1
Pervasive & Ubiquitous Computing
Introduction
Hao Chu (朱浩華 )
2/22/2005
2
Outline
• Faculty Intro• Student Intro• What is Pervasive / Ubiquitous Computing?• Course Topics• Course Format• Course Projects• Grading
3
Faculty Intro
• Ubicomp Lab, i-space Labs• New associate professor (8/1/2003)• Education:
– PhD (1999), Computer Science, University of Illinois at Urbana Champaign
– BS (1994), Computer Science, Cornell University
• Previous Work Experience: – Xerox – Intel– NTT DoCoMo USA Labs
4
Students Intro
• Please tell us about:– Grade level– Current (future) research area, interests, and faculty advisor– Background (rating: good, fair, none)
• English (reading, writing, and speaking)• Programming skills (C/C++, Java)• Hardware• Systems (OS or distributed)• Networking• User Interface• Vision
– Why are you interested in this course?
5
Terminology
• What is pervasive & ubiquitous computing (ubicomp)?– Ubiquitous = 到處存在的 ;遍在的– Other names: invisible computing, context-aware com
puting, everyday computing, ambient intelligence, embedded interactive computing, etc.
• How to realize ubicomp?– Move beyond desktop computing.– Embed computing into everyday objects.– Integrate (seamlessly connecting) physical objects wit
h virtual environment.– Networking everyday objects.
6
Toward Smart Everyday Objects
• Door can greet you by name upon entering.
• Wall can sense temperature, humidity, lighting, and adjust air conditioning, de-humidifier, lighting accordingly.
• Calendar can tell you meeting schedule.
• Pencils can record everything you write.
• Book shelf can tell you the location of the book/paper you need.
• Shoes can tell you where you have walked to.
• Refrigerator can offer recipes and dietary recommendation.
• Clothes can show the latest fashion or monitor your physical/mental health.
• Medicine cabinet can remind you when to take medicines.
• Dresser can give you fashion advices.
• Washing machines and dryers adjust to washing & drying instructions on dirty clothes.
• Credit card will warn if you are spending too much money.
• Smart chair, smart dining table, etc.
7
Ubicomp is the Future …
• “The (Computing) World is not a desktop!” ... Mark Weiser
• Ubicomp is how computing will be used in the future:– in everyday activities – invisibly through embedding in the physical objects
(requiring little user attention)– to create smart, everyday objects through
interconnections and cooperation (with other smart objects)
8
Related Fields of Ubicomp
• How does Ubicomp come about? (the evolution path)– Distributed Computing (PC + networks)
• Challenges: performance, scalability, server or network failures, open networks, performance,
– Mobile computing (Mobile devices + wireless networks)
• Challenges: resource-limitation, unpredictable network, power
– Ubicomp (Everyday objects + wireless networks)• Challenges: understanding user intention, heterogeneous
ubicomp environments, invisible user experience, more ..
9
Course Topics (Tentative)
• Vision & challenges• Software
infrastructure • Sensors • Context-aware
computing• Security and privacy
• Human experience
• Ubiquitous data access
• Coping with uncertainty
• Social computing• Project Aura• Project Oxygen• (Wearable
Computing)
10
Course Objective
• To prepare us (students and faculty) for research in ubicomp.
• (Try to) duplicate experience from similar courses taught at MIT, Stanford, CMU, and Georgia Tech.– Learning by reading papers
• Define problems & challenges• Understand state-of-art techniques & solutions• Identify limitations of state-of-art solutions
– Learning by doing projects• The project must have a research component.
11
Collaborative Learning
• This is a research seminar course, so everyone (faculty and students) will contribute to the learning process.– Paper discussion– Paper presentation– Project presentation
• Ubicomp is a new, fast changing field, so faculty may not know all materials!
12
(Unusual) Course Format
• Each lecture will discuss 4~5 papers on a specific topic.• For each paper,
– All will write a paper review before the class– Presenter will give an overview of the paper for 20 minutes.– Everyone will join the discussion for 10 minutes.
• Students will sign up for papers for presentation.
13
How to read a paper?
• For each paper, try to answer the following questions:– What is the problem?– What is the current state-of-the-art?– What is the key make-a-difference (new) method and
technique?– What is good/bad/ugly about this make-a-difference
method? – What has actually been done?– What is the future work?
14
Reading in Depth
• “Efficient Reading of Papers in Science and Technology”, Michael Hanson & Dylan McNamee
• Must challenge what you read!• Attack the paper (use your common sense)
– Are assumptions reasonable?– Is the method similar to other methods in related work?– Is the improvement marginal or significant?– Are arguments logically sound? – Are evaluation metrics reasonable?– Is conclusion drawn logically from measurements?
15
Develop Critical Thinking
• Critical thinking is "the examination and testing of suggested solutions to see whether they will work." Lindzey, Hall, and Thompson, 1978.
16
SWIFI
• We will setup a swifi website (collaborative website).– The assigned presenter must post presentation slides
and paper summary on swifi course page before the lecture starts.
– The assigned presenter will post a discussion summary on the swifi course page after the lecture ends.
17
Next Week Reading
• Mark Weiser. "Some Computer Science Problems in Ubiquitous Computing." Communications of the ACM, July 1993.
• Mark Weiser. “Some computer science issues in ubiquitous computing.” Communications of the ACM, 36(7):75-85, July 1993.
• Mark Weiser, John S. Brown. "The Coming Age of Calm Technology." 1996.
• M. Satyanarayanan. "Fundamental Challenges in Mobile Computing", Fifteenth ACM Symposium on Principles of Distributed Computing, May 1996.
• M. Satyanarayanan. “Pervasive Computing: Vision and Challenges”, In: IEEE Personal Communications. Carnegie Mellon University. (2001).
18
Reading Sources
• IEEE Pervasive Computing Magazine• ACM Ubicomp (2002, 2003, 2004)• Pervasive (2002, 2003)• IEEE Percom • ACM Mobisys, ACM Mobicom, ACM Mobihoc, ACM Sen
sys, SOSP, etc.• Can also be in any system & networking conferences, UI
conferences, etc.
19
Project Component
• Rapid research prototype of an ubicomp application in one semester.
20
Lifecycle of a Research Project
• Define motivation scenario (Tell an interesting story)– Emphasize the parts of scena
rio where it is currently not possible, but with your idea, it will become possible.
– Show me a proof-of-concept demo prototype
• Derive problem(s)– Assumptions, requirements, i
mplementation vs. research problems
• Survey related work
• Design solution(s) (new method, concept, and SW abstraction) – Differentiate your work from re
lated work– Must answer two questions:
What’s new? Why is it significant?
• Rapid prototype implementation
• Evaluation of Prototype Implementation (Experiments, user studies)
• Write a paper
21
Project Phases
• Phase 1: project idea presentation– Fun, realizable within one semester time framework
and computing equipments, has a research component.
• Phase 2: project proposal document– Form teams, define goals, plan, and needed
equipments.• Phase 3: project working prototype and final
report– Working prototype demonstration– Project report detailing motivation, objective, related
work, design, implementation, and evaluation.
22
Project Ideas
• References:– Georgia Tech “Mobile & Ubiquitous Computing” (Gregory
Abowd)• http://www.cc.gatech.edu/classes/AY2003/cs7470_spring/
– MIT Pervasive Computing (Larry Rudolph)• http://www.cag.lcs.mit.edu/classes/6.898/projects.htm
– CMU “Mobile Computing Systems & Applications” (Satya)• http://www-2.cs.cmu.edu/~15-821/
– Stanford “Mobile and Wireless Networks & Applications” (Mary Baker)
• http://www.stanford.edu/class/cs444n/projects.html
• IEEE Pervasive Magazines– Education & Training articles by Scott F. Midkiff
23
Project Idea (1) Weighting and RFID surfaces
• Ergo-chair: it can detect if a person is sitting in the correct posture
• Emotion-chair: it can detect a person’s mental state by the way how he/she sits on the chair
24
Project Idea (2) Pollution monitor
• Monitor your exposure to pollutants – Toxic Air– Acid rain– Toxic drinking water– Food chemical or bacteria– Environmental radiation (SUN)– Radiation cell phones & WLAN
25
Project Idea (3) Games
• Players’ physical context drives the games.– Players interact with the games and other players in both
physical and virtual spaces.– Physical world interactions are the “cool/innovative aspects”
(differs from PC-based games).• “Hide and Seek” (Nottingham):
– Runaways and policemen carry GPS-enabled cell phones in a city. Both can see their “approximate locations” on a city map. Policemen run around in a city to catch the runaways (caught if come within some small distance).
• “Ubicomp Doom” (MIT):– Game is projected on the wall.– No mouse and keyboard -> player’s physical movement moves
the virtual player.
26
Project Idea (4) Healthcare
• Smart refrigerator• Smart chopsticks• Smart robes• Smart pads
27
Project Idea (5) Everywhere Display
• Projector + mirror
28
Project Idea (6) Personal Experience Computing
• Lots of photos from your digital camera +• Digital photograph frame +• Context information
– Reminder– Mood pleaser– Reliving memory
29
Other project Ideas
• Interactive arts• Intelligent things• Indoor location systems• Location-based services• Security & Privacy protection• Using RFID• Anything …
30
Ubicomp Toys I
• Nokia and Windows CE Smart phones
• HP IPAQ 5500 (Bluetooth + 802.11)
• IPAQ Accessories: expansion pack, camera, memory card, GPS, GPRS, ..
• Sensors: light, tilt, temperature, accelerometer, pressure, weight, orientation, ultrasound, etc.
• Philgets RFID kits• Berkeley Motes
31
Ubicomp Toys II
• Passive RFID tags & readers (Skyetek and Alien)• Biometric sensors (heart rate, BVP, GSR, etc.)• Projector• WiFi based location systems (ekahau.com)• Ultrasound-based positioning systems (Navinotesd)• Use your imagination
– Anything you can find on the network or from your own lab …– However, budget may be limited
32
Grading (Tentative)
• Class Participation (30%)– Paper presentation– Paper review – Paper discussion
• Project (70%)
33
Course Requirements
• Class attendance (MUST!)– Please don’t register if you cannot wake up in the morning.
• Good programming skill• Some systems & networking background• Some creativity• Some research experience (if you are not my students)• Willing to spend extra time & efforts than what a regular
course would need• M1 and undergrads, please come & talk with me after
the class.
34
Course Homepage(is not up yet.)
I will setup a link from my homepage
http://www.csie.ntu.edu.tw/~hchu
(Click on course link)
Check on Wednesday
Watch for Course Announcements
Download papers
35
Thank you