spontaneous operation related to user mobility and to
proximate interactions.
‹
Innovative applications:
– The mobile phone is the best interface between users and ambient
information systems. It is used as a sensor.

WebWalker, a physical Web

Ubi-Bus, to help handicapped person in public transportation

Ubi-Q, a system to support efficient queue management (DAB, fast-food,…)

Ubi-Board, A fragmented and distributed display.
SPREAD : architecture
•Request application management, flow
•Request application management, flow
control
control synchronisation
synchronisation
•Spatial data manages with four operations (API)
SPREAD API
SPREAD
engine
Controller
••Two
Two tupple
tupple space
space management
management system
system
•Local : tuples published by the
•Local : tuples published by the
local
local node
node
••Remote
Remote :: tuples
tuples received
received from
from
neighbour
neighbour nodes
nodes
RemoteTM
TM
Remote
Readannouncer
announcer
Read
Tuplelistener
listener
Tuple
LocalTM
TM
Local
Tupleforwarder
forwarder
Tuple
Readlistener
listener
Read
Wireless interface
OUT <data>
••Read
Read Announcer
Announcer :: periodical
periodical
announced
announced of
of waiting
waiting read
read (UDP
(UDP
multicast)
multicast)
••ReadListener
ReadListener :: listent
listent to
to Read
Read type
type
of
of neighbour
neighbour nodes
nodes
••TupleForwarder
TupleForwarder :: send
send tuples
tuples
corresponding
to
a
corresponding to a read
read type
type
••TupleListener
TupleListener :: received
received tuples
tuples
related
related to
to aa read
read type
type
OUT
Controller
<data>
RemoteTM
TM
Remote
LocalTM
TM
Local
<info, new>
<data>
Readannouncer
announcer
Read
Readlistener
listener
Read
Tuplelistener
listener
Tuple
Tupleforwarder
forwarder
Tuple
LocalTM
TM
Local
Tupleforwarder
forwarder
Tuple
Tuplelistener
listener
Tuple
RemoteTM
TM
Remote
<tuple, 10>
Readlistener
listener
Read
Readannouncer
announcer
Read
DROP <data>
DROP
Controller
<data>
RemoteTM
TM
Remote
LocalTM
TM
Local
<info, new>
<data>
Readannouncer
announcer
Read
Readlistener
listener
Read
Tuplelistener
listener
Tuple
Tupleforwarder
forwarder
Tuple
LocalTM
TM
Local
Tupleforwarder
forwarder
Tuple
Tuplelistener
listener
Tuple
Readlistener
listener
Read
Readannouncer
announcer
Read
RemoteTM
TM
Remote
<tuple, 10>
READ Local
READ <data>
<data>
Controller
<data>
<data> ?
RemoteTM
TM
Remote
LocalTM
TM
Local
<info, new>
<data>
Readannouncer
announcer
Read
Readlistener
listener
Read
Tuplelistener
listener
Tuple
Tupleforwarder
forwarder
Tuple
LocalTM
TM
Local
Tupleforwarder
forwarder
Tuple
Tuplelistener
listener
Tuple
RemoteTM
TM
Remote
<tuple, 10>
Readlistener
listener
Read
Readannouncer
announcer
Read
READ Remote
READ <tuple,?int>
<tuple,10>
Controller
<tuple,?int> ?
<tuple,?int> ?
<tuple,?int> ?
RemoteTM
TM
Remote
Readannouncer
announcer
Read
<info, new>
<data>
<tuple,10>
Tuplelistener
listener
Tuple
Tupleforwarder
forwarder
Tuple
Readlistener
listener
Read
<tuple,10>
<tuple,?int> ?
Readlistener
listener
Read
<tuple,?int> ?
LocalTM
TM
Local
<tuple,10>
Tupleforwarder
forwarder
Tuple
Tuplelistener
listener
Tuple
<tuple,10>
RemoteTM
TM
Remote
LocalTM
TM
Local
Readannouncer
announcer
Read
<tuple,10>
CAPTURE
<tuple,10> <tuple,20>
<tuple,11>
READ <tuple,?int>
Controller
<tuple,?int> ?
<tuple,?int> ?
<tuple,10>
<tuple,11>
LocalTM
TM
Local
RemoteTM
TM
Remote
<tuple,10>
<tuple11>
<tuple,?int> ?
Readannouncer
announcer
Read
Tuplelistener
listener
Tuple
<tuple,?int> ?
Tupleforwarder
forwarder
Tuple
<tuple, 20>
<data>
Readlistener
listener
Read
<tuple,10><tuple11>
Readlistener
listener
Read
<tuple,?int> ?
<tuple,20>
Tupleforwarder
forwarder
Tuple
Tuplelistener
listener
Tuple
<tuple,10><tuple,11>
LocalTM
TM
Local
<tuple,10>
<tuple11>
RemoteTM
TM
Remote
Readannouncer
announcer
Read
Technological aspects (1)
‹
Context sensivity and context description have different requirements,
depending on the service/application
–
–
–
–
–
Range
Reactivity
Penetration / Radiation properties
Multicast support
Built-in security support
Reactivity [lower bound; upper bound]
10s
1s
10ms
HF
[Location accuracy; range]
UHF
1m
3m
10m
100m
Technological aspects (2)
RFID + Bluetooth
‹ Basic
principle
– Linking the handset (or user device) with other(s)
physical object(s).
Virtual link
Regular BT
comm
RFID : provides connexion parameters
and context descriptors (language, etc.)
Obvious benefits :
- bypass BT sluggish reactivity
- can address non-discoverable devices
QuickTime™ et un
décompresseur TIFF (non compressé)
sont requis pour visionner cette image.
QuickTime™ et un
décompresseur TIFF (non compressé)
sont requis pour visionner cette image.
Technological aspects (3)
Advantages
‹ Progressivity
– Leverage on existing technologies

BT was slow to take off, but has reach a significant user base
(> 250 Millions phones in 2006, ~1 billion expected by the end
of 2008)
– The tag can be used to address heterogeneous
communications technologies (SMS, BT, Wi-Fi…) and
various devices (phone, headset, … )
Bt addr,
accept streaming audio
Bt addr,
accept audio, image, …
Technological aspects (4)
Advantages
‹ Flexibility
– Support both personal and collective services
– Support spontaneous as well as user-initiated triggering
– Depending on the requirements of the service and the target
site, using external tags allows a fine tuning of the context
sensitivity :


Selection of the appropriate RFID technology (LF, HF, UHF)
Specific antenna design
Technological aspects (5)
Advantages
‹
Convenience
– Physical vs virtual, will the users really adopt “e-everything” ?

tangible objects are still preferred in many occasions: post-it, calculators,
map clips, photo prints…
– More convenient : simpler, easier, faster
– Too much services/functions packed into a single device: complexity in function
selection. External tags linked to user devices allow tangible and contextualized
access to the services.
– Tangible link with the end-user for the service provider / Branding

Physical cards / Tickets has value
QuickTime™ et un
décompresseur TIFF (non compressé)
sont requis pour visionner cette image.
Technological aspects (6)
Conclusion
‹
vs NFC
– NFC is very promising for many services where user initiative (including
payment) is the way to go.
– However,



‹
NFC-enabled phones will take time be mass deployed
How exactly users will adopt new usage is not known
Experimentation is important
RFID combined to existing communication technologies (in particular
Bluetooth) offers advantages
– Can be used today, on many existing devices
– External tags offers great flexibility for the service provider or site operator
Conclusion
Ambient computing and industry(1)
‹
Very hard to find truly ambient computing application in the
industry today
Ambient computing and industry(2)
‹ Why
is it not developed enough today?
– It is application domain dependent

Stick management
– RFID solutions very well deployed
‹ Mobile
phone as the user terminal
– Minority terminals (Bluetooth + Java), but it is changing
(50% by the end of 2008)
– It is a concurrent of operator technology, 3G/UMTS

Difficult to make ambient computing services compatible with the
operator « business model ». They prefer UMTS.
Ambient computing and industry (3)
‹
Who can be interested?
– Users who request a rich interaction with the physical world without any
dependency with the telecom operator (economical cost)


Advertising, …
Museum, electronic newspapers, …
‹
Ambient computing is a simple answer to handicap (physical, cultural),
but it hard to make this point important for the society.
‹
One solution: to find partners who are ready to invest in order to build
real system/application.
Ambient computing and industry(4)
‹ Examples
– The U-Japan project running in Japan
– Goals:
Ubiquitous
 Universal,
 User oriented,
 Innovative.

Perspectives industrielles (5)
‹
Exemple (2)-U-Japan
– Mainly based on RFID


Each object is equipped with an RFID tag or a RFID reader.
As soon as the object UID is read there is a connection to a server in order to
activate the associated process.
– Numerous experimentations (large scale aspects are taken into account)


NTT DoCoMo (Felica- micro paiement)
Pilots deployed in Tokyo (zoo, Ginza park,…)
Ambient computing and industry(6)
‹ Major
investments of the major telecom
companies. Nokia, samsung Sony Ericsson,
Apple…
New appliances (bluetooth, RFID/NFC)
 New services (geo localisation based, …)

– Operators (Orange, DoCoMo, …)

To define new services in order to generate more and
more bandwidth (remote server accesses).
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertising