2. Outline
Haptics Overview
Haptics for Mobile Devices
Localized Piezo Haptics
Haptics Characterization
Summary
Steve Dai, Haptics Purdue October 21, 2008
3. Haptics Overview
Definition: tactile feedback, more specifically active
feedback
Dominant Types of Active Feedback
Vibrotactile
Vibration sensed by nerves in the skin current focus
Kinesthetic
Sensations associated with body position, movement, or weight sensed by the nerves
in muscles, tendons, or joints
Steve Dai, Haptics Purdue October 21, 2008
4. Haptics Overview
Sample applications
BMW
mice I-drive
vibrating
gaming call alert
controllers
3D Design &
Interaction
medical training Braille displays
Steve Dai, Haptics Purdue October 21, 2008
5. Haptics Overview
Comparison with other senses
Neurons “Data” rate Temporal
(bits/s) Acuity
Touch ~106 102 5 ms
Hearing ~104-105 104 0.01 ms
Sight ~106 106-109 25 ms
Lynette Jones, MIT
Steve Dai, Haptics Purdue October 21, 2008
6. Haptics Overview
Frequency Dependence of Perception
SIGHT: 540 THz SOUND: 20~20,000 Hz TOUCH: 20~1000 Hz
Minimum audible=threshold
Fletcher-Munson curves, 1933
Verrillo-RT. Subjective Magnitude
Functions for Vibrotaction. IEEE
Transactions on Man-Machine Systems
MMS-11(1): 19-24. (1970
Steve Dai, Haptics Purdue October 21, 2008
8. Haptics for Mobile Devices
Why haptics?
“Data” rate Temporal Human to
(bits/s) Acuity device Input
Touch – 102 5 ms
fingertip
Audio – voice 104 0.01 ms
Sight – eye 106-109 25 ms
Steve Dai, Haptics Purdue October 21, 2008
9. Haptics for Mobile Devices
Value to users
FUNCTION SAMPLE USE CASES HAPTICS REQUIREMENTS
Vibe alert Simple vibration or patterns
Increasing Complexity
Notification
Confirma- Key press Short click-like response
tion status change Low latency
Localized preferred over global?
Amusement Base boost Rich variety of haptic contents
Entertain- Touchscreen UI event Visual/audio synchronization
ment Gaming “Essential but unnoticed” haptics
Communi- Sensorial Standards and infrastructure
cation communication “Hapton” capable phones
Steve Dai, Haptics Purdue October 21, 2008
10. Haptics for Mobile Devices
Global vs. Localized haptics
Global: vibrates the entire device
(phone)
Samsung F700 Samsung LG Voyager
Mot A1000 RAZR2 Mot & SCH-W559 Anycall 2008 (2008)
(2004) (2007) Krave (2007)
(2008)
Localized: vibration is localized to
an input surface (keypad, display)
ROKR E8 (2008)
Steve Dai, Haptics Purdue October 21, 2008
11. Haptics for Mobile Devices
Actuation Technologies
Global Localized
USE CASE &
IMPLEMENTATION
Rotary Linear Linear F Reactor Piezo
VIBE ALERT YES YES NO NO NO
KEYPRESS OK OK GOOD GOOD EXLT
ADVANCED HAPTICS LMTD LMTD OK LMTD EXLT
INTEGRATION NO NO LO LO MOD
COMPLEXITY
Steve Dai, Haptics Purdue October 21, 2008
12. POKR E8 Haptics
The review
“This new type of haptics really
works, and quite well. The effect is
best described as "spooky". It
works so well that if Motorola had
told us it had real keys under the
surface - and wasn't a touch
keypad at all - we would have
believed them and never doubted
it…. ”
Eric Lin, Eric M. Zeman and Rich Brome
January 7, 2008
www.phonescoop.com
Steve Dai, Haptics Purdue October 21, 2008
13. Localized Piezo Haptics
The challenge
“Morphing” keypad
= Display?
Lack of keyclick
feedback for key press
Sample photo placement
ROKR E8 ModeShiftTM keypad
Steve Dai, Haptics Purdue October 21, 2008
14. Localized Piezo Haptics
The solution
Direct bonding of piezo Proof-of-concept
element on phone
chassis behind keypad
Shrinkage/expansion of
piezo under electrical
field translates to
“buckling” motion of
keypad Force
haptics to fingertip sensing
resistor
Steve Dai, Haptics Purdue October 21, 2008
15. Localized Piezo Haptics
Why piezo?
Popple switch Optimized DC Optimized linear Piezoelectric
rotary motor motor actuator
Typical Popple Click (Press Only) Immersion VibeTonz vs. Click Linear Motor with Reverse Drive vs. Popple Click Piezo Actuator vs. Click
15 15 5
10 10
2.0 4.0 Popple Click Popple Click Popple Click
Acceleration (g) - Popple
Time (ms) Linear (Rev. Drive) Piezo Actuator
VibeTonz 5 5
Acceleration (g)
5
0 0
-5 5
-5
-10
MARGINAL -10
GOOD BEST
-15 5
-15
0 10 20 30 40 50 60 0 10 20 30 40 50 60
0 10 20 30 40 50 60 0 10 20 30 40 50
Time (ms) Time (ms)
Time (ms) Time (ms)
Global
Localized
Steve Dai, Haptics Purdue October 21, 2008
16. Haptics Characterization
What to measure for a keyclick?
Virtual tapping study (J Feine, Stanford Univ,
Haptics2006)
– Haptic simulations of tapping on a hard object
feels most real when hand acceleration
produced by virtual contact are matched to
those of contact with real object
Key click profile
– Push and release acceleration pulses 2.0
Time (ms)
4.0
– Each pulse: <5 ms, 40~100 g Accel_pp
Impact other than acceleration needs to be
further studied
Steve Dai, Haptics Purdue October 21, 2008
17. Haptics Characterization
Passive vibrotactile
Acceleration Comparizon Displacement Comparizon
0.6 10
SSRC SSRC
0.5
Average 8 Average
Dsiplacement (μ m)
Mountcastle
Acceleration (g)
Threshold
0.4
Mountcastle 6 1972
0.3 1972
4
0.2
0.1 2
0.0 0
0 100 200 300 400 0 100 200 300 400
Frequency (Hz) Frequency (Hz)
12 300
Comfort range
Max_g Max_d
10 Min_g
250
Min_d
Displacement ( μ m)
8 200
Accel_pp (g)
6 150
4 100
2 50
0 0
0 100 200 300 400 500 0 100 200 300 400 500
Frequency (Hz) Frequency (Hz)
Steve Dai, Haptics Purdue October 21, 2008
18. Haptics Characterization
Click threshold and comfort level
Test setup
Subjects control test processes
Up/down key to adjust voltage level
Space bar to register the desired voltage level
6 data points for each wave setting, 3 runs
starts from low and other 3 runs from high
Alternated test sequence on threshold and
comfort level for counter balance
Total 12 test subjects
12 levels at 2 dB a step
Steve Dai, Haptics Purdue October 21, 2008
19. Haptics Characterization
Waveforms and resultant acceleration profiles:
Step, sin300_1, sin300_2, sing150_1 and sing 150_2 generated in Audition
250 step w ave 250 Sin300_1 w ave 250 Sin300_2 w ave 250 Sin150_1 w ave 250 Sin150_2 w ave
200 200 200 200 200
150 150 150 150 150
Voltage (v)
Voltage (v)
Voltage (v)
Voltage (v)
Voltage (v)
100 100 100 100 100
50 50 50 50 50
0 0 0 0 0
-50 -50 -50 -50 -50
-0.005 0 0.005 0.01 0.015 0.02 -0.005 0 0.005 0.01 0.015 0.02 -0.005 0 0.005 0.01 0.015 0.02 -0.005 0 0.005 0.01 0.015 0.02 -0.005 0 0.005 0.01 0.015 0.02
Tim e (s) Tim e (s) Tim e (s) Tim e (s) Tim e (s)
15 step w ave p-p acceleratoin 15 15 15 15
sin300_1 w ave p-p acceleratoin sin300_2 w ave p-p acceleratoin sin150_1 w ave p-p acceleratoin sin150_2 w ave p-p acceleratoin
10 10 10 10 10
Acceleration (g)
Acceleration (g)
Acceleration (g)
Acceleration (g)
Acceleration (g)
5 5 5 5 5
0 0 0 0 0
-5 -5 -5 -5 -5
-10 -10 -10 -10 -10
-15 -15 -15 -15 -15
-0.005 0 0.005 0.01 0.015 0.02 -0.005 0 0.005 0.01 0.015 0.02 -0.005 0 0.005 0.01 0.015 0.02 -0.005 0 0.005 0.01 0.015 0.02 -0.005 0 0.005 0.01 0.015 0.02
Tim e (s) Tim e (s) Tim e (s) Tim e (s) Tim e (s)
Audio signal (Vpp = 0 ~ 2 V) to
20x piezo amplifier for threshold (max Vpp ~ 40 V)
100x Kepco voltage amplifier for comfort level (max Vpp ~ 200 V)
Wave forms played at both press and release of FSR
Steve Dai, Haptics Purdue October 21, 2008
20. Haptics Characterization
Click threshold 3
Threshold Acceleration
95% CI High
95% CI Low
Step wave
Threshold Accele_pp (g)
2 Mean
Accel_pp ~ 2 g
Single pulse? 1
Frequency effect 0
Step Sin150_1 Sin150_2 Sin300_1 Sin300_2
Accel_pp ~ 0.39 g at 150_1 Hz, Haptics
Accel_pp ~ 0.71 g at 300_1 Hz Vpp
20
95% CI High
Number of wave
Threshold Voltage (V)
15 95% CI Low
Mean
Accel_pp ~ decreases at double waves
10
Drive voltage 5
Voltage lower in double waves drive 0
Voltage lower as frequency goes up Step Sin150_1 Sin150_2 Sin300_1 Sin300_2
Haptics
Steve Dai, Haptics Purdue October 21, 2008
21. Haptics Characterization
Click “Comfort” Level Com fort Acceleration Level
25
95% CI High
Comfort Accele_pp (g)
Similar acceleration and voltage 20
95% CI Low
Mean
profiles as threshold 15
10
Frequency effect 5
Accel_pp ~ 3.2 g at 150_1 Hz, 0
Step Sin150_1 Sin150_2 Sin300_1 Sin300_2
Accel_pp ~ 6.7 g at 300_1 Hz Haptics
Com fort Voltage Level
Number of wave 150
95% CI High
Comfort Level Voltage (V)
Accel_pp ~ decreases at double 95% CI Low
100 Mean
waves
50
Drive voltage
Voltage lower in double waves drive 0
Voltage lower as frequency goes up Step Sin150_1 Sin150_2 Sin300_1 Sin300_2
Haptics
Steve Dai, Haptics Purdue October 21, 2008
23. Haptics Characterization
“Comfort” vs threshold levels
Accel_Cmf/Accel_Thr In dB (=20*log(Cmf/Thr))
Step 8.79 18.9
Sin150_1 8.29 18.4
Sin150_2 8.02 18.1
Sin300_1 9.44 19.5
Sin300_2 8.76 18.8
For current interactive click feedback test, the comfort level is
approximately 18~19 dB over threshold. The ratio is appx a constant, and
is independent of the wave forms, frequency and number of pulses
Steve Dai, Haptics Purdue October 21, 2008
24. Haptics Characterization
Issues of existing in-line test fixture
Fixture Hand
Frequency response
200 g wt on linear slider
Good tool for production line with good R&R
Discrepancy of fixture vs hand
Magnitude of pk-pk acceleration
Trend of frequency response
Steve Dai, Haptics Purdue October 21, 2008
25. Haptics Characterization
Literature -- human finger study
Multiple joints, muscles/tendons to actuate finger
Biomimetic finger: use SMA for actuation
* Vishalini Bundhoo and Edward J. Park, “Design of an Artificial Muscle Actuated Finger towards Biomimetic Prosthetic
Hands”, IEEE, 2005
Steve Dai, Haptics Purdue October 21, 2008
26. Haptics Characterization Rapid transient measurement
Literature – fingertip
Model
F (t ) = m&& + bx(t ) + kx(t )
x &
Findings
• Mass m: ~ 6 g
• Stiffness k: up linearly with force
• Damping b: large zero-f value, up linearly with force
* A. Haijun and R. Howe, “Identification of the mechanical impedance at the human
finger tip”, J Biomechanical Eng, Vol 119, P 109, Feb 1997
Steve Dai, Haptics Purdue October 21, 2008
27. Haptics Characterization
Internal Study -- finger impedance vs frequency
Model
Z(ω ) = F(ω ) x& (ω ) = (mω − k ω )j + b
16
14
12
|Z|, N-s/m
10
8
6
4
2
0
0 100 200 300 400 500 600
Frequency, Hz
2-spring system? 1 DOF?
• Both response to lower frequency
• One dominates at higher frequency
* C Fu and M Oliver, “Direct Measurement of Index Finger Mechanical
Impedance at Low Force”, World Haptics 2005
Steve Dai, Haptics Purdue October 21, 2008
Two springs?
28. Haptics Characterization
Improved test fixture – proof of concept
1 DOF
wt
Poron
A-meter
Probe
Phone
Springs
Schematic
• 2-spring system “Primitive” fixture
– Arm • stainless steel = spring 1
– probe • 200 g weigh: F @ contact ~ g
• 1 DOF at the joint • Accelerometer: MS ACH-01
• Weight and position to • Poron: Rogers 4790-92-15125-04 cellular
reach 150~200 g at the urethane foam, 3 mm, = spring 2
probe • Probe: Al block + screw with round tip
Steve Dai, Haptics Purdue October 21, 2008
29. Haptics Characterization
Results – Finger Press vs Fixture
Fairly good tracking
at 200, 250 and 300
Hz, nearly 1:1
Good tracking at all
voltage levels
Slightly off at 150 Hz
Overall the tracking is
very much improved
over the existing
inline test fixture
Steve Dai, Haptics Purdue October 21, 2008
30. Summary
Haptics is an essential functionality of mobile devices
Proper developed haptics could greatly enhance device
usability
Piezoelectric enabled localized haptics can provide a
nearly true keyclick experience
Haptics characterization is critical for technology
development
Steve Dai, Haptics Purdue October 21, 2008
31. Challenges
Haptics characterization
– Understanding of the physical parameter space responsible for a
wide range of tactile sensations that can be communicated from
a mobile device to the human hand and/or the wrist
– Provide psychophysical evidence for the design of tactile
patterns to be used with mobile devices
Haptics implementation
– Touchscreen solutions
– Novel UI enabled or enhanced by haptics in conjunction with
visual and audio effects
Steve Dai, Haptics Purdue October 21, 2008
33. Haptics Characterization
Voltage – Acceleration Calibration
Threshold Calibration Comfort Level Calibration
10 50
y = 0.216x - 0.033
Step R2 = 0.9996 Step y = 0.212x - 0.7498
sin150_1 sin150_1 R2 = 0.9986
8 40
Sin150_2 Sin150_2
Acceleration p-p (g)
Acceleration p-p (g)
sin300_1 sin300_1
6 Sin300_2 30 Sin300_2
Linear (Step) Linear (Step)
4 20
2 10
0 0
0 10 20 30 40 50 0 50 100 150 200 250
Voltage (v) Voltage (v)
• Observation
– Good linearity between Vpp to piezo and Accel_pp in all driving waveforms
• For sinusoidal wave, Accel ~ ω2*Displ ~ ω2*V
– Consistent slopes in both low and high voltage ranges
– Parameters from linear curve fitting are used to calculate the actual Accel_pp
Steve Dai, Haptics Purdue October 21, 2008
34. ACTUATORS
Multi-Function Transducer (MFT)
FOR SOUND
TOP PLATE COIL PROTECTOR1 DIAPHRAGM SUSPENSION FRAME
PROTECTOR2 MAGNET POLE PIECE SUSPENSION SPACER
FOR VIBRATION
Microelectronics & Physical Sciences Research