When Output meets Input…
Matthias RauterbergDepartment Industrial Design
Technical University EindhovenThe Netherlands, 2005
© Matthias Rauterberg, 2005
2/34
Matter
Information Energy
Attention
Gravity
Access?
Most Important Design Constrains
© Matthias Rauterberg, 2005
3/34
Output Space (system’s view)– The physical space where the
user’s [visual] attention is.Perception Space (user’s view)
Input Space (system’s view)– The physical space where the
user’s motor actions are.Action Space (user’s view)
input space= action space
output space= perception space
Visual Output Space and Input Space
© Matthias Rauterberg, 2005
4/34
How to design visual Output Space and Input Space in 2D?
MSc Thesis (1993) from Christian Cachin
Ref: Rauterberg, M. & Cachin, C. (1993). Locating the primary attention focus of the user. Lecture Notes in Computer Science, vol. 733, pp. 129-140.
© Matthias Rauterberg, 2005
5/34
Signal Detection Experiment (SDE)
time (ms)0 500 1000
circles
circles + signal (X or square)
N = 19; 11 women and 8 men took part in the experiment (mean age: 33 ± 14 years). 12 subjects were students of computer science at the ETH.
Dual task approach: (1) count circles, (2) detect signal X (given a distractor [])
Standard computer display: 14 inch, black&white
© Matthias Rauterberg, 2005
6/34
SDE Results: primary task‘Circle Deviation’ CD as a measure for task accuracy:
CD = |#CIRCLEScounted – #CIRCLESpresented| * 100% / #CIRCLESpresented
I II
III IV
CD=6.1% CD=6.8%
CD = 4.4%CD=6.9%Main Results:Quadrant IV outperforms all others
© Matthias Rauterberg, 2005
7/34
SDE Results: secondary task
0 3 6 90
10
20
30
40 [X] vs [no signal][X] vs [no signal or square]
Distance (inch)
‘Error Ratio’ ER:ER = (b + c) / (a + d) * 100%
Signal Detection Table:
© Matthias Rauterberg, 2005
8/34
Eye Recording Experiment (ERE)
PC
video camera
video recoder
NAC Eye Mark Recorder IV
mouse
projection screen overhead display subject
How to determine automatically the actual position of the user’s visual attention focus on a computer screen?
Subjects:N=6: 2 women and 4 men5 subjects were students of computer science at the ETH. 1 subject studied psychology at Uni Zurich.
Tasks:(1) Computer game; (2) Text formatting;(3) Hypertext navigation.
© Matthias Rauterberg, 2005
9/34
fixation region:= circle around fixation point with r = 3 inch
(1) without mouse operations: Mouse position in fixation region for 25% - 70%
(2) with mouse operations: Mouse position in fixation region for 49% - 97%
ERE Results
© Matthias Rauterberg, 2005
10/34
Design Recommendations
Outputspace
Input space
(1) Place e.g. the message left above the actual user’s focus of attention;(2) Place this message maximal 3 inch away of actual mouse position.
OS := Output SpaceIS := Input Space
Results:α ~ 135 grad∆ |OS-IS| < 3 inch
α∆
© Matthias Rauterberg, 2005
11/34
The Digital Desk from Pierre Wellner in 1991
Pierre Wellner
[00:05:88]
© Matthias Rauterberg, 2005
12/34
• Output Space– The physical space
where the user’s visual attention is.• Input Space
– The physical spacewhere the user’s motor actions are.
• Design Principle:– output space and input space must
coincide! [∆ =0]– “Interlacing the display and
manipulation space”(Djajadiningrat, 1998, TU Delft)input space
output space
∆
How to design Output Space and Input Space in 3D?
© Matthias Rauterberg, 2005
13/34
Tic-Tac-Toe with four interaction styles
© Matthias Rauterberg, 2005
14/34
[00:01:54]
User reactions with the DigitalPlayingDesk (1995)
© Matthias Rauterberg, 2005
15/34
Empirical Results : game playing time per dialog technique
0
50
100
150
200
250
300
350
400
450
500
Cel
l Mea
n fo
r tim
e of
use
r (s)
Cell Line Chart for "playing time"
Grouping Variable(s): Interface type
Error Bars: ± 1 Standard Deviation(s)
CI MI TI DPDP<.001P<.01P<.001
P<.001P<.001
P<.001
© Matthias Rauterberg, 2005
16/34
computer win
remis
user win Cell Line Chart for "winning chance"
Grouping Variable(s): Interface type
Error Bars: ± 1 Standard Deviation(s)
CI MI TI DPDP<.001P<.080P<.020
P<.802P<.001
P<.007
Empirical Results : winning chance per dialog technique
© Matthias Rauterberg, 2005
17/34
Design Recommendations(1) Output space and Input Space should maximally intersect.(2) Tangible interaction props do not interfere with primary task.
OS := Output SpaceIS := Input SpaceWS := Work Space
Result:WS = OS ∪ IS with max [OS3
visual ∩ IS3visual ]
© Matthias Rauterberg, 2005
18/34
The Build-It SystemFjeld, Bichsel & Rauterberg 1997
PhD Thesis (2001) from Morten Fjeld
[00:02:44]
© Matthias Rauterberg, 2005
19/34
Build-It: tangible interaction props
2D
3D
© Matthias Rauterberg, 2005
20/34
Interaction Props: user studyProps design factors:form, size, material and metaphor:
•An experiment was carried out to explore different design strategies.
•Tasks were based on initial planning of an interior architecture.
•Focus of the experiment was subjective opinion (n=12) about the bricks.
•The bricks were ranked by user performance before (first number) and after(second number) task solving activity.
© Matthias Rauterberg, 2005
21/34
Build-It: Spin-outs in Europe, Canada & JapanIPO
OMRON, Japan (2004)
© Matthias Rauterberg, 2005
22/34
[00:06:14]
Build-It: The PlanningTable in Japan by OMRON Inc.
© Matthias Rauterberg, 2005
23/34
Interaction ModelsBrygg Ullmer & Hiroshi Ishii, 2000
model
control view
INPUT OUTPUT
physical
digitalmodel
control Non graspablerepresentation
graspablerepresentation
INPUT / OUTPUT
© Matthias Rauterberg, 2005
24/34
How to design interaction props for 3D navigation?
PhD Thesis (2004) from Sriram Subramanian
2D interaction: 3D interaction:
© Matthias Rauterberg, 2005
25/34
VIP-3: Tangible Interaction PropsAliakseyeu, Subramaniam, Martens & Rauterberg 2002
[00:06:22]
© Matthias Rauterberg, 2005
26/34
Trends in Interactive System Technology
Mobile computing
Ambient rooms and Cooperative buildings
Transport [00:06:25]
© Matthias Rauterberg, 2005
27/34
Design Metaphors
Channel
Tool
Substitute
long time ago 2000 history
© Matthias Rauterberg, 2005
28/34
Trend in Interface Design
0
50
100
150
1970-1990 1990-2010 2010-
SW controlsHW controls
© Matthias Rauterberg, 2005
29/34
time1900 2000
mechanical style
electronic style
mechatronic styleDesign Styles
© Matthias Rauterberg, 2005
30/34
time1900 2000
mechanical style
electronic style
mechatronic styleactive forms(smart memory alloys)
given forms(ubiquituous computing)
channel forms(e.g. PC, TV, Radio, etc)
connected forms(ambient intelligence)
dedicated forms(e.g. typewriter, etc)
Design Forms
© Matthias Rauterberg, 2005
31/34
HomeLab: The Memory Browservan den Hoven, Eggen & Rauterberg, 2003
[00:01:09]
RFID tagged souvenirs
© Matthias Rauterberg, 2005
32/34
Interaction Props with Active Form
unloaded state loaded stateNitinol tubes
© Matthias Rauterberg, 2005
33/34
If Output meets Input:interactive holography might be the final solution
projection
smoke curtain (e.g. FogScreen)
© Matthias Rauterberg, 2005
34/34
Thank you for your attention.