Transcript slides
Marked Spaces
Interaction design and sensor integration for
ubiquitous location-based messaging and
communication
Mark Perry and Caroline Shangar
SISCM, Brunel University
contact: [email protected]
marked spaces
Communication in place
Many common instances of space in communication
notes on fridge doors and colleagues’ desks, paper pushed
under doors, pinned up in social areas for viewing at leisure
other instances where desirable (e.g. the bedtime SMS)
Location has an important role to play
comm’s technologies have largely failed to address this
“anytime, anywhere” vs. “now, here”
current communication and information is tied to devices, not
places and things
e.g. email, digital music and photographs, etc.
marked spaces
Previous work
Commotion (e.g. Schmandt, et al., 2000)
voice-based interaction on GPS platform
simple functionality: to-do lists, augmented with limited
location-based information services (context-aware
messaging)
Geonotes (e.g. Espinoza et al. 2001)
focus on volume problems in spatial messaging
based just on a single platform (GPS or Wireless LAN)
Urban Tapestries (e.g. Lane, 2004)
situated, public annotations, to build ‘urban histories’
map-based interaction, focus on ‘place’
user interaction simplified for the task with little
user customisation
All just stationery ‘locations’, all single sensor types
marked spaces
Designing to mark space
our goal:
not just locating in, but actively marking space
the space around things, as well as Cartesian space
utilise notion of proximity: ‘on’, ‘near’ or ‘around’
user-centred perspective
development efforts focus around use and
usability
marked spaces
Interaction and sensor
specificity
Three key granular areas of coverage
different capabilities and constraints
together - rich set of infrastructural resources allowing
determination of a device’s location at a different exactness
1. Microspatial (e.g. RFID tag)
coverage of a few cm to metres - precise sensitivity
un-powered, and allow ‘fit-and-forget’ placement
2. Mesospatial (e.g. ultrasound beacons/Bluetooth)
coverage of several metres (<5-100metres), powered
signal blocked by structures, allowing room-level sensitivity
3. Macrospatial (e.g. GPS)
Accurate to approx. 3-100 metres, powered
carried by the user, not embedded
marked spaces
‘Relative’ sensors (e.g. RFID, B/T)
Bluetooth
Bluetooth
RFID
RFID
<Sender: james>
<tag: john’s mobile>
remember to speak to
John - he’s got your
briefcase with him!
<Sender: caroline>
<tag: no-name>
Book due back in
the library on the
24th Feb 2005.
Check out the
reference to UbiNet!
marked spaces
‘Absolute’ sensors (e.g. GPS)
GPS tagged ‘boundary’ around school
(an X,Y co-ordinate + n metres)
<Sender: jim>
<tag: grid-reference>
ha, ha! get this!
<device plays music:
School’s out for the
summer!>
marked spaces
3 key research elements
1. understand the use of space in
messaging activity
support user needs
mesh with current practices
2. develop interaction
techniques
3. prototype technical sensor/
messaging infrastructure
lightweight (think SMS)
interactionally integrated
(multi-sensor)
personalisable
integrates a variety of sensors
into a single, unified system
marked spaces
Understanding space
…actually, understanding spatial communication
in particular, asynchronous messaging
little understanding of the role of space in communication:
why leave messages in spaces; what level of location specificity
is important; how are situated messages made visible?
need to examine current practices and values
role of spatial messaging in co-ordination and collaboration:
communication doesn’t occur in a vacuum
how do people use messages to communicate and maintain
relationships?
and communication breakdowns as current
practices fail
marked spaces
Interaction design
interaction model: text ‘tagging’ by Graffiti artists
‘sprayed’ onto people, places and things
extended to ‘media tagging’ (voice recordings, photographs,
music)
sensitivity at different levels of precision
how-to-do messaging at different target proximities…
…on an integrated system - with a single user interface?
interaction designer’s task:
low-effort message placement and retrieval
allow users to utilise the different resources and
constraints that the different sensors offer
without requiring a deep understanding of
underlying technologies
marked spaces
a complex design syntax!
(for
sender)
…compare to SMS…
present user with list of
available sensors
option <select
location/sensor>
menu selection:
sensor type, name,
other details
select <create new
message>
select media type
menu selection
create message
option <send>
menu selection, send
is the first option
select recipient/s
end
marked spaces
Interaction design issues…
Questions - how to do:
targeting (sender’s selection of a communication ‘space’)
addressing (how can spaces/sensors be named/identified?
as mobile or static? as personal or situated?)
filtering (which messages are relevant to this recipient?)
navigation (how to search through several messages?)
message editing (how to manipulate message settings, e.g.
editing message duration or deletion of old messages?)
system personalisation (how to configure the system for
ongoing needs?)
…compounded by the constraints of developing
on a handheld/small screened device.
marked spaces
Summary…
answer question: why might marking people, places
and things be useful?
how adopted in current patterns of activity, or
appropriated for new reasons?
what values might this form of communication hold for
users?
where might it fit into our current communications
ecology?
how to achieve this design is a non-trivial task!
its not just a matter of building the infrastructure…
…how can it be made useful and usable?
marked spaces
Simplified architecture
sensors ‘ping’ back a
unique identifier which
connects to the remote
message repository
LIM
(Location
Identifi cation
Module)
location data
location request
messages
messages and related ids
database
received messages,
devices in the area (ids)
Link
Processor
messages, related ids
User
Interface
marked spaces
High level architecture
location co-ordinates
e.g. GPS
Medium Range
Environmental
Sensor
e.g. B/T
device search
sensor poll
Portable Device
user id
tag id
sensor poll
Long Range
Environmental
Sensor
location info
location request
Long Range
Integrated
Sensor
Short Range
Integrated
Sensor
Portable Device
user id
unique
sensor id
unique tag id
unique
sensor id
e.g. RFID
Short Range
Environmental
Sensor
marked spaces