instrument science and technology artificial tactile sensing and haptic … · 2006-02-28 · meas....

Meas. Sci. Technol. 2 (1591) 1003-1016. Printed in the UK

INSTRUMENT SCIENCE AND TECHNOLOGY

Artificial tactile sensing and haptic perception

Danllo De R o d Centro 'E Piaggio', School of Engineering, University of Pisa, Via Diotisalvi. 2. 56100 Pisa, Italy

Received 25 March 1991, accepted for publication 25 July 1991

Abclracl. A critical evaluation of t h e state of t h e art in tactile sensing is presented. Various fields of application such as prosthetic and teleoperation are covered, but particular attention is paid to advanced robotics, where most of t he recent developments have taken place.

In addition, tactile sensing modalities are discussed in some detail through reference to basic issues of contact mechanics to provide cluesfor the conception and design of artificial haptic systems capable of surface texture discrimination, stable object grasping, fine-form detection, hardness evaluation and thermal sensing.

1. fntroductlon

The high degree of dexterity that characterizes grasping and manipulative functions, and the sophisticated capa- bili(4 of recognizing object features in humans, are the result of a powerful sensory-motor integration which fully exploits the wealth of information provided by the cutaneous and kinesthetic neural afferent systems.

In neuroscience literature it is customary to refer to tactile sensation when dealing with cutaneous spatio- temporal discrimination Of mechanical stimuli. The term haptics is used primarily when referring to Sensations involved in stereognostic functions or to other perceptual operations performed by using, in addition to cutaneous sensing, afferent and efferent kinesthetic responses (Loomis and Lederman 1986).

Haptic perception implies integration of skin sensing with force, displacement and position sensing located in muscles, tendons and joints (kinesthesia) which lead to active touch operations when performed under the control of the sensory-motor system (Gordon 1987).

In this paper, the present status of research on tactile sensing (artificial touch) is outlined with reference to different fields of application.

In the world of engineering, the continuous scanning of variable contact forces that occurs during object expioration and manipulation by means of miniaturized Sensors and Sensor arrays is becoming increasingly important in a,jvanced robotics, teleoperation and telepresence, medical prosthesis and orthotics.

The human tactile system necessarily constitutes the ultimate model inspiring research activity in tactile sensing. However, the various applications present different constraints and demand different specifications and design options.

Research and development on tactile sensing has experienced an impressive growth in the last decade; system performance and limiting design specifications have been laid out (Harmon 1982), several kinds of materials and sensing techniques have been exploited (Pennywitt 1986) and methodologies and algorithms for tactile data analysis have been proposed (Nicholls and Lce 1989).

In this paper major emphasis is given to problems and solutions in advanced robotics, where tactile sensing is very actively investigated and most of the recent advances have originated.

In this context, relevant aspects of contact mechanics and grasp kinematics are discussed, transduction effects

1003

Danllo Dm Rossl was born in Genova. Italy. in 1949. H e received t h e Laurea degree in chemical engineering from the University of Genova. Italy. From 1976 to 1981 h e was a Research Fellow in the Institute of Clinical Physiology. Italian National Research Council, Pisa, where he is currently leader of the Sensor Group. Since 1982 h e has been with the School of Engineering, University of Pisa. Italy. where he is currently an Associate Professor of Biomedical Engineering. He has also worked in France, USA. Brazil end Japan.

His scientific interest is focused on the physics of organic and polymeric materials for signal transduclion and modelling, and on the design 01 sensors and actuators lor bioengineering and anthropomorphic robotics.

W57-0233/91/111003+14 $03.50 0 1991 IOP Publishing Lld

Danilo De Rossi

and materials are mentioned and the most significant developments in tactile sensing are illustrated.

2. Dinerent fields of application

2.1. Tactile sensors for prostheses and orthoses

The development of more efficient prosthetic and orthotic systems intended to substitute for or help in restoring lost mechanoreceptive and manipulative functions because of injury or disease now depends on major breakthroughs in artificial tactile sensing technology and biological interfaces.

Strict requirements of cosmetic and functional acceptability severely limit the usefulness of currently available devices. More substantial limits are set by the inability to transfer tactile information, in synthetic form, to the aRerent neural system and by the impractical use of sensory substitution channels. As a result most of today's hand prostheses are simple devices made of grippers with one or two degrees of freedom, working under visual control.

Localized force sensors, limiting-force switches and even low resolution tactile sensor arrays have been developed, but at present are little used.

Only if current efforts in the development of nerve guidance channels and neural connectors one day prove successful will the use of tactile sensors for hand prostheses be worth reconsideration.

Peripheral (limb) neuropathy caused by disease or trauma affects a relatively large fraction of the adult population. Of particular concern in developed countries are the secondary pathologies associated with the loss of peripheral sensation in chronic diabetic patients.

the available technology still being at a quite primitive level.

2.2. Haptics in teleoperators and telepresence systems

Teleoperation and telepresence enable humans to extend their manipulative ability or project their active presence to remote locations (Bejczy 1980). Teleoperators (remote manipulators) definitely need an expansion of their functional performance with respect to the ability to perceive haptic information and to convey it to a human operator in order to fully benefit his sensory-motor skill and decision-making capability.

Teleoperation and telepresence necessarily imply a strict interaction between human operator and machine.

Increasing demands of superior performance for currently available teleoperators, usually provided only with vision feedback, and for advanced computer interfaces (Foley 1987) motivate efforts to find new methods and techniques to convey kinesthetic information through man-machine interfaces. Sensorized gloves are now available (Data-Glove) to provide real-time information on operator hand orientation and the motion of finger joints to a computer. Force feedback control of teleoperators is increasingly used in underwater and space operations (Heer and Bejczy 1983).

Tactile sensing in these areas is thought to be important, However, the real-time presentation of remote tactile information to the operator for a prompt response still faces very fundamental difficulties. Major advances in this field are also hampered by our fragmentary under- standing of human haptic perception.

2.3 Robotic tactile sensing The to be sensed are the pressure distri- The development of future generations of robots capable

butions on the foot during standing and walking (Hen- ning et all9821 01 the finger and palm contact pressures in object grasping (Johnson and Huffman 1974). Practi-

of operating in an unstructured environment ~r intended to for man in hazardous or inaccessible locations demands the implementation of sophisti-

cal considerations, however, necessitate a rather small number of sensing sites.

Many different transduction techniques and materials with the essential requirements of reliability, robust-

matching with body tissues (Brand and Ehner 1969) have been used.

In the last few Years, tactile sensing has begun to be used in functional neuromuscular stimulation (FNS), where orthotic systems have been developed to electri- callv stimulate Daralvsed muscles through feedback from

cated sensory capabilities, far beyond those available in today's industrial robots.

Harmon (1982) followed up a survey on automated tactile sensing among developers and potential users,

several of application (see tables 1 and 2). The analysis of Harmon has also generated tentative specifications for tactile which are widely referenced by operators in the field,

ness, minimum encumbrance and good compliance providing a general view of needs and requirements in

These specifications are: . . - sensors located on the insensate hand or foot (Crago et a/ 1986).

Both single-element compliant pressure' sensors and arrays of tactile sensors based on capacitance effects have been developed, hut they do not yet satisfy the strict specifications dictated by this particular application in terms of mechanical compatibility, reliability and patient acceptance.

The search for more appropriate tactile sensors in this area of application, however, is actively pursued,

1004

The sensor surface or its covering should combine compliance with robustness and durability.

s The sensor should provide stable and repeatable output signals. Loading and unloading hysteresis should he minimal. Since some degree of viscoelasticity is always present in plastics and elastomers, the mechanical loss tangent should he independent of frequency in the range of use. Linearity is important, although only monotonic

response is absolutely necessary. Some degree of non-linearity can be corrected through signal processing.

e The sensitivity of each individual sensing unit should be of the order of 1 gram. A dynamic range of at least 1OOO:l is thought to be needed.

e The sensor transduction bandwidth should not he less then 100 Hz, intended as tactile image frame- frequency. Individual sensing units should accordingly possess a faster response, related to their number, when multiplexing is performed.

e Spatial resolution should be at least of the order of 1-2 mm, as a reasonable compromise between gross grasping and fine manipulation tasks. The area covered, and hence the number of individual sensing units, also depend upon the end-effector geometry and kinematics.

It should be clear, however, that the exact requirements of a particular sensor depend on its specific use.

The requirements listed above refer to skin-like sensing performed by bidimensional, miniaturized tactile sensors in an array, whose single sensitive sites are called ‘taxes’.

To perform object manipulation using dexterous end- effectors, the force and motion states of grasped object and finger links should be controlled, possibly using a minumum set of sensors capable of determining in real

Table 1. Major sensory needs for robotic tasks.


time the location, orientation and type of contact. This problem has been approached by using resultant force and moment miniature sensors located inside the fingertips of the end-effector to provide the robot with a sense that is more analogous to human kinesthetic senses. The integration of skin-like and kinesthetic-like sensing will definitely lead to new robotic systems equipped with artificial haptic perception.

Indeed the quality and quantity of information required to properly command a dexterous robotic end- effector very much depend on the tasks and goals it is intended to perform.

Sensorial needs of a haptic nature, allowing robotic multifingered hands to perform fine manipulative tasks and object exploration, have evolved to the complex set listed below:

(a) monitoring finger coordinates; (b) measuring point contact forces and torques

(resultants); (e) estimating contact area extension and position; ( d ) incipient slip detection; (e) fine-fonn discrimination; (f) probing material bulk properties (hardness, vis-

coelasticity); (g) evaluation of surface properties (texture,

roughness); (h) thermal sensing.

Function Simple Tactile touch sensing Vision Proximity

1. 2. 3. 4. 5. 6. 7. 8.

9. 10. 11. 12. 13.

14. 15. 16.

17. 18. 19. 20. 21. 22. 23. 24. 25.

Assembly Gauging (quantitative) Grinding and deburring Harvesting Inspection (qualitative) Medical examination Pick and place (manipulation) Prosthetics and orthotics Sensory Orthopedic Sorting Space (assembly, prospecting, repair) Surgery Teleoperators (biological. nuclear, space) Underwater (prospecting, repair, salvage)

Machining Stacking Welding

Casting Forging Mining Moulding Painting Polishing Pouring Stamping Transporting (gross movement)

X

X

X

X

X

X

X

X X

X

X

X

X

X

- - -

X

X

X

X

X

X

X

X - X

X

X

X

X

X

X

- - - - X

X - - X

X

X

X - X

x X

X - X

X

X X

X

X

X

X

X

X

X

X

X

X

X

X

X ~ ~~

NOTE: x =definitely useful; -=somewhat useful; .=useless. Source: Harmon (1982, Map 1). (Reprinted from Int. J. Robotics Res.1 (2) L D Harmon, Automated tactile sensing, by permission of the MIT Press, Cambridge, MA, 0 1982 Massachusetts Institute of Technology.)

1005

Danilo De Rossi

Table 2. Estimates of transducerlsystem properties required to provide functions (rank-ordered by increasing parametric demand, D) .

Spatial Time Force Dynamic Pattern resolution resolution sensitivity range recognition

Function Coarse Fine Low High Low High Small Large Simple Complex Dt

1. Space X X X X X 1 2. Underwater X X X X X 1 3. Grinding X X X X X 2 4. Harvesting X X X X X 3 5. Sorting X X X X X 3 6. Gauging X X X X X 3 7. Inspection X X X X X 3 8. Teleoperators X X X X X 4 9. Pick and place A x j: x X 4

10. Assembly X X X X X 5 11. Prosthetics X X X X X 5 12. Medical examination X X X X X 5 13. Surgery X X X X X 5

t S u m of t h e entries for t h e more demanding of the two classes in each category; hence, the s u m of t h e X entries in fine space resolution,high time resolution, high force sensitivity, large dynamic range and complex pattern recognition. Source: Harmon (1982, Map 4). (Reprinted from Int. J. Robotics Res.l(Z) L D Harmon, Automated tactile sensing, by permission of the MIT Press, Cambridge, MA. @ 1982 Massachusetts Institute of Technology.)

Sensing and processing all this information in an integrated system represents a formidable task, over- whelming in its complexity and far beyond our present knowledge and technical capabilities. However, all these sensing modalities have been investigated as isolated tasks and several devices conceived and implemented to show the feasibility of detecting single tactile sensing features.

A device has been shown (Dario et a/ 1984) to be capable of sensing multiple features, although it requires information extraction procedures that are still unman- ageable these days, at least where sensory-motor integration is concerned.

Haptic sensing by machines is an emerging field involving a wide variety of disciplines which range from materials science, mechanics, sensor technology, micro- electronics and signal analysis to the most sophisticated spheres of artificial intelligence.

3. Transduction principles and materials

The wide variety of configurations and design options adopted in tactile sensing technology originates from the exploitation of many different transduction effects and materials capable of mechano-electric, mechano- magnetic and mechano-optic conversion. Thermal effects also deserve consideration in this context because of their use in thermal sensing and their unavoidable interference in mechanical sensing.

In table 3 a synoptic illustration of intrinsic effects is given with reference to the field of tactile sensing.

In table 4 indirect (geometric) sensing effects used in tactile sensing are illustrated with reference to the most significant advances.

1006

Due to the very specific nature of tactile sensing, some of the traditional figures of merit in mechanical sensing (Neubert 1963) are rendered insufficient or inap- propriate. Possibly the most peculiar requirements of truly ‘skin-like’ sensing consist in the mechanical compliance of the transducer itself, for increased grasp stability, its conformability to curved shapes and the possibility of realizing high-density miniature sensing arrays.

Not surprisingly, plastics, rubbers and carbon fibers have found widespread use in tactile sensor technology because of their high mechanical compliance and robustness, combined with their ability to be processed into complex shapes. Some polymers and plastic composites also possess intrinsic transduction properties (piezoresis- tivity, piezoelectricity, photoelasticity, magnetoelasticity, etc) which render these materials particularly attractive for tactile sensing.

Recent reviews on tactile hardware (Nicholls and Lee 1989, Jayawant 1989) are available for the interested reader and tactile sensor materials and design issues are also discussed extensively in a book edited by Webster (1988), hence no further details on these aspects are reported here.

Moreover, a broad survey of the technology support- ing general purpose manipulation has been published by Crupen et a/ (1989), providing useful information on dexterous end-effectors and tactile sensing.

Much less attention has been paid to the analysis of grasp kinematics and contact mechanics and their very fundamental role in tactile sensing and manipulation. These aspects are now discussed with reference to the implications and constraints they impose on the design and implementation of haptic hardware and software.


Table 3. Intrinsic transduction effects.

Ferro- electric

Piezo Strain (stress)- Ferroelectric High dynamic Lack of DC Dario et al (1984) polarization polymers (PVDF, range response

P(VDF-TrFE)) Good De Rossi et a/ (1987) mechanical properties

Pyro Temperature- Ferroelectric Stress Multiplexing is Petlerson and polarization polymers (PVDF, components difficult Nevill (1986)

P(VDF-TrFE)) selectively sensed

Broad Thermal and frequency mechanical response effects are

difficult to separate .

4. Basic mechanlcs and tactile sensing

Considerations based on kinematics and dynamics o f

clarifying design issues and dimensioning criteria in haptic sensing.

rigid bodies and contact mechanics are absolutely hnda- mental t o the analysis and design of artificial haptic systems. Although these disciplines are well established, their application t o machine grasping and tactile sensing has been limited, and only in the last few years have reports of specific studies on the subject appeared in the literature.

The considerations reported here are l imited in scope, addressing specific tactile sensing modalities in some-

4.1. Grasp kinematics and kinesthetic-like sensing

M u c h of the early work on grasping has looked at contact between objects and end-effectors on the basis of purely kinematic arguments (Salisbury and Craig 1982, Ker r and Roth 1986). The primary goal of these studies was t o define criteria for selecting internal grasp forces (those to be applied by the end-effector actuators) to ensure stable grasp, thus avoiding object sliding or

what idealizkd conditions and purposely directed at damage.

1007

Danilo De Rossi

Tabla 4. Geometrical changes induced by contact forces.

Geometric Coupled Working change in: parameters principles Pros Cons Typical desians .. - Capacitance Force-voltage Compressive Large dynamic Chun and Wise

thickness of a High spatial dielectric, resolution increasing Good frequency Noise capacitance response susceptibility

(or frequency) force reduces range (19W

Compatible Fearing et a l with VLSI (1 986) technology

Resistive contact Force-voltage Contact Simple Some hysteresis Hillis (1982) drop resistance of construction

conducting Compatible with Only normal Raibert (1984) elastomers VLSl technology force detected changes upon High spatial squeezing resolution

Inductance Displacement- A moving core, Very large Poor reliability Sat0 et a/ (1986) induced energized by a dynamic range Bulky voltage drive coil, Low spatial

induces resolution voltage in a sense coil

induced range susceptibility (1983) displacements Robustness change High frequency Scarcely Kinoshita et a/ magnetic flux response investigated (1983) detected by Stress magnetoresistive component Vranish (1986) o r Hall effect selectivity transducers possible

Magnetic flux Force-voltage Contact-force- Large dynamic Noise Hackwood et a/

Ultrasonic time of Force-us pulse Force causes a - Non linear Shoemberg et a/ flight delay rubber layer to response (1 983)

deform and us Complex circuitry propagation High hysteresis time to vary accordingly

Optical Force-light Through Reliability Bulky Rebman and Morris

of light-pass Immunity from Low spatial Maalej and Webster transmittance intensity displacement (1984)

occluders EM1 resolution (1988)

Optical Force-light Deflection of Good spatial Narrow dynamic Bejczy (1981)

elements Immunity from Diflicult Schneiter and changes EM1 calibration Sheridan (1984) intensity of reflected light

reflectance intensity reflecting resolution range

Optical coupling Force-light Totally internally High spatial Difficult Tanie et a/ (1985) intensity reflected light resolution calibration

is modulated Good frequency Only normal Begej (1988) by compliant, response force detected grooved pad Immunity from

EM1

Contact models assumed in such analyses are ideal- ized point, line, planar and ‘soft-finger’ contacts. Salis- bury (1984) proposed a method o f obtaining the set of contact information that i s most relevant to those models. Salisbury’s original method consists in contact sensing based on the measurement o f resultant force and

1008’

torque remote from the contact area, and applies exactly t o only point-type contacts.

The theory o f ‘intrinsic’ contact sensing has been largely improved upon recently, and exact results for the general case of compliant fingertip-object pairs (soft- finger type contact) are now available (Bicchi et al 1990).


The salient features of intrinsic contact sensory systems are that the contact force is sensed in both its normal and tangential (friction) components, as well as the friction torque resisting rotational slippage. More- over. concise information on the location of contact is given, in terms of a characteristic point (the ‘contact centroid’) which lies inside the smallest convex portion of the fingertip surface enclosing all of the contact area.

As a consequence, however, when extended line or planar contacts occur, the line or surface of contact cannot he uniquely located (Salisbury 1984) unless ‘active sensing’ is performed. This procedure implies successive sensor readings during small exploratory motions between the tip and the object, whilst mantaining contact.

Typical implementations of such sensors consist of a multiaxis, miniaturized forceitorque sensor, placed inside and coupled with the fingertip cover that contacts the environment (figure I).

Brock and Chiu (1985) have designed and constructed a tactile sensor based on these premises using 16 semiconductor strain-gauges mounted onto the struc- tural support of the terminal phalanx of an artificial finger. The strain-gauges are bonded onto the four legs of a steel Maltese cross and the contact wrench system is calculated by inverting a redundant linear equation system in which strain readings, geometrical factors and material elastic coefficients are known entities.

Bicchi and Dario (1988) have described dimensioning criteria and implementation of a ‘minimal’ (minimum number of required sensing elements) tactile sensor in which six strain-gauges have been mounted onto a thin- walled cylindrical cantilever beam which has been incorporated into the tip of a multijoint robotic finger.

Bicchi (1990) has developed a systematic tool for the optimal design of force/torque sensors, based on the analysis of measurement error propagation.

4.2. Modelling friction and incipient slippage detection

Kinematic design is a powerful methodology in grasp modelling as discussed in the previous section.

However, neglecting dynamic aspects of the contact

U Figurel. Schematic drawing of an intrinsic tactile sensor mounted on a linger and exploring an unknown object surface.

and adopting oversimplified assumptions on the defor- mation and frictional behaviour of finger-tip materials necessarily leads to unrealistic predictions of the internal forces in end-effectors and poor grasp preformance.

More recent research work has therefore addressed the problem of analysing sliding phenomena when more realistic frictional models of interaction between compliant finger-tips and objects and dynamic factors of contact are considered.

Only after taking these aspects into account may more stable grasp modalities be conceived and may controlled sliding (Fearing 1986) between fingers and objects, necessary in dexterous manipulation, he better quantified.

The requirements of relatively large contact areas and deformable gripping surfaces, made of ruhher-like materials for stable grasp, force the analysis beyond the single-point and line contact modalities and coulombic friction.

The price to be paid is a rather complex analysis and additional difficulties related to the determination of contact conditions that are no longer uniquely determined.

Cutlosky and Wright (1986), having discussed non- coulombic frictional models for compliant, rubber-like materials in contact with rigid bodies, have performed a simplified dynamic analysis of contact modalities; obtaining limiting conditions which prevent rolling and slipping under linear shear or torsional stresses.

Howe et al (1988) examined the behaviour of soft and very soft finger contacts under combined torsion and linear shear loading and identified slipping limits using a model in which friction is assumed to he proportional to normal force to the power 213. They reported a simple, operative sliding limit as:

wheref; andf, are respectively the tangential and normal contact forces, m, is the spinning moment, p is the coefficient of friction and A is a proportionality constant between the torsion and shear limits, which is a function of the contact radius, r. Different modelling approaches permit the estimation of A , which is strongly dependent on contact geometry.

A more rigorous analysis of incipient sliding under tangential loading (Johnson 1987a) takes into account the irreversible nature of these phenomena and the onset and dynamic changes of ‘stick’ and ‘slip’ contact regions. When, in a non-conformal contact with friction, tangential loading is superimposed on normal contact forces, tangential surface tractions arise, causing ‘microslip’ to occur at the edges of the contact area where tangential traction is high and normal force is low. A ‘stick’ region develops in the central zone where tangential traction is lower and normal force higher. A further increase of tangential forces under constant normal loading will cause the ‘slip’ region to propagate inward and the ‘stick’ region to shrink until surface tangential tractions can no longer be sustained and the object starts to slip (see figure 2).

1009 . . .

Danilo De Rossi

Figure 2. Surface tractions and 'stick and 'slip' regions for two coaxial cylinders in contact (tangential forces superlmposed on a constant normal load). Curve A refers to the case of no slip; curve B shows the case of partial slip ('slip' and 'stick' regions develop). (Moaitied atter Johnson (1987a); figure 7.7. Reprinted by permission of Cambridge University Press.)

From the discussion above it follows that incipient slippage detection, for ensuring stable grasp by servo gripping forces, cannot be performed by kinesthetic-like sensing alone.

Distributed skin-like tactile sensors, capable of resolving normal and shear contact forces and permitting the estimation of size and location of contact area, are necessary for ensuring stable grasp and controlled sliding in manipulation.

Several devices have been described, capable of detecting slip between a grasped object and the end- effector, Tomovic and Stojiljkovic (1975) reported the construction and use of a slippage sensor made of miniature spheres incorporated into compliant protec- tive laycrs, whosc rotary motions, induced by a sliding contact, can be detected electronically. Other authors used different transduction effects to reveal sliding contacts, such as acoustic emission (Dornfeld and Handy 1987), vibrations detected by photoelastic elements (Crowder 1988) and skin accelerations (Howe and Cutlo- sky 1989). All these sensors, however, detect a signal that is generated only when relative motion has already occurred, considerably reducing their usefulness in servo- loop grasp control.

Detection of incipient slippage, which is essentially what is needed of stable grasp and dextreous manipulation, is essentially an unexplored area. However, a few sensor arrays capable of selectively measuring stress components, and hence potentially of use in monitoring not only normal load but also tangential shear and torque in contact, have been described.

Hackwood et al(1983) proposed the use of an array of magnetic dipoles, embedded into an elastic pad, whose position and orientation can be detected by magnetoresistive elements. Sensing normal and tangential contact forces as well as torque was implied to be possible.

A stress-component-selective tactile sensor array, made of a suitable combination of piezoelectric polymer elements, has been described (De Rossi et a/ 1987) and its ability to resolve normal and shear components (Domenici et a/ 1989) demonstrated experimentally. Although the primary motivation for its development

1010

was fine-form discrimination for object shape recognition (see section 4.3), the sensor is now under test to assess its performance in detecting surface tractions and 'stick- slip' region dynamics before gross relative motion occurs.

Work along similar lines has been reported by Novak (1989), in which a preliminary analysis and design of a capacitive sensor array, potentially capable of tensorial strain detection, is given.

4.3. Contact mechanics and fine-form diserimination

Tactile spatial discrimination in humans reaches its peak in performance at the fingertips, where the high density U1 qJ~'C'd.""C" rlr=LrranuLrc=p,ruL> L l l l V W J rrrg,, LrJul"llu,,

sensing and sophisticated coding of neural patterns (Johnson 1983).

The motivation for the development of high-density tactile sensor arrays resides in the intent to replicate cutaneous sensing features dedicated to resolving and categorizing fine spatial details of an object's profile, i.e. its fine form.

Fearing and Hollerbach (1985) have introduced and analysed problems and suggested methodologies to detect the actual contact stress generated by object contact at the surface of a compliant pad covering the finger. Determining contact force was shown to be more significant than indentation profiles in planning grasping forces. The contact stress at the finger surface was calculated by using the measured, discrete spatial distribution of stress or strain in a plane located beneath or embedded in the compliant elastic pad.

The problem, so formulated, belongs to the class of inverse problems of elastic contact, where the force distribution acting on the boundary of the sensor should be inferred from spatially discrete knowledge of the stress field over a surface inside the sensor. This problem, however, is an ill-posed one in which the existence, uniqueness and stability of the solution are not guaran- teed, and only through suitable constraints, formulated on the basis of physical considerations or complementary information, may a solution eventually be found (Tikhonov and Arsenin 1977).

A few research groups have recently addressed this problem by conceiving and developing both suitable sensor arrays and inversion methodologies and algorithms, The problem is usually modelled as a frictionless, non-conformal contact occurring between a rigid indenter and a semi-infinite elastic, homogeneous, isotropic medium (see figure 3).

Limited classes of indenter shapes are used paradig- matically and direct problems of elastic contact are usually formulated as mixed boundary value problems and solved to obtain the strain in the medium in terms of surface stresses, indenter profile and resultant load.

Linear functional equations are obtained in the form.

-r __I- :Fl:--A -..-I- "------- A _ _ - - I t L:-L ----#..A: - ..

T x = y

where x and y are real functions and T is a bounded integral operator with smooth kernel.

Since the tactile sensor array can provide only a


solution by defining

x i a , B j = a , T+a,H

. I = SENSOR

Figure 3. Schematic drawing of an axially symmetric indenter loading t h e surface of a tactile sensor. On the right side t he shapes (flat-base cylinder, conical, parabolic) of the indenters used by De Rossi et a/ (1990) are depicted.

spatial sampling of stress (or strain) on a plane, and the surface stresses or indenter profiles are to be obtained as discrete samples, the inversion method consists in solving linear equations associated with discrete convol- ution as a generalized inverse solution:

xo=T-' y

where xo is a vector of dimension equal t o the number of surface samples and y is a vector whose dimension is equal to the number of sensors. T-' is the inverse operator.

Pati ef al (1988) have described an approach to tactile inversion based on neural network principles. A triaxial piezoresistive silicon array and a dedicated neural processor on a VLSI chip were implemented. Recon- struction of surface stresses generated by a cylindrical indenter, accomplished by a neural network employing a maximum entropy deconvolution, was fast and accurate.

De Rossi et al (1990) reported faithful reconstruction of axially symmetric indenter profiles by numerical algorithms based on regularized solutions of inverse elastic contact problems, through discrete inverse operators. A stress-component-selective sensor array made of piezoelectric polymer elements and hybrid circuitry for ampli- fication and multiplexing is used in this study.

Solutions of the inverse problem were based on the definition of a class of approximated inverse operators:

R,=(T*T+xia iBt Er)-' T*

where the asterisk denotes the adjunct operator, a, are real positive numbers and Bi are constraint operators through which a priori knowledge of the solution can be used to stabilize the inversion process. R, is an approximation of T-' in the sense that

lime-o I R. T x - x I = O V X € X

where X is the domain of T . The crucial issue with this method is the choice of the operator E;; one possible choice is to impose smoothness constraints on the

where I is the identity operator and H the first derivative operator. Making this choice means to favour, among all the functions that give a good enough fit to the data, the one with minimum norm and first derivative. a , and a, play the role of trade-off weights that enforce or relax the a priori constraints.

Simulation performed using input data from the theoretically solved direct problem showed accurate reconstruction of indenter profiles (conical, parabolic, flat-base circular) even in the presence of Gaussian noise added to the data (see figure 4).

Further work along these lines is expected to address the relaxation of several simplifying assumptions of the contact model (contact without friction, axially symmetric shape, semi-infinite elastic medium, small defor- mations etc), the distinction between errors in the model and noise in the data, and improvements in hardware and algorithms.

4.4. Probing and grasping of textured surfaces

Humans make extensive use of surface textural information for object classification, feature extraction and for adapting gripping forces.

Waviness and roughness sensing rely upon active tangential exploration in which mechanoreceptors and their associated neural processing extract averaged topographical features of surface asperities. The characteristic spatial wavelength of the surface features is much less than the spatial acuity of the tactile sensing apparatus. This form of sensing may be interpreted in terms of cutaneous sensing and peripheral processing, being wholly determined and coded by the relative discharge rates in different populations of skin mechanoreceptors (rapidly adapting, slowly adapting and Pacinian) (John- son 1983).

Some psycho-physiological studies of texture detection, aimed at identifying exploratory procedures adopted by humans performing specific recognition tasks, have also been motivated by the need to implement artificial haptic systems (Klatzy et al 1987) in robots operating in an unstructured environment.

Machine sensing of regularly textured and rough surfaces, however, has been seldom studied though often mentioned.

A few reports of a descriptive nature have addressed the capacity of different tactile sensors to probe surface texture.

Bajcsy (1985), in an articulated approach addressing the problem of reconstructing shape from touch, reported modalities and results on texture detection of different surfaces and materials using tactile sensor arrays with low spatial resolution moved tangentially across the object surfaces. Coarse texture was well discriminated through the temporal pattern analysis of sensor element responses.

Dario et al (1988) described a single-element piezo-

1011

-6 -4 -2 0 2 1 6 r lmml

Shear stress

2 0

-10 10 ' r ~ l -20

-30

- 4 0

- 5 0 - 6 - 4 -2 0 2 1 6

r 1m1

Figure 4. Shape reconstruction for a parabolic indenter. The two upper pictures represent the calculated normal and shear stress components at the sensor depth. The two lower pictures represent the shape of indenter reconstructed from normal or shear data. Full curve, actual shape; dotted curve, reconstructed shape (no noise); broken curve, reconstructed shape (5% Gaussian noise added).

electric sensor whose compliant, dome-shaped tip acts as a sort of distributed pick-up capable of converting texture spatial wavelengths into temporal signals under sensor-object sliding motion. A small set of grooved and embossed metal surfaces were probed and discriminated.

Work on these lines, however, has not provided any clue to the parameters and features to be extracted by roughness detection. Moreover, no report on the information analysis and coding needed for their actual use in object manipulation by machines appears to be available.

In contrast, careful studies of a general nature on the analysis and classification of surface texture have been performed and their effects on the results of classical contact theory (smooth surfaces in continuous contact) carefully evaluated (Johnson 1987h).

Waviness and roughness (see figure 5 ) detection is, to be sure, useful in object discrimination, but its role is far more relevant in planning and performing gripping and controlled sliding in manipulation.

In the case of objects with wavy surfaces, it is crucial to determine the evolution of real contact area under varying contact forces. Changes in the ratio of real to apparent (nominal) contact area with normal contact loads have been analysed theoretically and experimentally (Johnson 1987b).

For an isotropic, wavy surface in contact with an elastic half-space, experimental data and numerical solu-

1012

Weviness (secondary texture)

Roughness (primary texture)

W Figure 5. Major surface texture components.

tions of contact equations have been reported (Johnson et al 1985), permitting the evaluation of the load-varying contact area.

It is worth noting that the classical Hertz theory of contact predicts, in this case, that the real area of contact grows with the normal load to the power 2/3. It should not he surprising that the same law holds approximately for rubber in light contact with rigid bodies. Hertz theory, however, only provides asymptotic results at very low surface traction; at higher contact pressures


the area of contact grows faster than predicted by Hertz theory.

Surface texture can be analysed using this methodology when the amplitude of the surface undulations is small with respect to their characteristic spatial wavelength.

In the case of grasping objects with rough surfaces, assessing topographical features which condition contact behaviour would again be a valuable sensorial modality for an artificial manipulation system.

Roughness detection for explorative and discriminat- ive purposes (for example in object selection) should be based upon sound topographical descriptions of rough surfaces and realistic models of transduction phenomena.

Average roughness along a line segment L is defined as

where z is the local asperity height referred to the surface centre-line.

Useful parameters are the standard deviation us of the height of the surface from the centre-line, the root- mean-square slope U,,, and the root-mean-square curvature uk of the surface profile (or the mean summit curvature &).

When rough surfaces are pushed into contact, their behaviour is governed by these (or derived) quantities and by the asperity density qs. The analysis of profil- ometer traces and the derivation of the quantities above have been investigated by Greenwood (1984), and the reader should refer to this paper for a detailed account.

It is quite unlikely that data for an accurate analysis of surface roughness could be provided by even sophisticated and specialized tactile sensors. However, more pragmatic approaches should be adopted by identifying relevant features in spatio-temporal patterns of the signal generated by surface tangential exploration.

It is useful to note here that, as opposed to the few artificial roughness detection systems that have been proposed in advanced robotics, humans do not use spatial-to-temporal frequency coding in perceiving roughness (Taylor and Lederman 1975).

Much more relevant to artificial manipulation is the analysis of the effects of roughness on conformal and non-conformal contact. In the case of conformal (planar) contact with a randomly rough surface the effects are well known (Greenwood and Williamson 1966). Amon- ton’s law of friction states that the real area of contact grows in direct proportion to the normal load, and a coefficient of limiting friction can be used in defining criteria for slip avoidance.

A more complex situation occurs in elastic contact of rough non-conformal surfaces. Greenwood and Tripp (1967) extended the Hertz theory of elastic contact to rough surfaces and found that Hertzian results are valid at sufficiently high loads, but at lower loads the effective pressure distribution is much lower and extends much farther than for smooth surfaces.

In the case of a smooth sphere of radius R in contact with a plane, whose rough surface follows a normal Gaussian distribution of asperity heights about a mean surface, the pressure distribution and effective radius of contact have been shown to depend on two independent dimensionless parameters

where a. is the contact radius calculated for flat surfaces (Hertz radius).

Deviations from Hertzian behaviour caused by surface roughness can be large (see figure6), depending primarily on a and less markedly on p.

Although these two factors may influence planning and control of grasp forces considerably, their inference from measurement and quantification of surface roughness appears to be problematic in tactile sensing and strongly affected by modelling complexity and inac- curacy.

A more reasonable approach to account for surface roughness in object manipulation would possibly reside in monitoring the dynamics of ‘stick and ‘slip’ regions, as discussed in section 4.2.

4.5. Rheology and assessment of bulk material properties

Assessing bulk material properties of objects, such as hardness, bas attracted some interest because of its relevance to the tasks of classification, selection and manipulation.

The level of ongoing activity does not appear to be significant, however, and the few reports addressing these aspects are very preliminary and often only of a qualitative nature.

Even scarcer is the research and development work performed by the robotic community on machine psycho-rheology. Psycho-rheology, a discipline whose methodologies and techniques are aimed at interpreting haptic sensations associated with properties such as ‘consistency’, ‘body’, ‘tack’, etc and at replacing subjective

\‘ 0.4 4 -7

. 0 1 ,I 1 Roughness parameler a

Flgureb. Elastic rough sphere in contact with a fiat, smooth plane. Influence of roughness on the maximum actual pressure p(0 ) compared with the maximum Hertz pressure po. (Modified after Johnson (1987b); figure 13.12. Reprinted by permission of Cambridge University Press.)

1013 . - -

Danilo De Rossi

human judgements with instrumental measurements, has experienced major progress mostly with the pioneering work of Scott-Blair (1969).

Hardness evaluation was carried out by Bajcsy (1985) by pressing a robotic finger, sensorized with a low spatial resolution tactile sensor, against the object: this was performed in small incremental displacement steps, and by reading the sensor output during the loading and subsequent unloading processes.

Material hardness was ranked according to the slopes of the linear parts of the loading and unloading sensor outputs. No details were given about loading and unloading speeds.

Although discrimination of different materials was reported to be successful, it is unlikely that the procedure and devices that have been used will lead to true measurements, even in terms of ordinal scales.

Work along similar lines was reported by Bardelli e t al (1983) using a single-element sensor made of a piezoelectric polymer pressed against flat sheets of rubbery materials of different compliance and backed by a reference load cell. Hardness ranking was associated with the slope of the straight line obtained in the sensor output-reference cell signal plane under loading.

More significant work has been reported by Kato e t al (1977), as an initial step towards realizing artificially the recognizing function of softness similar to that in human hands. They prepared samples of polymeric materials with different viscoelastic behaviour, and psychophysical test responses were associated with the output of an automatic indenting apparatus purposely designed and constructed to classify and quantify viscoelasticity.

A good correspondence between ordinal and ratio scales (Stevens 1951) constructed in the psychophysical study and the output of the apparatus led the authors to formulate claims of the machine’s perception of softness.

De Rossi et al (1988), after investigation of intrinsic mechanoelectrical transduction properties of skin tissue under compression, proposed the use of charged polymer hydrogels as materials useful in tactile sensing, in particular for softness perception, because of their ideal compliance matching with human skin.

4.6. Heat transfer and thermal sensing

Human thermal sensing is an additional exteroceptive function for material discrimination, as it is not only used for temperature sensing. Differences in thermal conductivity and diffusivity in objects at the same temperature are perceived by thermal receptors since they govern the heat flux through non-isothermal skin.

Thermal sensing in robotics, as suggested by Harmon (1982), has been investigated by using single-element (Bardelli et al 1983) and array sensors (Dario and De Rossi 1985, Russel 1984). Pyroelectric and thermistor arrays, backed by a thermostated heat source, have been used to measure dynamic maps of temperature changes generated by contact between sensor and object. Dis-

1014

crimination among different materials was obtained with fairly good accuracy.

Thermal sensing through contact heating was suggested to be useful also in object shape recognition (Russel 1985).

In addition, shape recognition by a pyroelectric tactile sensor array was demonstrated through object shadowing under radiant heating (Wang e t al 1984).

5. Conclusions

The last decade has seen considerable effort applied tn research and development activities related to the design of tactile sensors, particularly for robotics applications. Much less attention has been paid to other important aspects having a crucial role in the implementation of artificial haptic systems,

Processing tactile data, sensory data fusion and sensory motor integration have not been specificially treated in this survey. Research on these topics is rapidly expanding, but the current status of our knowledge is still quite primitive.

Most of the work related to processing methodologies and algorithms of tactile data has been focused on the analysis of static tactile images, following the lines developed in the field of machine vision (see Nicholls and Lee 1989). This approach is limited in scope and does not consider a major asset of tactile sensing which lies in the processes of active touch and manipulation.

Planning active touch procedures and analysing the pertinent sensor data was recognized early on as being important (Gurfinkel e t al 1974), but progress in this area has been quite slow.

Integrating tactile and kinesthetic sensing with proximity information and visual images is a very challenging task and dedicated effort to tackle these aspects is beginning to be consistent (see J . Int. Robo t i c s Res. 1988).

Sensory-motor integration in robotic systems provided with haptic sensibility is almost unexplored as it constitutes the higher hierarchical level of complexity in analysis and implementation.

As a final comment, it should be pointed out that although the rapid growth in interest in the field of tactile sensing has initiated significant progress in haptic hardware, very little fall-out in real applications has occurred, and the present market for these devices is still very marginal, despite some early optimistic fore- casts. Future widespread use of tactile and haptic systems is still foreseen, but the time scale for these events to occur should be realistically correlated with the great theoretical and technical difficulties associated with this field, and with the economic factors that ultimately drive the pace of its development.

References

Bajcsy R 1985 Shape from touch Advances in Autonlalion and Robotics ed G N Saridis (London: JAI)


finger tip for a dextrous hand. SPIE Conf on Intell. Robotics and Como. Vision (Cambridae. M A ) oo 1-10

Bardelli R, Dario P, De Rossi D and Pinotti P C 1983 Piezoelectric polymers: new sensor materials for robotic applications Proc. 13th Int. Symp. Industrial Robotics 18/45-56

Bejczy A K 1980 Sensors, control and man-machine interface for advanced te!eoperaticn Sciencc 208 1327-35. - 1981 Touch sensor responds to contact pressure, NASA

Tech. Briefs Summer 207-8 Begej S 1988 Planar and finger shaped optical tactile sensors

for robotic applications IEEE J . Robotics and Automation 4 472-84

force sensors M I T A.I. Memo 1263

artificial hands Robotic Research ed R C Bolles and B Roth (Cambridge, MA: MIT)

from force measurements M I T A . 1. Memo 1262

for denervated hands and feet J . Bone Joint Surg. 51 109-16

Brock D and Chiu S 1985 Environment perception of an articulated robot hand using contact sensors Proc. A S M E Winter Annual Meeting (Miami, FL) pp 1-8

Chun K and Wise K D 1985 A high-performance silicon tactile imager based on a capacitive cell IEEE Trans. Electron. Deuices 32 1196-201

Crago P E , Chizeck H J, Neuman M R and Hambrecht F T 1986 Sensors for use with functional neuromuscular stimulation IEEE Trans. Biomed. Eng. 33 256-68

Crowder R M 1988 Sensors, touch, force and torque measurement International Encyclopedia of Robotics ed R C Dorf (New York Wiley)

Cutlosky M R and Wright P K 1986 Friction, stability and the design of robotic fingers Int. J . Robotics Res. 5 (4) 20-37

Dario P and De Rossi D 1985 Tactile sensors and the gripping challenge IEEE Spectrum 22 ( 8 ) 46-52

Dario P, De Rossi D, Domenici C and Francesconi R 1984 Ferroelectric polymer tactile sensors with anthropomorphic features Proc. Int. Conf Robotics (Washington, D C ) (IEEE Comp. Soc.) pp 332-40

Dario P, De Rossi D, Giannotti C, Vivaldi F and Pinotti P C 1988 Ferroelectric polymer tactile sensors for prostheses Medical Applications of Piezoelectric Polymers ed P M Galletti et a1 (New York: Gordon and Breach) pp 151-66

Data-Glove, VPL Research Inc, Redwood City, CA, USA De Rossi D, Canepa G, Bacci A and Caiti A 1990 Inversion

of tactile data through a skin-like sensor sensitive to stress components Proc. Annual Conf IEEE Eng. Med. Biol. 12 1946-1948

sensing skin mimicking electromechanical transduction properties of human dermis IEEE Trans. Biomed. Eng. 35 83-92

De Rossi D, Nannini A and Domenici C 1987 Biomimetic tactile sensor with stress-component discrimination capability J . Mol. Electron. 3 173-81

Domenici C, De Rossi D, Bacci A and Bennati S 1989 Shear stress detection in an elastic layer by a piezoelectric polymer tactile sensor lEEE Trans. Electr. Insul. 24 1077-81

Dornfeld D and Handy C 1987 Slip detection using acoustic emission signal analysis Proc. IEEE Int. Conf Robotics and Automation pp 1868-75

Fearing R 1986 Implementing a force strategy for object reorientation Proc IEEE Int. Conf Robotics and Automation pp 96-102

Fearing S and Hollerbach J M 1985 Basic solid mechanics for tactile sensing lnt. J . Robotics Res. 4 (3) 40-55

Fearing R S, Rise A and Binford T 0 1986 A tactile sensing

Bicchi A 1990 A criterion for the optimal design of multi-anis

Bicchi A and Dario P 1988 Intrinsic tactile sensing for

Bicchi A, Salisbury K J and Brock D L 1990 Contact sensing

Brand P W and Ebner J D I969 Pressure sensitive devices

De Rossi D, Domenici C and Nannini A 1988 Artificial

1 . ... Foley J D 1987 Interfices for advanced computing Sci. Am.

257 (41 83-90 Gordon G (ed) 1987 Active Touch the Mechanism of

Reccgni!io.e $1 Objects h.' , .'. Wxipule t ion (Oxford: Pergamon)

Greenwood J A 1984 A unified theory of surface roughness Proc. R . Soc. A 393 133-57

Greenwood J A and Tripp J M 1967 The elastic contact of rough spheres Trans. ASME - J . Appl. Mech. 34 153-9

Greenwood J A and Williamson J B P I966 Contact of nominally flat surfaces Proc. R. Soc. A 295 300-19

Grupen R A, Henderson T C and McCammon I D 1989 A survev of eeneral ouroose manioulation lnt. J . Robotics . . Res. 8 ( I ) 78-62

Gurfinkel V S, Shneider A, Kanavev Ye M and Curfinkel Ye V 1974 Tactile sensitizingbf manipulators Eng. Cybernetics 12 (6) 47-56

1983 A torque sensitive tactile array for robotics 1111. J . Robotics Res. 2 46-50

Harmon L D 1982 Automated tactile sensing Int. J . Robotics Res. I (2) 3-31

Heer E and Bejczy A K 1983 Control of robot manipulators for handling and assembling in space Mech. Mach. Theor. 18 23-35

Henning E H, Cavanagh P R, Albert H T and MacMillan N H 1982 A piezoelectric method of measuring the vertical contact stress beneath the human foot J . Biomed. Eng. 4 213-22

Hillis W D 1982 A high resolution imaging touch sensor In[. J . Robotics Res. 1 (2) 33-4

Howe R D and Cutlosky M R 1989 Sensing skin acceleration for slip and texture perception Proc. IEEE Int. Conf Robotics and Automarion pp 145-50

robot fingers under combined torsion and shear loading Proc. I E E E Int. Conf Robotics and Automation pp 103-5

Jacobsen S C, Wood J E, Knutti D F and Biggers K B 1984 The Utah/MIT dexterous hand work in progress Robotic Research ed M Bradv and R Paul (Camhridee. M A MIT

Hackwood S, Beni G, Hornak L A, W o k R and Nelson T J

Howe R D, Kao I and Cutlosky M R 1988 The sliding of

~,

Press) pp 601-53 Javawant B V 1989 Tactile sensina in robotics J . Phvs. E Sci. -

Instrum. 22 684-92

device for the denervated hand Biomed. Sci. Instrum. 10 9-10

Cambridge University Press) pp 202-32

Johnson J B and Huffman T R 1974 A tactile prosthetic

Johnson K L 1987a Contact Mechanics (Cambridge:

- 1987b Contact Mechanics (Camhridee: Cambridee University Press) pp 397-423

- I

Johnson K L. Greenwood J A and Hieginson J G 1985 The contact of elastic wavy surfaces In;.-J. Mech. Sci. 27 383-96

Johnson K 0 1983 Neural mechanisms of tactual form and texture discrimination Federation Proc. 42 2542-7

Journal of International Robotics Research 1988 Special issue on sensor data fusion 7 (6)

Kato I, Kudo Y and Ichimaru K 1977 Artificial softness sensing - an automatic apparatus for measuring viscoelasticity Mechanism and Machine Theory 12 11-26

Kerr J and Roth B 1986 Analysis of multifingered hands Int. J . Robotics Res. 4 (4) 3-17

Kinoshita G, Hajika T and Hattori K 1983 Multifunctional tactile sensors with multielements for fingers Proc. In( . Conf Ado. Robotics pp 195-202

exploration in one and two fingered robots Proc. IEEE In!. Conf Robotics and Automation (Washington, DC) (IEEE Comp. Soc.) pp 1806-9

Klatzky R L, Bajcsy R and Lederman S J 1987 Object

Loomis J L and Lederman S I 1986 Tactual perception

1015

Danilo De Rossi

Handbook of Perception and Human Performance vol2, ed K Bo8 et al (New York Wiley-Interscience) pp 1-41

Luo R C, Wang F and Liu Y 1984 An imaging tactile sensor with magnetostrictive transduction Proc. SPIE 521 264-70

Maalej N and Webster J G 1988 A miniature electrooptical force transducer IEEE Trans. Biomed. Eng. 35 93-8

Mitchell E E and Vranish J 1985 Magnetolastic force feedback sensors for robots and machine tools -an update, Proc. 5th Inl. Conf on Robot Vision and Sensory Control (Amsterdam) (Amsterdam: IFS) pp 187-94

Neubert H K P 1963 Instrument Transducers (Oxford Clarendon)

Nicholls M R and Lee M H 1989 A survey of robot tactile sensing technology Int. J . Robotics Res. 8 (3) 3-30

Novak J L 1989 Initial design and analysis of a capacitive sensor for shear and normal force measurement Proc. IEEE Int. Con5 Robotics and Automation (IEEE Comp. Soc.) pp 137-44

Pati Y C, Friedman D, Krishnaprasad P S, Yao C T, Peckerar M C, Yang R and Marrian C R K 1988 Neural networks for tactile perception Proc. IEEE Con$ Robotics and Automation pp 134-9

Pennywitt K E 1986 Robotic tactile sensing Byte Magazine 11 177-200

Petterson R W and Nevill G E 1986 The induced vibration touch sensor - A new dynamic touch sensing concept Robot Sensors Vol. 2 Tactile and Non-Vision ed A Pugh (New York: Springer) pp 219-28

Proc. Int. Conf Robotics (Washineton. DC: IEEE Como. Raibert R H 1984 An all digital VLSI tactile array sensor

I .

Soc.) pp 314-9 Raibert M H and Tanner J E 1982 Desien and -

implementation of a VLSI tactile sensing computer Int. J . Robotics Res. 1 (3) 3-18

Rebman J and Morns K A 1984 A tactile sensor with electro-optical transduction IEEE Spectrum 44 (9) 210-6

Robertson B E and Walkden A J 1986 Tactile sensor system for robotics Robot Sensors Vol. 2 Tactile and Non-Vision ed A Pugh (New York Springer) pp 89-97

Russel R A 1984 Thermal touch sensing J . EEE Austr. 4 (41) 68-70

~ 1985 A thermal sensor array to provide tactile feedback for robots Int. J . Robotics Res. 4 (3) 35-9

Salisbury J K 1984 Interpretation of contact geometries from

force measurements Proc. Int. Conf. Robotics (Washington, DC) (IEEE Comp. Soc.) pp 240-7

Salisbury J K and Craig J J 1982 Articulated hands: force control and kinematics issues Int. J . Robotic Res. 1 ( I ) 4-17

Sato N, Heginbotham M and Pugh A 1986 A method for 3D part identification by tactile transducers Robot Sensors, Vol. 2: Tactile and Non-Vision ed A Pugh (New York Springer) pp 133-43

sensor for manipulators Robot Computer - Integrated Manufacturing I ( I ) 65-71

Scott-Blair G W 1969 Elementary Rheology (London: Academic) pp 81-9

Shoemberg A A, Sullivan D M, Baker C D, Booth H and Galway C 1988 Ultrasonic PVF2 tranadimrs fnr ncnsing tactile force Medical Applications of Piezoelectric Polymers ed P M Galletti et al (New York: Gordon and Breach) pp 245-56

Stevens S S 1951 Handbook of Experimental Psychology (London: Chapman and Hall) ch 1

Sugiyama S, Kawahata K, Yoneda M and Igarashi I 1990 Tactile image detection using a IK-element silicon pressure sensor array Sens. Actuators A21-3 397-400

Tanie K, Komoriya K, Kaneko M, Tachi S and Fujikawa A 1986 A high resolution tactile sensor Robot Sensors Vol. 2: Tactile and Non-Vision ed A Pugh (New York: Springer) pp 189-98

Taylor M M and Lederman S J 1975 Tactile roughness of grooved surfaces: a model and the effect of friction Perception Psychophys. 17 23-36

Tikhonov A N and Arsenin V Y 1977 Solution of l l l -Posed Problems (Washington, D C Winston-Wiley)

Tomovic R and Stojiljkovic Z 1975 Multifunctional terminal device with adaptive grasping force Automatica 11 567-70

Vranish J M 1986 Magnetoresistive skin for robots Robot Sensors Vol. 2 Tactile and Non-Vision ed A Pugh (New York Springer) pp 99-1 1 I

Wang L R, Dario P, De Rossi D, Francesconi R and Pinotti P C 1984 Pyroelectric sensing station for automatic recognition of position and orientation of objects Int. J . Mach. Tool Des. Res. 24 (2) 145-60

Medicine (New York Wiley)

Schneiter J L and Sheridan T B 1984 An optical tactile

Webster J G (ed) 1988 Tactile Sensors for Robotics and

1016

instrument science and technology artificial tactile sensing and haptic … · 2006-02-28 · meas....

Documents