Tactile Mary partx

fangscaryΤεχνίτη Νοημοσύνη και Ρομποτική

13 Νοε 2013 (πριν από 3 χρόνια και 6 μήνες)

41 εμφανίσεις

PIEZORESISTANCE

Piezoresistance is
a widely used technique in tactile sensing. Piezoresistance

refers to the process of
change in electrical resistance upon application of mechanical strain. Piezoresistive tactile sensors utilize
this principle whereby the change in resistivity due to an applied load assists in detecting the presence and
characteri
stics of an object. The resistance of a
resistor is given by the formula
l
R
A


, where
l
is the
length of the resistor,
A

is the cross
-
sectional area, and
ρ

is the bulk resistivity. Hence, there are two ways
by which resistance ca
n chan
ge, through a change in dimensions
, or a change in
resistivity
.
In the first
case, stretching a material causes its length to increase, and its cross section to decrease in a manner
governed by poisson’s ratio
, hence increasing its resistance. In the seco
nd case, resistivity changes as a
function of strain. The resistivity of a material depends on its charge carrier mobility, which is given by
*
e
m



, where
e

is the charge per unit carrier,
τ

is the average scattering time, and
m
*

is the effective
mass per unit carrier. Both the scattering time, and the effective mass are related to the atomic spacing in
the material. Thus, when deformation
causes a change in interatomic spacing, a change in resistivity is
observed.


Typical piezor
esistive materials include various semiconductors and metals such as silicon, germanium,
and aluminum.
A variety of piezoresistive tactile sensors have been reported in literature

consisting of
these materials. Micromachining was particularly attractive du
e to the ease of integration with signal
-
processing circuits on a single chip.

However, a drawback of these semiconducting and metallic materials
is the difficulty of packaging sensors made from these materials onto
curved surfaces such as robotic
hands. I
n addition, for the case of silicon processing, the tactile array would be limited by the finite size of
silicon wafers. As a result, many groups have explored polymer
-
based tactile sensors. For example, Park
et al. embedded carbon fibres in polydimethylsi
loxane
(PDMS)

to achieve a flexible piezoresistive sensor
(Fig
XXX
).
I

Conductive polymers are also sometimes used
II
, however, they can exh
i
bit large hysteresis.
Resistive sensors tend to perform better at low frequency measurements.


Figure
1
. Piezoresistive tactile sensing
array

with carbon fibres connected to electrodes, in a PDMS
casing

Mukai et al. developed a human
-
interactive robot “RI
-
MAN” (
Figure XXX
) with piezoresistive pressure
sensing elements by FUJIKURA (
figure XXX).
III

These elements are capable of detecting absolute



I
A piezoresistive tactile senso
r based on carbon fibers and polymer substrates

II

Conductive polymer
-
based sensors for biomedical applications

III

Development of the Tactile Sensor System of a Human
-
Interactive Robot “RI
-
MAN”


pressures between
42 and 437 kPa
, with virtually linear output from applied pressure. The circuit is
fabricated as a flexible printed circuit so that it can be fitted over a curved surface. The

tactile sensors
were robust enough to last over 10 months despite experiencing hard contact with dummies.


Figure
2
. RI
-
MAN carrying a dummy


Figure
3
. Structure of sensor element employed in RI
-
MAN, ado
pted from FUJIKURA’s piezoresistive tactile sensor element.




CAPACITANCE

Capacitance between two plates is given by
0
r
A
C
d


, where
0

and
r

are vacuum and dielectric relative
permittivity respectively,
A

is the area of a plate, and
d

is the interpolate spacing. When using capacitive
sensors for tactile sensing applications, a load on the capacitor leads to either a change in plate area, or
cha
nge in distance between the plates, causing a change in capacitance. This change in capacitance is
typically on the order of femtofarads


hard to detect without complex signal conditioning electronics.
Hence, capacitive tactile sensors are difficult to mi
niaturize due to crosstalk and stray capacitances.
However, an advantage of capacitive sensors is that they can be easily adapted to curved surfaces, and
they can be easily scaled up.

To maximize the change in capacitance with a certain applied load, a hi
gher permittivity dielectric, such
as polyvinylidene fluoride, can be used. Capacitive based sensors typically come in two configurations:
coaxial (Figur
e XXXX
) and parallel plate (Fi
gure XXXX)
. Parallel plate capacitors are easier to fabricate,
however, c
oaxial cylindrical capacitors can yield a larger response if the plate area to distance ratio is less
than 1.


Figure
4
.a) Cylindrical capacitive tactile sensor cross section
IV
; b) Three
-
plate capacitive sensor
V

By adjusting the dimensions of the capacitive sensor, one can tune the operating range of the sensor.
Paschen et al. fabricated parallel plate sensing elements
using polysilicon, and by adjustin the top
membrane diameter, he could effectively control the s
ensor to operate at different ranges, up to tens of
MPa, although linearity of response is compromised (
Figure XXXX
).
VI




IV

http://www.gmrv.es/~motaduy/Eurohaptics08/Goethals_paper1.pdf

V

A Robust, Low
-
Cost and Low
-
Noise Artificial Skin for Human
-
Friendly

Robots

VI

A novel tactile sensor system for heavy
-
load applications based on an integrated capacitive pressure sensor


Figure
5
. Signal vs. pressure curves for capacitive tactile sensors with membrane diameters of a) 120 um, b) 7
0 um, and c) 25
um.


















Tactile sensing in robots:

With the rapid advancements of robotics for various applications, more and more efforts have been
devoted to increasing robot autonomy. A semi to fully autonomous robot should have the ability

to extract
information about its surrounding environment, and move part or all of itself in the environment without
human interference to effectively manipulate its targets. Thus, the robot should be equipped with sensing
hardware and appropriate accompan
ying circuitry. In robotic sensing, tactile sensors are extremely useful
for activities such as navigation, manipulation, and construction. A good tactile sensor can provide
information about the object being touched, such as its texture, hardness, fricti
on, etc. This knowledge is
essential for enabling reliable handling of unknown objects in uncontrolled environments by the robot.
For example, a robot designed to pick up a certain object from a pile of different objects should be able to
distinguish wheth
er the object being touched is the one of interest, and know the appropriate amount of
force to apply to pick up the object based on the feedback of its tactile sensors.

Human tactile sensing has served as the main inspiration for the development of roboti
c tactile sensors.
Many different types of receptors are located in the human skin, which respond to a wide variety of
stimuli. These stimuli range from Hz to hundreds of Hz, and can sense skin stretching, curv
ature,
vibration, muscle force, and more
. This

suggests that an advanced humanoid robot requires tactile sensors
for various ranges and parameters. For example, tactile array sensors are appropriate for determining
pressure distribution and the local shape of the object. Force
-
torque sensors are appro
priate for
determining contact force and torque, suitable for placement near robotic finger
-
tips. Dynamic tactile
sensors are more appropriate for detecting vibration, slip, stress changes etc.

Various different sensor
structures and transducing technologi
es have been investigated, such as piezoelectric, piezoresistive,
capacitive, resistive percolation, and optical. They will be described in more detail in the next section.














Introduction into our application

Among the
numerous

useful applications of robotics
,
applications to missions dangerous to
human (eg.
fire
-
fighting
)

have received significant attention
.

Fire
-
fighters put their lives at risk
by entering
emblazed

buildings to find victims.

According to results published by the United
States Fire Administration, approximately 100 firefighters are killed on duty every year in the
US. Using semi
-
autonomous
robots instead of humans in dangerous rescue situations could
significantly decrease the

number of fire
-
fighter deaths.

There are many approaches to employing a robot for fire rescue. For example, a small, quickly
navigating robot could be used for finding trapped victims
in such a situation, or a large robot
could be used for carrying victi
ms out from the danger site

or carrying out fire distinguishing
measures
.
V
arious robots for fire
-
fighting have been reported in literature.
Penders et al.
designed a robot swarm to assist fire
-
fighters in searching a warehouse
VII
. The main role of the
swarm

is to
facilitate searching by fire
-
fighters in a smoke
-
filled, large warehouse. Thick smoke
from a fire can significantly impair human vision, poor tactile awareness can be caused by thick
safety clothing, and hearing can be impaired due to headgear and a
mbient noise. Thus
communication between the robot swarm and fire
-
fighters can improve navigation for fire
-
fighters through a crowded warehouse.

Ito et al.
proposed

a

fire
-
rescue
system

where
rescue staff
control the rescue robots from outside to quickly f
ind victims
VIII
. A snake
-
like robot prototype
was presented, capable of avoiding obstacles using wires at the sides of its body, which act as
tactile sensors.


The application detailed herein is using a large semi
-
autonomous robot with tactile sensors
capable of finding victims within a certain area, and carrying them to safety.
The semi
-
autonomous nature of the robot imparted by its tactile sensors enable them t
o effectively remove
obstructing objects and handle the victim.

Apart from tactile sensors, the robot can als
o contain a
camera
, proximity sensors,
temperature sensors
, and other equipment

to help its operation.

M
any
robotic systems previously reported in
literature require one
-
to
-
one operation by a well
-
trained
professional operator. However, in disasters such as a fire, it may be difficult to gather a group of
professional operators together in a short amount of time. Thus
,
by equipping robots with a
tact
ile sensory system, they can carry out much of the work autonomously, allowing
multiple
robots
to

be handled by a single human. In the case of a fire,
these robots can be sent into the
danger site. The operator can determine the approximate locations
of th
e victims visually
, then
direct a

robot towards that area, so that it can carry out the task of finding the victim, removing
weights or coverings on the victim, then carry them to safety by navigating out of the danger



VII

A Robot Swarm Assisting a Human Fire
-
Fighter

VIII

A rescue robot system for collecting information designed for ease of use


a proposal of a rescue

systems
concept

zone. The tactile sensors will help t
he robot

determine if the object they are touching is alive,
whether it is a human, its size and shape,

as well as

its orientation
.

Although little has been reported in literature for humanoid robots with tactile sensors
specifically for bringing victims out of a burning building, much has been reported regarding
robots with tactile sensors
that are
capable of lifting heavy weights, wh
ich are critical
to

this
application. Ohmura et al. described a robot capable of lifting a 30 kg object by full
-
body

contact

tactile feedback

(f
igure XXX)
IX
.

The robot contains 1864 tactile sensor elements on its surface,
and a total of 46 degrees of freedo
m.

Vidal
-
Verdu et al. showed the design of a large area tactile
array for a rescue robot. The robot is capable of lifting at least 120 kg. Force/torque sensors are
also incorporated to assist it in manipulating humans safely without hurting them.
X

As the d
esign
tactile sensory system on the robot is a critical aspect of the robot’s performance, the type of
tactile sensor and its materials must be carefully chosen. In the next section, different types of
tactile sensors will be outlined and compared to selec
t the most appropriate choice for this
particular application in fire
-
fighting. These sensing mechanisms fall largely into the categories
of piezoelectric, piezoresistive, op
tical, resistive percolation, and capacitive.


Figure
6
.

A humanoid robot capable of handling a 30 kg weight
IX




IX

Humanoid Robot which can Lift a 30kg Box by Whole Body Contact and Tactile Feedback

X

Large area smart tactile sensor for rescue robot