HWG 1

1
Pervasive Computing Facilitating Everyday
Environments

• Hans-W. Gellersen
• Lancaster University
• Department of Computing
• Ubiquitous Computing Research

2
1. Facilitating Everyday Environments

• Pervasive Computing / Ambient Intelligence
• Mass deployment of computing in everyday life
• At the periphery of peoples attention
• New Devices
• Information appliances, continuous service,
  contextualized
• Augmented Artefacts
• Leverage existing artefacts/structures
• Embed new services in the background of everyday
  activity

3
Facilitating Everyday Environments

• Design Perspective
• Focus on foreground activity / interaction with
  accustomed physical environments
• Introduce computing in the background
• Utilize the rich affordances of physical
  artefacts and structures
• Dont break with accustomed uses and familiar
  concepts
• Technical Perspective
• Computers as secondary artefacts
  (embedded/situated)
• Sensing/context to tie computers to the primary
  artefacts
• The primary artefacts are the user interface
• Networking to enable coherent smart environments
• Proximate networking, high density, very dynamic

4
Facilitating Everyday Environments

• Technology Research Challenges
• Embedded platforms for augmentation of artefacts
• Architectures for ad hoc composition of networked
  artefacts
• Current Related Projects at Lancaster
• Smart-Its (EC, Disappearing Computer Programme)
  A Platform for Digital Augmentation of Everyday
  Artefacts
• PinPlay (EC, Future and Emerging Technologies)
  A Networking Concept for Objects attached to
  Surfaces
• Load-Sensing Surfaces (EPSRC, Equator IRC)
  Lightweight infrastructure for tracking of
  objects and activity

5
2. Smart-Its

• Concept
• Post-hoc augmentation of artefacts
• Integration of sensing and perceptual computing
  to track artefact state/context
• Integration of wireless communication for
  artefacts to share context, and to interface
  existing networks/systems
• Technology Design
• Modular embedded device

6
Smart-Its

• Toward tighter integration of things with
  information

Information
Files
Databases
Objects
ActiveArtefcats
ManualRegistr.
Barcode scannen
RFID Tags
Things
Active Artefacts Dynamic state Self-controlled
presence
Barcodes Explicit Registration
RF Identification Implicit, unintentional
7
Smart-Its

• Smart-Its Background
• Mediacup Project, 1999-2001
• Demonstrator for Active Artefacts Model
• Mediacup Implementation
• Cup with embedded sensors, processor, wireless
  comm. and wireless charging
• Autonomous computation of user-level context
  (filled up, gone cold, ...) from sensor data
  
• Continuous digital presence
• gt95 reliable context prediction in everyday use

8
Smart-Its

• 1G Device Implementation
• PIC Microcontroller, RFM 868 MHz, Light, Audio,
  Accel., Temp. Sensors
• About 4x4x1 cm
• 150 Devices in use
• Revised design 4x1.5x0.5cm
• Application Research
• Ubiquitous computing prototyping
• Context proximity applications
• Aware Goods in enterprisemanagement
  (collaboration with SAP)

9
3. PinPlay

• Concept
• The wall as network bus for the things attached
• A new type of network to connect everyday objects
  on common surfaces such as boards and walls
• Use of familiar mechanism pinning objects to
  the wall pinning nodes to the
  network
• Components
• Surface Common surface augmented with conductive
  material to create a large-scale two-dimensional
  network medium
• Connectors pushpin-like physical connector for
  socket-less attachment of objects to the network
  medium
• Objects any type of device/object with embedded
  computing
• Network protocols not pre-determined, however
  has to support PinPlay behaviour

10
PinPlay

• A PinPlay Noticeboard
• Fully functioning prototype for proof of concept

11
PinPlay Noticeboard

• PinPlay Surface
• Corkboard augmented with two conductive sheets
• Ground layer on top, data/power layer hidden,
  cork as insulator
• Low cost, off-the-shelf,deployable at
  large-scale
• PinPlay Connectors
• Simple connector board with pushpin for two
  separate connection points

12
PinPlay Noticeboard

• PinPlay Objects
• 1-wire bus, Dallas MicroLAN
• 16300 bits/s
• PinPlay Objects
• Smart Notification Pin iButton and switchable
  LED
• Time-in-a-can iButton memory, internal calendar
  and clock

Switch
Switch
Time-in-a-Can
Time-in-a-Can
13
PinPlay Pinboard Scenario

• User interaction
• insert or remove pin
• network detects change
• protocol to determine pin with highest
  priority

• Network control
• External laptop connected as 1-wire network
  node
• runs network controller
• used to pre-set pins with priority and
  deadlines

14
PinPlay

• Technology Research
• Network surface development
• Simple and robust protocol design, zero
  maintenance
• Scalability and density (initial target 25
  nodes/sqm)
• Application Research
• Augmented noticeboards and other interactive
  surfaces
• Embedded home control buses
• Networking and free placement of controls (light
  switches, appliance controls etc.)
• Communication bus for wall-attached artefacts
• Clocks, calendars, sensors, digital picture
  frames,

15
4. Load-Sensing Surfaces

• Concept
• No physical thing can escape gravitation
• Use load-sensing as interface between the
  physical and the virtual
• Augment common surfaces (floor, tables, shelves)
  this is where gravitation pulls objects to
• Principle
• Augment surface at the corners
• Force applied (e.g. by weight of an object, or
  explicit pressure) is detected as load depending
  onposition of the pressure point
• i.e. surface detects weight/pressure and position

16
Load-Sensing Surfaces

• Weight Lab
• Various augmented surfaces
• Floor 240 x 180cm, up to 800kg load
• Larger table up to 200kg
• Coffee table up to 8kg, highly sensitive
• Shelves and trays

Floor with embeddedS-load cell
17
Load-Sensing Surfaces
Sensor board with wireless communication
Augmented tables
18
Load-Sensing Surface

• Context Acquisition
• Weight of objects
• Detection depends on sensor range (i.e. small
  weights not detectable on heavy-load surface)
• Application object identification (classes/
  instances)
• Position of objects
• cm-level accuracy
• Table can be pre-loaded
• Multiple objects can be positionedif placed
  non-simultaneously

19
Load-Sensing Surface

• Context Acquistion
• Beyond weight and position events derived from
  signal analysis over short time

New object
knocked over
Object put down
removed
20
Load-SensingSurfaces

• Context Acquistion
• Tracking of people/objects
• Prediction ofactivities

21
Load-Sensing Surface

• Surfaces as Interaction Device

22
Load-Sensing Surface

• Surfaces as Interaction Device

moveleft
moveleft
touch and move right
cup
book
click andrelease
click
click
23
Load-Sensing Surface

• Surfaces as Interaction Device

24
5. Summary

• Facilitating Everyday Environments
• Build on familiarity and meaning of existing
  artefacts / structures
• Introduce digital added value in the background
• Smart-Its
• Toward platforms for augmentation of everyday
  artefacts
• Close coupling between physical entity and
  information processing
• Pin Play, Load-sensing surfaces
• Examples for network/tracking infrastructure
  integrated with everyday use
• Low-tech, unobtrusive design