Sensory feedback from touch is critical for many tasks carried out by robots and humans, such as grasping objects or identifying materials. Electronic skin (e-skin) is a crucial technology for these purposes. Artificial tactile skin that can play the roles of human skin remains a distant possibility because of hard issues in resilience, manufacturing, mechanics, sensorics, electronics, energetics, information processing, and transport. Taken together, these issues make it difficult to bestow robots, or prosthetic devices, with effective tactile skins. Nonetheless, progress over the past few years in relation with the above issues has been encouraging, and we have achieved close to providing some of the abilities of biological skin with the advent of deformable sensors and flexible electronics. The naive imitation of skin morphology and sensing an impoverished set of mechanical and thermal quantities are not sufficient. There is a need to find more efficient ways to extract tactile information from mechanical contact than those previously available. Renewed interest in neuromorphic tactile skin is expected to bring some fresh ideas in this field. This article reviews these new developments, particularly related to the handling of tactile data, energy autonomy, and large-area manufacturing. The challenges in relation with these advances for tactile sensing and haptics in robotics and prosthetics are discussed along with potential solutions.

For very rough surfaces, friction-induced vibrations contain frequencies that change in proportion to sliding speed. Given the poor capacity of the somatosensory system to discriminate frequencies, this fact raises the question of how accurately finger sliding speed must be known during the reproduction of virtual textures with a surface tactile display. During active touch, ten observers were asked to discriminate texture recordings corresponding to different speeds. The samples were constructed from a common texture, which was resampled at various frequencies to give a set of stimuli of different swiping speeds. In trials, they swiped their finger in rapid succession over a glass plate, which vibrated to accurately reproduce three texture recordings. Two of these recordings were identical and a third differed in that the sample represented a texture swiped at a speed different from the other two. Observers identified which of the three samples felt different. For a metal mesh texture recording, seven observers reported differences when the speed varied by 60, 80, and 100mm/s while the other three did not reach a discrimination threshold. For a finer leather chamois texture recording, thresholds were never reached in the 100mm/s range. These results show that the need for high-accuracy measurement of swiping speed during texture reproduction may actually be quite limited compared to what is commonly found in the literature.

One human finger explored plastic Braille dots using a variety of velocity and force profiles. The fingertip friction forces were measured. Characteristics of the interaction were studied to explore the manifestation of the amplitude/duration interdependence of signals across velocity, normal force and dot height. Both amplitude, defined here as maximum tangential force, and duration, were seen to vary with velocity and normal force, however the integral of the tangential force over time was found to not have a strong dependence on either variable. When three consecutive dots of varying height were examined, the tangential force integral was not constant, but increased in proportion to height. We propose that the nervous system may use the tangential force integral as an invariant to recognise the same spatial asperity explored under different velocity and force conditions.

In previous vision-to-touch sensory substitution approaches, including most `electronic white canes', typical approaches include mapping space-to-space, space-to-intensity, or space-to-frequency. To our knowledge, however, mapping space to time-delay has not been considered. Yet, because organisms must anticipate impending collisions with obstacles or anticipate being contacted by approaching objects, many organisms have developed computational short-cuts where distance-to-target is assumed to be proportional to a time-span. This short-cut often manifests itself in low-level sensorimotor behaviours and perceptual mechanisms. We studied whether untrained humans would spontaneously employ such a short-cut to estimate distance-to-obstacle in the absence of vision. The observers pressed a push button and a tactile pulse was delivered to the hand with a delay proportional to the distance to an obstacle detected by an optical range finder that they wore. The observers were not informed of the nature of the coding but could freely probe the obstacle while walking toward the target. Upon randomized presentation of obstacle distances, the observers quickly calibrated their judgement of distance-to-obstacle and were able to estimate this distance within a range of four meters for a proportionality factor corresponding to a velocity of one m/s.

We have performed a psychophysical experiment to investigate differences in perceived object size when exploring the inside or outside of objects. The experiment consisted of five conditions, in which ten blindfolded subjects compared the size of circular disks and holes using either the index finger, two different probes, the finger-span method, or an infinitesimal virtual probe. The result showed significant negative biases for the conditions with the large probe and the finger-span method, meaning that an object felt on the inside should be larger than an object felt on the outside in order to be perceived as the same size. This indicates that subjects are unable to sufficiently correct for the diameter of the probe when exploring objects. At the same time, a general tendency was observed in all conditions that involved movement to feel the inside of objects as larger than the outside. This suggests that, in order to obtain a neutral estimate of object size in a virtual environment, one should use a virtual probe diameter of about 4% of the size of the object to be explored.

Microscale teleoperation with haptic feedback requires scaling gains in the order of 104−107. These high gains impose a trade-off between stability and transparency. Due to the conservative approach used in most designs, transparency is reduced since damping is added to the system to guarantee stability. Starting from the fact that series, negative feedback and parallel connection of passive systems is a passive system, a new approach is addressed in this work. We propose here a complete teleoperation chain designed from the ground up for full transparency and stability, including a novel self-sensing probe and a high fidelity force-feedback haptic interface. By guaranteeing the passivity of each device and assuming that the human operator and the environment are passive systems, a homothetic direct coupling can be used without jeopardizing the stability and provides best transparency. The system is experimentally demonstrated in the complex case of a probe interacting with a water droplet under human control, while accurately transcribing the interaction back to operator.

Vection refers to the illusion of self-motion when a significant portion of the visual field is stimulated by visual flow, while body is still. Vection is known to be strong for peripheral vision stimulation and relatively weak for central vision. In this paper, the results of an experimental study of central linear vection with and without vibrotactile feet stimulation are presented. Three types of vibratory stimuli were used: a sinusoidal signal, pink noise, and a chirp signal. Six subjects faced a screen looking at a looming visual flow that suggested virtual forward motion. The results showed that the sensation of self-motion happened faster and its intensity was the strongest for sinusoidal vibrations at constant frequency. For some subjects, a vibrotactile stimulus with an increasing frequency (a chirp) elicited as well a stronger vection. The strength of sensation of self-motion was the lowest in the cases when pink noise vibrations and no vibrotactile stimulation accompanied the visual flow stimulation. Possible application areas are mentioned.

This paper addresses the issue of producing localized tactile stimuli on a transparent surface. An approach based on time reversal of acoustic waves is presented and implemented on a thin glass surface actuated by piezoelectric transducers located at the periphery. The physical performance measurement, completed by a user study demonstrates the ability to provide localised and perceivable tactile stimulation.

Human grip forces are automatically adjusted upon occurrence of an external disturbance experienced by an object that is held by a thumb and index finger. We investigated some of the cues that may be used by the brain to perform rapid grip restabilization. To this end we ask subjects to grip and hold an instrumented and actuated parallelepiped-shaped handle between the index finger and the thumb. Under computer control, the handle could be jerked from the still grip and could independently provided vibration of 250 or 100 Hz to the gripping fingers. We found that the latency of the motor corrective action was 139 ms on average, but when a vibrotactile stimulation was applied 50 ms before the application of the pulling force, the latency was reduced on average to 117 ms. The average latency of the conscious response to the vibrotactile stimuli was 230 ms, suggesting that vibrotactile stimulation was capable of influencing the reflex action.

With the aim of augmenting auditory sensation by tactile stimuli, we investigated cross-modal relationships between the two modalities, focusing on frequency. The results showed that frequency consonance between tactile and audio stimuli depends on the relationship between harmonics, in a manner similar to auditory waves, but with broader peaks.

We investigated the perception of differences between direction-dependent, movement-opposing forces. The magnitude of these forces changed in whenever the direction of motion reversed. They were felt by participants during an experiment that required them to scan a virtual surface, represented by a planar haptic interface, via left-right motions of the index finger. We found that individuals are surprisingly insensitive to changes in opposing force magnitude that are contingent on reversals in direction of motion, despite large contrasts in force magnitude. Forces of 1 N failed consistently to be discriminated from forces of 0 N during sequential presentation at the highest speeds. As the mean scanning speed of the digit was reduced, the effect progressively vanished. The effect we observed is simple and robust enough to be demonstrated on virtually any haptic force-feedback interface. We suggest possible interpretations based on temporal information processing in the nervous system, on physiology and biomechanics, and through inferences that the nervous system may rely on to relate motor commands to sensory input during dynamic haptic interaction. The results obtained raise fundamental questions about the perceptual interpretation of kinesthetic stimuli involving rapid movement, and may also suggest a reconsideration of requirements for haptic interfaces.

Grasping is one of the most common forms of dexterity. So far, most research has focused on slow-varying loads which can be resisted by anticipatory grip adjustments. There are common cases, however, when a rapid, unexpected increase in the load occurs and where the central nervous system must re-adjust the grip dynamically to prevent slippage. During such events, the central nervous system reactively updates the grip force to minimize further escape of an object. While existing theories postulates that the shear strain of the finger pads caused by the load force is a primary source of information for detecting a new load condition, vibrations induced by even minute object slip in the hand might more effectively signal the occurrence of unwanted movement of the object relatively to the hand. With the help of a high-sensitivity force sensor interposed in the load-path of a fast traction-creating device, we recorded the fluctuations of the force projected onto the fingertip when a rapid perturbation was applied to a grasped object. These fluctuations are indicative of slip. The results highlight the existence of a correlation between the amplitude of the vibrations and the grip force modulation, when textural features are present. The study provides promising evidence that the central nervous system exploits vibrations to detect the onset of unwanted movement of an object relatively to the hand to optimally scale the grip force in response to unexpected, rapid load variations.

We evaluated the use of rich tactile feedback in the task of scrolling through a long list of items. We used a hand-held device having a tactile transducer that could provide sensations with temporal and spatial content. These capabilities were put to use in an interaction metaphor where input and tactile feedback were tightly coupled. We measured time- to-target and error rates, but also measured the time spent by participants to look at the screen. We found a 28% decrease of reliance on vision when tactile feedback was enabled.

Force-feedback interaction with a bowed string model can suffer critically from noise in the velocity signal derived from differentiating position measurements. To address this problem, we present a model for bowed string interaction based on a position-constraint friction. We validate the proposed model by comparing to previous work using off-line simulations, and show measurements of interaction on haptic hardware. This noise-free excitation signal leads to cleaner string motion than previous models, thereby improving the quality of force and audio synthesis in the presence of noise.

We propose extending the concept of spatial constancy to haptic perception. In vision, spatial constancy refers to the conversion of retinotopic signals into spatiotopic representations, allowing the observer to perceive space independently of his or her own eye movements, or at least partly so. The problem would seem at least as important in haptic perception, where sensory surfaces undergo even more complex movements in space. Here we develop a methodology for studying haptic spatial constancy, which involves a tactile display mounted on a mobile platform, and which allows us to decouple movements of the sensory surface - in this case the fingertip - from movements of objects on the fingertip. Using this apparatus, we find evidence for only weak haptic spatial constancy.

Fluctuations of the frictional force arising from the stroke of a finger against flat and sinusoidal surfaces are studied. A custom-made high-resolution friction force sensor, able to resolve milli-newton forces, was used to record those fluctuations as well as the net, low-frequency components of the interaction force. Measurements show that the fluctuations of the sliding force are highly non-stationary. Despite their randomness, force spectra averages reveal regularities. With a smooth, flat, but not mirror-finish, surface the background noise follows a 1/f trend. Recordings made with pure-tone sinusoidal gratings reveal complexities in the interaction between a finger and a surface. The fundamental frequency is driven by the periodicity of the gratings and harmonics follow a non-integer power-law decay that suggests strong nonlinearities in the fingertip interaction. The results are consistent with the existence of a multiplicity of simultaneous and rapid stick-slip relaxation oscillations. Results have implications for high fidelity haptic rendering and biotribology.

The tactual scanning of five naturalistic textures was recorded with an apparatus that is capable of measuring the tangential interaction force with a high degree of temporal and spatial resolution. The resulting signal showed that the transformation from the geometry of a surface to the force of traction and, hence, to the skin deformation experienced by a finger is a highly nonlinear process. Participants were asked to identify simulated textures reproduced by stimulating their fingers with rapid, imposed lateral skin displacements as a function of net position. They performed the identification task with a high degree of success, yet not perfectly. The fact that the experimental conditions eliminated many aspects of the interaction, including low-frequency finger deformation, distributed information, as well as normal skin movements, shows that the nervous system is able to rely on only two cues: amplitude and spectral information. The examination of the “spatial spectrograms” of the imposed lateral skin displacement revealed that texture could be represented spatially, despite being sensed through time and that these spectrograms were distinctively organized into what could be called “spatial formants.” This finding led us to speculate that the mechanical properties of the finger enables spatial information to be used for perceptual purposes in humans with no distributed sensing, which is a principle that could be applied to robots.

We describe a system which simulates in realtime the auditory and haptic sensations of walking on different surfaces. The system is based on a pair of sandals enhanced with pressure sensors and actuators. The pressure sensors detect the interaction force during walking, and control several physically based synthesis algorithms, which drive both the auditory and haptic feedback. The different hardware and software components of the system are described, together with possible uses and possibilities for improvements in future design iterations.

In the above titled paper (ibid., vol. 24, no. 3, pp. 527-536, Jun 08), it is stated in Sec. IV-D3 that the characteristic number of the first variant of the "god object" method applied to texture synthesis is 1. This statement is not always true, and clarification of this issue is presented here.

This article presents research on making schoolbook illustrations accessible for students with visual impairment. The MaskGen system was developed to interactively transpose illustrations of schoolbooks into tactile graphics. A methodology was designed to transpose the graphics and prepare them to be displayed on the STReSS2, a refreshable tactile device. We experimented different associations of tactile rendering and audio feedbacks to find a model that children with visual impairment could use. We experimented with three scientific graphics (diagram, bar-chart and map) with forty participants: twenty sighted adults, ten adults with visual impairment, and ten children with visual impairment. Results show that the participants with visual impairment liked the tactile graphics and could use them to explore illustrations and answer questions about their content.

A contact-mechanics analysis was used to explain a tactile illusion engendered by straining the fingertip skin tangentially in a progressive wave pattern resulting in the perception of a moving undulating surface. We derived the strain tensor field induced by a sinusoidal surface sliding on a finger as well as the field created by a tactile transducer array deforming the fingerpad skin by lateral traction. We found that the first field could be well approximated by the second. Our results have several implications. First, tactile displays using lateral skin deformation can generate tactile sensations similar to those using normal skin deformation. Second, a synthesis approach can achieve this result if some constraints on the design of tactile stimulators are met. Third, the mechanoreceptors embedded in the skin must respond to the deviatoric part of the strain tensor field and not to its volumetric part. Finally, many tactile stimuli might represent, for the brain, an inverse problem to be solved, such specific examples of “tactile metameres” are given.

Advanced, synthetic haptic virtual environments require textured virtual surfaces. We found that texturing smooth surfaces often reduces the system passivity margin of a haptic simulation. As a result, a smooth virtual surface that can be rendered in a passive manner may loose this property once textured. We propose that any texture algorithm is associated with a characteristic number that expresses the relative change in loop gain. We further found that a passive virtual interaction can have severe unwanted artifacts if the synthesized force field is not conservative. The energy characteristics of seven algorithms are analyzed. Finally, a new texture synthesis algorithm, which operates by modulating a friction force during scanning, is shown to have several advantages over previous ones.

This paper describes the design, construction, and initial evaluation of a handheld information device that supports combined tactile and graphical interaction. The design comprises a liquid crystal graphic display co-located with a miniature, low-power, distributed tactile transducer. This transducer can create electronically-controlled lateral skin deformation patterns which give the sensation of sliding over small shapes. It is integrated within a slider mechanism to control scrolling. It also functions as a detent when pushing on it. Tactile feedback and the combination of visual and tactile feedback in a mobile context enable the development of new functions, such as multimodal navigation within large graphic spaces

Mobile interaction can potentially be enhanced with well-designed haptic control and display. However, advances have been limited by a vicious cycle whereby inadequate haptic technology obstructs inception of vitalizing applications. We present the first stages of a systematic design effort to break that cycle, beginning with specific usage scenarios and a new handheld display platform based on lateral skin stretch. Results of a perceptual device characterization inform mappings between device capabilities and specific roles in mobile interaction, and the next step of hardware re-engineering.

We describe the redesign and the performance evaluation of a high-performance haptic device system called the Pantograph. The device is based on a two degree-of-freedom parallel mechanism which was designed for optimized dynamic performance, but which also is well kinematically conditioned. The results show that the system is capable of producing accurate tactile signals in the DC-400 Hz range and can resolve displacements of the order of 10 /spl mu/m. Future improvements are discussed.

When a progressive wave of localized deformations occurs tangentially on the fingerpad skin, one typically experiences the illusion of a small object sliding on it. This effect was investigated because of its potential application to the display of Braille. A device was constructed that could produce such deformation patterns along a line. Blind subjects' ability to read truncated Braille characters (‘○○’, ‘○•’, ‘•○’, and ‘••’) using the device was experimentally tested and compared to their performance with a conventional Braille medium. While subjects could identify two-character strings with a high rate of success, several factors need to be addressed before a display based on this principle can become practical.

Passivity theory is applied to the creation of synthetic, complex multidimensional haptic environments. It can be shown that under appropriate conditions, sufficiently high rendering rates can guarantee the passivity of a simulation produced by a haptic device coupled to a discrete-time realization of a nominally passive environment. The creation of a passive, globally defined, virtual environment is either analytically complex or computationally costly. A method is described whereby a passive environment is created from transitions between locally defined force models that encode static conservative force fields. This is applied to the haptic rendering of tool contact with deformable bodies, in which sparse force-deflection responses are used to define local models. Passivity, continuity, and fidelity are provided by response-function interpolation rather than by interpolation of forces, as in previous methods. The work also includes an illustrative example.

A method is described whereby the virtual haptic interaction with deformable elastic objects is created in terms of two processes: a slow process which carries on the simulation, and a fast process to render forces. Passivity theory is used to design an update strategy which reproduces exactly pre-computed responses between a tool and an object. This yields a design procedure for adjustable local models which guarantee the passivity of the interaction while preserving fidelity. Two examples of local models are given and some experimental results are reported.

Current visualization methods of volume angiograms are limited in their ability to display vessel connectivity and depth information readily available in the data set. In this paper, we introduce a hybrid graphic-haptic display technique that allows visualizing vessel connectivity as well as provides user assistance in tracing vessel branches. Most importantly, our display provides three-dimensional shape cues in both visual and haptic domains that eliminate the need for a rotational display for 3D perception. Issues related to haptic rendering, implementation, and experimental validation of the developed system are reported. This proposed multimodal visualization approach is independent of the devices used and can be applied to current visualization platforms for angiograms due to the small computational load it requires.

This paper presents a novel approach to compute the rank-deficiency locus of nonsquare Jacobian matrices. This algorithm is based on the computation of the singular vectors associated to zero singular values of the Jacobian. Examples are provided to illustrate the application of the algorithm. Results an shown for a four degree-of-freedom and a seven degree-of-freedom manipulator.

There are many engineering challenges that must be addressed in order to successfully integrate haptics with the environmental characteristics found in surgery. The most fundamental of these challenges is to achieve update rates of solid nonlinear deformable objects that are acceptable to the human haptic system. The paper presents a software architecture that is designed to meet these challenges by analysing the task, the haptic, and the hardware constraints of surgery simulation.

Experimental results are presented to show that SMA actuators are able to control forces both rapidly and precisely. An antagonistic pair of constrained actuators is shown to be capable of rapidly changing the set-point throughout a /spl plusmn/7.00 N range. Although the system entered into a limit cycle at the set-point, the limit cycle magnitude was small. The peak to peak amplitude of the limit cycles was only 0.07 N with a worst case average offset of 0.016 N at high force rates up to 75 N.s/sup -1/. Simulation results also indicate that accuracy could be improved with an increased sampling rate. The SMA actuator pair was also shown to be capable of accurate tracking. The maximum tracking error for a f=0.500 Kz, 2.00 N sine wave was 0.04 N. For the 2 Hz tracking, the maximum error was approximately equal to the boundary layer width which was set at 0.30 N.

This paper introduces a methodology to simulate the dynamics of deformable viscoelastic 3D bodies in real-time for haptic interaction. The method is based upon a finite element approach. The central idea in this scheme is to reduce the computation required in regions which are to the periphery of the region of interaction between the virtual haptic device and the virtual body. This is accomplished by implementing a multilayer finite element mesh. The top layer, or parent, consists of a coarse mesh of the entire body; child meshes represent sub-regions of the coarse mesh, but have a much finer resolution. By using equivalent impedances to relate the two meshes, it is possible to decouple the coarse and fine regions; this enables the system to not only to have different resolutions in different regions, but also allows the parent and child meshes to be updated at different frequencies. The multilayer mesh also addresses numerical integration issues.

This paper addresses the problem of reducing the hysteresis found in the actuation of most smart materials. They are divided in two groups: systems with no saturation (e.g. piezoelectric actuators), and systems with saturation (e.g. shape memory actuators). For the control of the first group the concept of phaser is introduced, an operator which shifts the phase of a periodic signal but keeps its magnitude unchanged. Since it is possible to approximate phasers with linear filters, it is possible to design practical compensators. The design of a phaser requires the knowledge of one parameter /spl phi/, easily identified from experimental transfer function estimates. For the second group, two phasers are used in a tandem connection. One phaser is designed as described before, and the second is designed so as to vary with the input. This compensation reduces the hysteresis to a single saturation. To show its effectiveness, simulation results are provided using the hystery model, then the method is applied to an SMA actuator.

An algorithm is presented which computes feasible manipulator trajectories along fixed paths in the presence of kinematic singularities. The resulting trajectories are close to minimum time, given an inverse kinematic solution for the path and bounds on joint velocities and accelerations. The algorithm has complexity O(M log M), with respect to the number of joint coordinates M, and works using "coordinate pivoting", in which the path timing is generated locally with respect to whichever joint coordinate is changing the fastest. This allows the handling of singularities, including linear self-motions (e.g., wrist singularities), where the path speed is zero but other joint velocities are non-zero. Examples involving the PUMA manipulator are shown.

The autonomous parameter estimation of a manipulator is considered with respect to both dynamic and joint sensor properties. Using methods based on adaptive control, a new formulation is introduced such that bench calibration of the robot joint sensors and actuators is no longer necessary. This method is unique because estimation is done with respect to invariant forces due to gravity loading. The method also guarantees convergence to the true values from arbitrary initial estimates; consequently, the algorithm can also be used for manipulator self test. Experimental results are presented which were performed on two links of a six degree of freedom hand-controller. Results show that angles can be recovered to an accuracy of /spl plusmn/1/spl deg/ in the absence of initial estimates.

Further development of a three level switching controller (Grant and Hayward, 1995) is presented. The controller was originally designed to drive a novel shape memory alloy (SMA) actuator consisting of a number of thin NiTi fibers woven in a counter rotating helical pattern around supporting disks. While the original controller performed satisfactorily, it was hampered by the presence of limit cycles at higher gains. By allowing the upper switching level to be proportional to the velocity if is possible to achieve a damped response, analogous to PD control for linear systems, effectively quenching the limit cycle. With this damping, it is possible to decrease the rise time by almost half and maintain the same steady state accuracy.

It is shown that in hybrid position/force control as well as in multi-robot cooperation and teleoperation, the desired behavior of the system can be defined in terms of a massless mechanism whose joints act as ideal position or force sources. Based on this remark, a simple and robust controller is proposed. Experiments validate the approach.

A simulator is developed to model and design high strain shape memory alloy (SMA) tension actuators. The simulator may be used predict characteristics of a given actuator, or to design its geometry under specifications such as force, speed, stroke and size. The accuracy of the model is verified experimentally in reference to an existing NiTi shape memory alloy prototype actuator. Having developed some confidence in the model, the performance of the proposed actuation mechanism is compared to other existing technologies. In particular, the force-displacement and speed characteristics of a micro-solenoid electro-magnetic actuator and a muscle-size pneumatic actuator are compared to those of the SMA actuators with same dimensions.

An algorithm is presented for computing the necessary time-scaling to allow a non-redundant manipulator to follow a fixed Cartesian path containing kinematic singularities. The resulting trajectory is close to minimum-time, subject to bounds on joint velocities and accelerations. The algorithm assigns a series of knot points along the path, increasing the knot density in the vicinity of singularities. Appropriate path velocities are then computed for each knot point. Two experiments involving the PUMA manipulator are shown.

We address the problem of constructing a boundary model of an object when the input consists of a set of points that lie on its surface. We assume that the points are acquired using telemetric techniques. Such data constitute a discrete sampling of the surface: a "cloud" of points. Supplemental connectivity information between those points is necessary if one is to reconstruct an approximation to the underlying object. We use the implicit information provided by the acquisition procedure itself to achieve this goal.

A novel shape memory alloy (SMA) actuator consisting of a number of thin NiTi fibers woven in a counter rotating helical pattern around supporting disks is first described. This structure can be viewed as a parallel mechanism used to accomplish a highly efficient transformation between force and displacement. The actuator overcomes the main mechanical drawback of shape memory alloys, that being limited strain. Two variable structure controllers are applied to a pair of antagonist actuators. The first involves a switching control input creating a sliding mode in conjunction with a linear control activated within a boundary layer in the vicinity of the set point. The second involves a multistage switching control that simplifies amplifier construction. Experimental performance results in the time domain are discussed.

The characteristics and the nonlinear dynamics of a high performance hydraulic actuator produced by ASI Inc. are described and modeled. When a feedback is applied for the regulation of output force, a limit cycle is observed. The existence of the limit cycle can a priori be attributed to one, or to a combination of, the four dominant nonlinear effect that were identified in these actuators. In order to pinpoint its origin, successive approximations are made to apply the describing function principle, so as to predict the onset of the limit cycle as function of the feedback gain. Given the experimental data, this method allows us to attribute beyond any doubt its origin to the electromagnetic hysteresis in the valve, which is based on jet-pipe technology. A multiple term lead-lag controller is designed and implemented to quench the limit cycle and improve the rise time of the force control by more than an order of magnitude.

The development of a haptic interface to address the tasks performed by people with small tools is described. Design issues are considered from the requirements, in terms of actuation, kinematics, motion transmission, sensing, and concept design.

A method is presented for autonomous kinematic calibration of a 3-DOF redundant parallel robot. Multiple closed loops are used in a least squares optimization method. Ill-conditioning, column scaling of the gradient matrix, and observability indices for the best pose set of robot calibration configurations are discussed. Experimental results are presented and compared with the results using an external calibration device.>

Some experiments as well as some applications relative to an algorithm for collision prediction, the dynamic urgency algorithm (DUA), are presented. The purpose of the experiments is to gain some insight into the behavior of DUA and illustrate its validity. A brief explanation of the algorithm is presented, and the tests which have been made are described. A discussion of the results is presented. Comparing DUA to the full sorting algorithm where tau is calculated, a lower growth of the computational time (up to a ratio of 1:80) is noted, with in turn some erroneous alerts. On the other hand, the sequencing scheme seems to be efficient in 2/3 of the time for the prediction of the most urgent pair. Furthermore, comparing DUA to the full sorting algorithm where the distance function is calculated, it is confirmed that including the rate of change of the environment is more reliable as far as the belief of the most imminent pair is concerned.>

A kinematic is presented that is fully parallel and actuator redundant. Actuator redundancy refers to the use of more actuators than are strictly necessary to control the mechanism without increasing the mobility. The uses of this form of redundancy include the ability to partially control the internal forces, increase the workspace, remove singularities, and augment the dexterity. Optimization takes place based on several objective functions. The kinematic dexterity, the forces present at the actuators, and the uniformity of the dexterity over the workspace are all investigated as potential objectives. Global measures are derived from each of these quantities for optimization purposes. Instead, optimization of several factors is done simultaneously by specifying a primary objective and minimum performance standards for the secondary measures.>

Traditional software design methodologies have been shown to have drawbacks in designing and implementing software systems for robotics. A novel dual-hierarchical object-oriented design methodology is presented, which is well suited to problems of this type. A practical example of the application of this methodology is presented, utilizing CLOS as the implementation vehicle. The methodology developed is shown to facilitate the programming and planning of complex robot tasks, and the provision of generic recovery procedures for exception handling.>

A techniques for transitioning between path segments that is tolerant to dynamic changes arising from sensor inputs is described. The main idea is to blend the segments together in a way that does not require advance knowledge of the paths. It is also possible to decompose the transition into an action which brings to rest the motion along the final path. By adjusting the timing of these two components, one may control the shape of the transition, in both time and space, so as to satisfy different task constraints.>

Consideration is given to a world made up of a collection of objects which are all moving with respect to each other. The goal is to design a system capable of reporting predicting all possible object collisions, given that all relevant information is available in due time. Previous approaches are based on the notion of a distance function that reflects the closest distance between objects in the world at any given instant in time. Explicitly including time in the representation makes it possible to obtain an algorithm based on the shortest possible time before the next possible collision. The algorithm deals with all pairwise interactions between objects, sorts the pairs with respect to their predicted collision time, and maintains the most-likely-to-collide pairs at the top of a stack. A novel kind of hierarchy in the representation of the world is thus introduced. To find the shortest possible time before a collision, the trajectory of objects is constrained by imposing bounds on the object's acceleration and velocity. All interacting pairs are classified into buckets that reflect the imminence of the collision. The computing cost is kept constant by reclassifying only one pair from each bucket at each time sample.>

The authors consider the problem of aggregating the geometrical information provided by sensors such as range finders, and its applications to scene modeling. The information consists of collections of three-dimensional surface points that form a discrete subspace of the objects-to-free-space boundaries within the world to be modeled. The ultimate goal is to obtain a valid surface model which can in turn be transformed into an efficient volumetric representation for solid-interference-detection algorithms. This representation should achieve search efficiency, compactness, and parallelism. The authors first introduce such a volumetric representation which uses a tetrahedrization of space. Then they show that surface connectivity is a nontrivial and essential element of a valid representation. To that end, they introduce a formal graph-theoretic definition for model validity which they use to guide the process of aggregating the different views. Finally, the topological problems posed by objects nonhomeomorphic to spheres, such as multiholed tori, are introduced.>

An expert system is developed which determines the best grasp configuration to pick up a rectangular box. This two-level system first chooses, from a predetermined set of grasps, the ones best suited for the task using a set of rules, and then computes a quality index for each possible configuration. The program is written in LISP and was tested for many different objects and situations. The grasps found are not necessarily optimal but are efficient since they are similar to the grasps a human would choose in the same situation. Because the program gives fast results and doesn't perform tedious computations, the best grip found can easily be used as a suboptimal solution or as a starting point for an optimizing program.>

This paper explains a simple method for fast collision detection in manipulator tasks. We show from examples taken in the literature that solutions to this problem can be chosen among a continuum of schemes, according to the method selected for representing the workspace and the robot, and the amount of computations performed before testing a particular trajectory. We then describe a method based on a recursive decomposition of the workspace, also referred to as an octree model, as a good tradeoff for a class of applications.

RCCL is a robot programming system that enables a user to specify robot manipulator tasks employing a set of primitive system calls similar in spirit to those of the UNIX input-output system. The goals addressed in the RCCL system are: manipulator task description; sensor integration; updatable world representation; flexibility; wide range of applications; medium level robot programming; off-line programming; efficiency; manipulator independence; portability; foreground-background programming; Cartesian path programming; arbitrary path specification; tracking; force control.

An original system to perform an automatic analysis of assembled programs is presented. Executable programs are analyzed from the description of the machine on which they run and are translated into an intermediate language taking into account the particularities of the considered machine. The system was primarily designed as the first step of a project for transferring programs from one machine to another. The final goal of the project is to achieve an even utilization of computer resources for a real-time controlled robot, on the basis of partially dedicated processors. At the present time, the actual implementation provides a tool for studying the theoretical aspect of machine-level program analysis. Nevertheless, other applications can be found in program debugging and assembled program validation.