1 1 Spatialized Haptic Rendering: Providing Impact Position Information in 6DOF Haptic Simulations Using Vibrations 9/12/2008 Jean Sreng , Anatole Lécuyer, Claude Andriot, Bruno Arnaldi [email protected]
1
1
Spatialized Haptic Rendering: Providing Impact
Position Information in 6DOF Haptic Simulations
Using Vibrations9/12/2008
Jean Sreng, Anatole Lécuyer, Claude Andriot, Bruno Arnaldi
2
2
Introduction
• Context: Manipulation of solid objects in Virtual Reality
• Applications: Industrial virtual assembly / disassembly / maintenance
3
3
Enhancement of the information of contact
• Use of visual cues of contact (Sreng, Lécuyer, et al., IEEE TVCG 2006)
• Use of auditory cues of contact (Sreng, Lécuyer et al., ACM VRST 2007)
4
4
Haptic rendering of the information of contact
• Use of 6DOF haptic devices
• Computation of contact force : haptic rendering• Collision detection
• Force feedback
5
5
Limits of haptic rendering ?
• Importance of contact information
• In a real world • In a virtual world
6
6
Spatialized haptic rendering
• We propose to superimpose the vibrations corresponding to the 3D contact position to the classical haptic rendering
7
7
Outline
• Spatialized haptic rendering
• First experiment: Determining the optimal vibration parameters
• Second experiment: Preliminary evaluation
• Conclusion
8
8
Haptic rendering of contact position
• The impact between objects:• A reaction force
• A high-frequency transcient vibrations
• This high frequency transcient vibrations depends on:• The object material (Okamura et al. 1998)
• The object geometry
• The impact position
9
9
Haptic rendering of contact position
• How can we use vibrations to convey impact position information ?
• Different vibration models can be used (Sreng, Lécuyer, et al., EH 2008)
• Realistic model of a vibrating cantilevel beam
• Simplified model
10
10
Simulation of vibrations: Realistic model
• Realistic simulation based on the Euler-Bernouilli model
• General solution
11
11
Simulation of vibrations: Simplified model
• Simplified patterns based on the physical behavior based on an exponentially damped sinusoid:• Amplitude changes with the impact position
• Frequency changes with the impact position
• Both Amplitude and frequency changes
• Main benefits:• Easier perception
• Simplified computation
12
12
Am
Fr
AmFr (Consistent)
Near impact Far impact
Simplified vibration patterns
13
13
6DOF Spatialized haptic rendering
• Generalizing the previous approach for 6DOF manipulation:• Virtual beam model
14
14
6DOF Spatialized haptic rendering
• Two types of information can be conveyed:• A distance of impact
• A direction providing the orientation of the impact
15
15
6DOF Spatialized haptic rendering
• The impact force
• The wrench sensed by the hand
16
16
6DOF Spatialized haptic rendering
• The wrench is then modulated by the vibration model:
• In particular the vibration torque can be expressed:
Vibration pattern Vibration orientation
17
17
Manipulationpoint
Manipulationpoint
Manipulation point and circle of confusion
• Different impact positions can generate the same haptic feedback
18
18
6DOF Spatialized haptic rendering
• The total torque applied to the device is obtained by superimposing:• The classic torque obtained with closed-loop rendering
method
• The impact vibrations torque
19
19
Outline
• Spatialized haptic rendering
• First experiment: Determining the optimal vibration parameters
• Second experiment: Preliminary evaluation
• Conclusion
20
20
First experiment: determining the optimal vibration parameters
• Objectives1. Perceptual study : “Is it possible to perceive the contact position
in 3D space? Is it possible to perceive the vibration direction ?”
2. Technological aspect : Determine the optimal range of amplitude / frequency parameters
• Tests among: ( 4 amplitudes a ) x ( 4 frequencies f )
• Population: 10 male subjects (22 – 27 years old)
0.005 rad to 0.02 rad 12 Hz to 40 Hz
21
21
Apparatus
• Haptic device: Haption Virtuose6D 35-45
• Vibrations applied around the 3 axes
• Update rate of 1kHz
22
22
Procedure
• “On which axis is the vibration applied ? Where is the impact located ?” ● ● ● (3AFC)
• 15 blocks of 4 x 4 x 3 = 48 vibrations: total of 720 trials (35min)
●
●
●
23
23
Results: Effect of frequency
• Average performance : around 80% of correct responses
• Best performances achieved with low frequencies
4 frequencies f12 Hz 18 Hz 27 Hz 40 Hz
24
24
Results: Effect of amplitude
• Average performance : around 80% of correct responses
• Best performances achieved with high amplitudes
4 amplitudes a0.02 rad 0.015 rad0.01 rad 0.005 rad
25
25
Discussion
• Participants were able to perceive the vibration directions among the three axes
• Most participants reported that they did not use a particular strategy
• Best performances were achieved with low frequencies• Suggest the importance of kinesthetic cues over tactile cues
• However some participants reported that they relied on tactile cues for small amplitudes
• Best performances were achieved with high amplitudes• However some participants reported that high amplitudes made
their perception more difficult
26
26
Outline
• Spatialized haptic rendering
• First experiment: Determining the optimal vibration parameters
• Second experiment: Preliminary evaluation
• Conclusion
27
27
Second experiment: perliminary user evaluation
• Objective: Subjective evaluation of Spatialized Haptic Rendering in a real case
• Population: 11 naive subjects (8M, 3F) (25 – 43 years old)
• Task: 6DOF manipulation of 3D object• Subjective ratings
- Realism of the impact
- Feeling of impact position
- Overall comfort of the manipulation
• Procedure: Participants were asked to test the two rendering techniques successively (without and with vibrations) in a random order
28
28
Procedure
• Virtual scene: two 3D objects
• Spatialized Haptic Rendering parameters:
0.005 rad 0.02 rad
40 Hz 15 Hz
Manipulation point
29
29
Results
• Mean ratings from 1 (worst) to 6 (best) without and with superimposed vibration
30
30
Discussion
• Better feeling of impact position obtained with Spatialized Haptic Rendering
• Several participants spontaneously reported that the vibrations enabled them to perceive the impact position
• Most participants pointed out that they perceived different materials between the two conditions• A « crisper » or « harder » feeling with vibrations (Okamura et al. 1998,
Kuchenbecker et al. 2006)
• A feeling of « vibrating metal »
• Most participants enjoyed the manipulation using the vibrations
• Two participants (familiar with haptic rendering) reported that they associated the vibration with an unstable, an potentially harmfull rendering algorithm
31
31
Conclusion
• We proposed a 6DOF spatialized haptic rendering method to provide impact position directly on the haptic channel• using vibrations based on a vibrating beam
• We conducted two experiments to evaluate this method• Experimental study on the perception of vibration direction
- Participants can identify the vibration direction, i.e., the position of contact in 3D space
- Optimal range of model parameters: Low frequencies / High amplitudes have better results
• Subjective study on a 6DOF case- Better subjective perception of impact position
• Further work• Investigate more deeply the perceptive characteristics of vibrations• Conduct an objective evaluation on a virtual prototyping context
32
32
Thank you. Questions ?
?
33
33