Transcript CSM-15_Lecture_10

Component-Based
Software Engineering
Oxygen
Paul Krause
Lecture 10 - Oxygen
Contents
 Introduction to the Oxygen Project
 Technological themes
 User technologies
 Conclusions
Beyond Objects,
Beyond Components
 Ubiquitous



Embedded devices
Mobile devices
Ad-hoc, dynamically changing networks
 Ambient



Computing
Intelligence
Sensitive and responsive
As effective away as at home
“Social” user interfaces
PERVASIVE, HUMAN-CENTRED COMPUTING
OXYGEN
&
HYDROGEN
MIT LABORATORY FOR COMPUTER SCIENCE
MIT ARTIFICIAL INTELLIGENCE LABORATORY
PHILIPS RESEARCH
Oxygen Partnership

Massachusetts Institute of Technology








Laboratory for Computer Science
Artificial Intelligence Laboratory
Acer (Taiwan)
Delta Electronics (Taiwan)
Hewlett Packard (USA)
Nippon Telephone and Telegraph Inc. (Japan)
Nokia Research Centre (Finland)
Philips Research (Netherlands)
Pervasive, Human-Centred
Computing
Pervasive, Human-Centred
Computing
 Computation
will be pervasive (“like
Oxygen in the air we breath”)


In hand-held and personal devices
In our homes, offices, cars…
 Each
device will adopt our “information
personality” when we interact with it
 We will interact with them using speech,
gestures, facial expressions
Requirements for Pervasive
Computing
 Must
be accessible anywhere
 Must adapt to change


In user requirements
In operating conditions
 Must

never shut down or reboot
Individual components may come and go, but
Oxygen must be available all the time
Addressing Human Needs
 Speech
and Vision technologies to enable
us to communicate with Oxygen
 Individualised access and collaboration
technologies
 Devices embedded in home, office and
cars to sense and affect our needs
 Hand-held devices to empower us to
compute and communicate wherever we
are
Technological themes
 Distribution

People, resources and services
 Semantic

content
What we mean, not what we say
 Adaptation

and change
To enable us to act in a dynamic world
 Information

and mobility
personalities
Privacy, security and form of interactions
Semantic content
 Time
flies like and arrow
 Fruit
flies like a banana
Necessary Information
 Information

Uses whole lifestyle to identify information
needs
 Adaptation

personality
and change
Information and advice will be tailored to
context
Devices and Networks
 Universally


accessible devices
E21s, embedded in the environment
H21s, Hand-held devices
 Adapt
automatically, or
 Modified explicitly
To address specific user preferences
Stationary Devices
 Linked



to local sensors and actuators
Monitor and change room temperature
Check and close doors
Redirect e-mail
 Can
act for us, even when we’re
thousands of miles away
But?
 How



do we:
Register sensors with all and only the relevant
listeners
Ensure actuators do not receive conflicting
requests
Safeguard the networks against attacks from
viruses, attempts to access secure
information, …
Hand-held devices
Multiple modes of
Communication
Serve as:
• cellular phones
• radios
• GPS
• television
• PDA
Configurable to a range of
communication protocols
May offload
communication and
computation to nearby
E21s
Networks
 Support
dynamically changing
configurations of devices
 Able to identify devices and services by
how we intend to use them
 Enable us to access information and
services securely and privately
 Enable dynamic creation of “collaborative
regions” - local; building-wide; district-wide
User Technologies
 Speech
and Vision
 Knowledge Access
 Automation
 Collaboration
Speech and Vision
 Main
modes of interaction will be spoken
language and visual cues
 Use of vision to augment speech
understanding



Facial expressions
Lip movements
Gaze
 Graceful
domains
switching between different
Knowledge Access
 Individualised
Knowledge Access
 Universal access to information


Content-based searches/manipulations of
data
Information stored in different formats: text;
graphics; video clips …
 Searches
must return small amount of
most relevant information
Automation
 Processes
for automating and tuning
repetitive information and control tasks
must be




Natural
Easy-to-use
Customisable
Adaptive
 E.g.
set up the house for your arrival home
 Set up equipment for a serious operation
Collaboration
 Enable
people to engage in group
activities

Secure “collaborative regions”
 Trace

group actions
Annotated trail of issues, decisions,
documents, conversations …
 All
entities in the traces linked in a
semantic web
(Their) Conclusions
The Oxygen technologies will lead to:
 Profound leap in human productivity
 Anywhere, anytime, people will be able to



Automate routine tasks
Access knowledge
Collaborate with others
 Significant
amplification of human
capabilities throughout the world