Upload
others
View
9
Download
0
Embed Size (px)
Citation preview
Computer Science and Engineering -‐ University of Notre Dame
Wearable Compu=ng
Steve Mann
• 1970s, pre-‐laptop, early computer era. • Building computers he could wear. • Inventor of wearable compu=ng.
Computer Science and Engineering -‐ University of Notre Dame
Steve Mann
• 1991: Started the ”Wearable Compu=ng Project” at MIT.
• 1995: World’s first covert wearable computer – camera and display concealed in ordinary eyeglasses.
• 1997: PhD from MIT in the field he himself had invented.
• Today: Works at University of Toronto.
Computer Science and Engineering -‐ University of Notre Dame
Steve Mann
Computer Science and Engineering -‐ University of Notre Dame
What is Wearable Compu=ng?
Seven aTributes of wearable compu=ng [Steve Mann, 1998]: 1. Unmonopolizing of the user’s aTen=on. User can aTend to other events. 2. Unrestric2ve to the user. Allows interac=on while user carries out normal
func=ons. 3. Observable by the user. As the system is being worn, there is no reason why
the wearer cannot be aware of it con=nuously… but this contrasts with 1! – BeTer phrasing: User can iden=fy computa=onal and non-‐computa=onal
components of their clothing.
4. Controllable by the user. User can take control at any =me. 5. A<en2ve to the environment. Can enhance the user’s environment and
situa=onal – Awareness.
6. Communica2ve to others. Can be used as a communica=ons medium. 7. Shares the same physical and situa2onal context as the user.
Computer Science and Engineering -‐ University of Notre Dame
Wearable vs. Ubiquitous
• Ubiquitous compu=ng – Computer/sensors embedded in the environment
• Wearable compu=ng: – Computers/sensors on people
• Complimentary
Wearable Compu=ng
intelligent, personal assistant • Proper=es
– Invisibly embedded into the ou]it – Context-‐aware – Always on and proac=ve – Connected
• Func=onality – Mobile access – Just in =me informa=on delivery – Personalized informa=on collec=on
• Architecture: ubiquitous, distributed system – Deeply embedded intelligent sensors network – Embedded communica=on and power genera=on – Electronic appliances as unobtrusive accessories
Research Areas
• Wearable computer architectures – heterogeneous, distributed architectures – context sensi=ve resource management – experimental wearable pla]orms
• Applica=on systems • Context recogni=on systems
– mul=modal sensor systems – context models and recogni=on
• Packaging for wearability – tex=le interconnec=ons – electronic miniaturiza=on
• Other – low power wireless communica=ons – display technology
display context sensor
array: camera, light, microphone, GPS
distributed reconfigurable computer body area
network: wireless
communication: WLAN, GSM
Hardware Features
• Light-‐weight (small) • Durable • Comfortable • Long baTery life=me • Easy to use • Affordable • Cool (invisible, hidden, disappearing)
Applica=on Features
• Person-‐to-‐person communica=on • Personal organiza=on/remembrance aid • Context awareness • Effortless usage
– natural, intui=ve
Big Picture
Application I/O
Communication Heat Power
I/O Interface
• Visual – Input: computer vision; cameras – Output: overlaying things
• Audio – Input: speech recogni=on, background noise separa=on, speaker iden=ty (voice fingerprint)
– Output: speech synthesis
Example: Eyeglass Display
• Human factors studies – Health and safety – Social acceptance
• Bi-‐ocular displays – Viewing 2D text and images
• 3D glasses
Big Picture
Application I/O
Communication Heat Power
Local Communica=on Examples
• Low range • Low power
• Transfer data between: – wearable and handheld – wearable and desktop – wearable and environment
Big Picture
Application I/O
Communication Heat Power
Power Requirements
• The tradi=onal bigger ones: 5 W – head mount display, 2GB hard disk, 133 MHz Pen=um, 20 MB RAM
• The improved smaller ones: 0.7 W – MicroOp=cTM eye-‐glass display, Flash memory (.5 GB), StrongArm
microprocessor (.3 W at 115 MIPs) – without communica=on
• May have to come from the person wearing the devices (hTp://www.youtube.com/watch?v=F0ck2Qynjrc)
Big Picture
Application I/O
Communication Heat Power
Heat Dispensing
• requirement – -‐15 intolerably cold – 15 -‐ 34 OK – 34 -‐ 39 hot – 39 -‐ 43 pain – 43 -‐ =ssue damage
Applica=on Areas
• Warehouse picking • Inspec=on • Maintenance • Repair • Security • Military (health, equipment, maps, terrain, infrastructure)
• Technicians (blueprints) • Field workers (remote experts) • Researchers
Applica=on Examples: Miniature Head-‐up Displays
Examples
hTp://www.versa=lemobile.com/hardware/motorola/mobile-‐terminals/ws.html
hTp://opera.media.mit.edu/levis/jacket.mpg hTp://www.media.mit.edu/galvac=vator/index2.html
Examples
• Wearables for the military: Future Force Warrior (FFW)
• Onboard physiological/medical sensor suite to accelerate casualty care
• NeTed communica=ons to maximize robustness and integra=on of small teams
• Embedded training (similar to mar=al arts example?)
• Enhanced situa=onal awareness (heads-‐up display?)
• Synchronized firing of weapons from team.
• Bone conduc=on technology: “talking and speaking without sound or hearing”
Computer Science and Engineering -‐ University of Notre Dame
Example
• Wearables for sports training • Karate trainees are instrumented with
accelera=on sensors. • Sensor data is translated directly into
sound output. • Trainees can now ‘hear’, as well as see
instructor’s movements. • Trainees can also hear themselves:
aTempt to match own sound to sound of instructor.
• Mar=al arts training is about reproducing paTerns over =me, not just matching sta=c poses; therefore, sound is a useful sensory s=muli to introduce to training.
• Result: Trainees with system tended to learn faster than trainees without system.
Computer Science and Engineering -‐ University of Notre Dame
Applica=ons – Augmented Memory
• Elderly or people with poor memory. – Remember name and face of people.
• Image processing can recognize a face and map it to the person’s name and affilia=on.
– How should it be presented?
Computer Science and Engineering -‐ University of Notre Dame
Applica=ons – Aiding the Visually Disabled
• User wears non-‐transparent glasses with integrated displays, experiences the world through a camera.
• Computer processed video stream. – Enhance contrast. – Adjust colors. – Night vision. – Enlarged view.
Applica=ons – Aiding the Visually Disabled
• Fisheye lense for reading text.
• Remapping around blind spots.
Computer Science and Engineering -‐ University of Notre Dame
Applica=ons – Addi=onal Vision Tricks
• ”Edgertonian” eyes – Freeze-‐frame effect, fast shuTer. – This enables
• Reading text on a =re of a speeding car. • Clearly seeing the rotor blades of a helicopter. • Coun=ng the number of bolts holding an airplane rotor together in mid-‐air.
• Plus lots of other interes=ng effects.
Applica=ons – Addi=onal Vision Tricks
• Giant’s eyes. – Enhances depth percep=on of distant objects.
Computer Science and Engineering -‐ University of Notre Dame
Applica=ons – A-‐Life, Avalanche Rescue
• Help rescue avalanche vic=ms. • Survival chance
– 92% aver 15 minutes. – 30% aver 35 minutes.
• Time is not the only factor – Orienta=on, head up or down. – Air pockets, air channels.
• How can wearables help rescuers?
Applica=ons – A-‐Life, Avalanche Rescue
• Each vic=m wears a small sensor. – Tilt sensor – Heart rate – Blood/oxygen satura=on
• iPAQ gets readings. – Sorts and priori=zes.
• Rescuers get advice on where to start digging. – Increased survival chances.
Computer Science and Engineering -‐ University of Notre Dame
Applica=ons – Social Sovware
Usually designed for urban sewngs. Interface to groups or individuals.
• Safety net • Friend finder • Familiar Strangers
Applica=ons – Safety Vests
• Social network of wearable users. • Biosensors monitor user’s body.
– Heart rate, perspira=on, breath rate. – Alert friends in case of abnormal values.
• Example – Start sending video in scary situa=ons.
• Such ”safety vests” exist already.
Wearable Cap=oning
Computer Science and Engineering -‐ University of Notre Dame
• Wireless, wearable personal cap=oning device to assist hearing-‐impaired individuals in interac=ng in a speech-‐centric sewng
Smart Shirt
Computer Science and Engineering -‐ University of Notre Dame
• “Smart Clothing” • Smart Shirt
– Vital signs and bullet wounds • Smart Arc=c Clothing,
– For ac=vi=es in harsh winter weather
Wayfinding
• Using a wearable device to help people navigate in their environment
• IR beacons for street crossing – Audio and tac=le cues
• Augmented Reality for Naviga=on – Visual cues from GPS and compass overlaid on world
Wayfinding
• Spa=al Language as Naviga=on Aid, Loomis[2002]
Remote Interfaces
• Gesture Pendant – Universal Remote Control via gestures
Telemedicine/Telehealth
• Gesture Pendant – Also could be used for medical monitoring
Challenges for the Wearable PC
• Seamless connec=on – across different kinds of
network • Occasional connec=on
– in and out of range • Local communica=on
– ad-‐hoc peripherals
n Modes of interac=on n visual and vocal
n Health and safety n strain on the
senses
n Unobtrusive n socially acceptable
Connectivity Usability
Situatedness
n Awareness n capturing context
n Interpreta=on n use of context
data
n Augmenta=on n personal assistant
Technical
Social