Proceedings of the 2000 IEEE
International Conference on Robotics & Automation
San Francisco, CA - April 2000
Human Tactile Spatial Sensitivity for Tactile Feedback *
G. Moy, U. Singh, E. Tan, R.S. Fearing
Department of EE&CS
University of California
Berkeley, CA 94720-1770
Abstract
In this paper, we quantify several spatial capabil-
ities of the hmnan tactile system needed for tactile
feedback, or teletaction. Psychophysics experiments
measure the amplitude resolution of the human tactile
system, the effects of shear stress on grating orienta-
tion discrimination, and the effects of viscoelasticity
{creep and relaxation) on tactile perception for static
touch. The results are used to determine teletaction
system design parameters. We find that 10% ampli-
tude resolution is sufficient for a teletaction system
with a 2 mm elastic layer and 2 mm tactor spacing.
1 Introduction
Information about texture, local compliance, and
local shape is important in applications such as
telesurgery or handling of fragile objects in teler-
obotics. Figure 1 shows a general configuration of a
tactile feedback (or teletaction) system. One possi-
ble application is on a robotic laparoscopic telesurgery
system. The tactile sensor is mounted on the end ef-
fector (the laparoscopic instrument), and the tactile
display is mounted on the master manipulator (the
user interface). The tactile display presents informa-
tion recorded by the tactile sensor to the user. Ideally,
the patterns felt by the user would be indistinguish-
able from direct contact with the environment. The
tactile display needs to generate surface stresses that
realistically represent data collected by the tactile sen-
sor. To fully control surface stress, the ideal tactile
display system would be an infinite density array of
3 DOF actuators.
Teletaction systems are composed of a tactile sen-
sor, a tactile filter, and a tactile display. We discuss
fingertip teletaction systems with high spatial detail
but low temporal resolution. Most teletaction work
has focused on tactile sensors and displays [Shimoga
1992; Howe et al 1995]. Tactile sensors are compar-
atively well understood [Howe 1994], but are gener-
ally designed for shape recognition and manipulation
tasks, not teletaction. Typical tactile sensors range in
size from a 1 mm square with 8 × 8 elements [Gray and
Fearing 1996] to a 16 mm square with 8 x 8 elements
*This paper is a synopsis of "Human Psychophysics for Tele-
taction System Design" which will appear in haptics-e, The
Electronic Journal of Haptics Research.
Environment _ ~ Measured Strain ~ Stress Pattern
-is ..... I =lF''r I -,, . u..
Figure 1: A block diagram representation of a general teletaction
system.
[Howe et al 1995] to a 25 mm diameter cylinder with
3 × 16 elements [Nicolson and Fearing 1995]. Tactile
sensors typically respond to the normal component of
strain though there are tactile sensors that measure
normal and shear stress [Domeniei and DeRossi 1992].
A tactile filter converts tactile sensor data to tac-
tile display data. Some concerns of tactile filter de-
sign include spatial and temporal sampling differences
between the sensor and display, anomaly and noise
filtering of the sensor data, and conversion of strain
profiles from the sensor to normal and shear displace-
ment profiles or normal and shear force profiles for the
display. The filters are typically computers with A/D
and D/A boards.
Tactile displays originated with tactile reading aids
for the blind using piezoelectric-driven pins and direct
pneumatic actuation [Bliss 1969]. Progress in tactile
displays has been slow, due to the demanding mechan-
ical requirements. An ideal display requires 50 N/cm 2
peak pressure, 4 mm stroke, and 50 Hz bandwidth;
that is, a power density of 10 W/cm 2 with an actua-
tor density of 1 per mm 2. The performance require-
ments are a result of the 70 SA I mechanoreceptors
per cm 2 [Johansson and Vallbo 1979] and force and
displacement for compression of the finger [Serina et
al 1997].
Tactile display designs have used solenoids [Fischer
et al 1995], shape memory alloy [Howe et al 1995; Has-
ser and Daniels 1996], pneumatics [Cohn et al 1992;
Caldwell et al 1999], and MEMS [Ghodssi et al 1996].
Voice coil actuators have also been used [Paw[uk et al
1998], but result in a large apparatus. Electrocuta-
neous stimulation [Kaczmarek et al 1991] is mechani-
cally quite simple; however, the perceptual effects are
hard to analyze. Typically, tactile displays control
either displacements or forces. In a displacement dis-
play, an array of pins is shaped into a contour. In
a force display, the pin array will produce a surface
stress distribution representing the data. The tactile
display's spatial density is limited by actuator size.
Currently, the spacing between the centers of the pins
is around 2 mm [Cohn et al 1992; Howe et al 1995].
0-7803-5886-4/00/$ 10.00© 2000 IEEE 776