Get Complete Project Material File(s) Now! »

Requirement for a good haptic feedback design

In order to produce a good haptic feedback device, several requirements need to be taken into account. It is difficult to produce a complete set accurate for all devices as the ideal requirements vary for different devices and depend on the application. Some requirements have been compiled from the following source (Massie & Salisbury 1994). A common goal as set out by (Mahvash & Hayward 2005) describes the need to allow unimpeded motion but to be able to exert high fidelity forces and torques. In other words, when the user moves through free space, nothing should be felt, and when a hard virtual surface is contacted, it should feel perfectly rigid. To do this, the device must be able to resolve extremely small changes in position and force. Also, the mechanical constraints for the device should include low inertia, backlash, weight and friction.
When using a device it is important that if the contact point of the device is not colliding with anything in the virtual world then negligible friction should be perceived. A friction level of 5% of the force range was set by (Adelstein & Rosen 1992) during their design phase. Low frictions enable the user to move the device freely when this one is not in contact. Obtaining negligible frictions can be a problem particularly when high stiffness is required (Hayward & Armstrong 2000). In fact, high stiffness implies a stiff mechanical interface which needs to be constructed from metal. These heavy materials increase the friction as well as increasing the overall weight of the device. This provides a conflict between obtaining high stiffness while keeping low friction. When considering the construction of a force feedback device, two designs are possible, i.e. serial and parallel mechanisms. Serial mechanisms have a larger work space and are easier for control implementation. The downside is that as each section is added to the chain, the total inertia increases and the total stiffness decreases. Parallel mechanisms do not exhibit the above problem and have a much higher stiffness. The disadvantage over serial mechanisms is that the mechanism’s elements can physically interfere.
Ensuring the device is statically balanced is another design consideration. This means that the centre of mass of the moving parts remains stationary regardless of movement. If this property can be achieved then there is no need for active gravity compensation and the average torque required from the actuators is zero. The materials used in such devices need to be considered. Some parts of the device will need to be strong enough to take the stress and strain applied by the user and the actuators.
Another consideration, which affects the overall weight of the device, is the type of actuators incorporated. Actuators, as will be discussed in detail in section III.1, provide the forces for the device. In most cases, a good actuator should be compact and light as well as capable of producing the necessary power to deliver the forces. There are tradeoffs between power, volume and weight since actuators capable of producing large forces are generally heavier and are larger in size than those capable of small forces. The power-to-weight ratio of the actuators used is a critical factor when considering portable devices, particularly when multiple degrees of freedom (DOF) are required. For certain devices these constraints can be relaxed. For example, on motion platforms used in flight simulators, the weight and size of the actuators are less critical since low power output is required.
Often haptic devices will be used for long periods of time and so they must be comfortable to use. The weight of a device needs to be taken into consideration. If some components are very heavy then moving them around may cause fatigue in a short space of time, particularly if the device is carried out the user. The position in which the user is expected to operate the device also needs to be considered. For small desktop devices, it is important that the hand and wrist can be positioned comfortably. Many small devices allow the wrist to be supported by the surface on which they sit, similar to operating a mouse. According to (Mahvash & Hayward 2005), a study of wrist motion was undertaken to show that a square region of 150 mm side could be utilized as the workspace area and that 50 g masses were acceptable for applications requiring approximately 30 min to complete. For larger devices, ensuring a comfortable position for operation can be more of a problem. The workspace of such systems should be sufficient and should not restrict the user in natural movement. For the motion of larger devices such as arm exoskeletons, the characteristics of the human arm need to be noted. Moreover, in order for humans to perceive forces smoothly, the device must match or exceed the human sensing resolution.
The following section will examine closely the human sense of touch since it is necessary for the haptic feedback design.

Human haptic perception

Human haptic perception, comprised of tactile and kinesthetic perceptions, is the process of acquiring, interpreting, selecting, and organizing haptic sensory information. Kinesthetic perception refers to the sense of force and motion within the muscles and tendons, whereas tactile perception specifically concerns the acquisition and interpretation of sensations realized through the mechanoreceptors of the skin.
Many scientists have studied human perception thresholds in order to understand the limits of our abilities. Current studies of the just-noticeable differences (JND) for kinesthetic and tactile senses have focused on discernment of geometries, textures, and volumetric properties of objects held by the human (Allin et al. 2002). The JND (also referred to as the difference limen or the differential threshold) is the smallest change in a specified modality of sensory input that is detectable by a human. The Weber fraction defined as the JND divided by stimulus intensity is a common parameter used to evaluate the performance of the discrimination. It is usually assimilated to JND (%) in literature.
Early kinesthetic studies by (Clark & Horch 1986) investigated human perception of limb positions. The authors concluded that humans are capable to detect joint rotations of a fraction of a degree performed over a second of time interval. (Jones et al. 1992) also reported the JND for limb movement as 8 % (of stimulus intensity). Further psychophysical experiments conducted by (Tan et al. 1994a) determined the JND for the finger joints as 2.5 %, for the wrist and elbow as 2 %, and for the shoulder as 0.8 %.

READ  Characterization of potential MtNODULEROOT1 and MtNODULEROOT2 interacting partners participating in nodule and aerial organ development

Sensory motor control

In addition to tactile and kinesthetic sensory channels, human haptic system also includes a motor subsystem which is based on control body postures, motions as well as forces of contact with objects. In performing manual tasks of real or virtual environments, contact force is perhaps the most important variable that produces an effect both tactual sensory information and motor performance.
The key aspects of human sensory motor control are maximum force exertion, sustained force exertion, compliance, hand mechanical impedance, and sensing and control bandwidth.

Maximum force exertion

Several studies have been conducted to measure the controllable manual maximum force exertion. (An et al. 1986) found a maximum power grasping force of 400 N for males and 228 N for females. Power hand grasping force measurements are illustrated in Tab. 1.1.
Note that the maximum force exertion capability is dependent on the user’s posture. It was found that maximum force exertion grows from the most distal joint in the palm to the most proximal one (shoulder). In order to ensure user safety, a haptic interface should never apply forces that the user can not successfully counter.

Sustained force exertion

It is important to remember that humans can exert a maximum force only for a short period of time before the onset of fatigue. Muscle fatigue in turn adversely affects both sensing and motor control, and eventually leads to discomfort and pain. Another negative effect of prolonged exertion of high forces is the shift in force perception. It is therefore necessary to determine what force can be sustained comfortably by users for long durations. This information will help to determine the forces that haptic interfaces need to produce during task simulations. Tab. 1.2 shows the sustained forces of human hand, determined by (Tan et al. 1994b).

Table of contents :

I. Introduction
II. Specifications of Haptic Interface
II.1. Haptic rendering algorithms
II.2. Requirement for a good haptic feedback design
II.2.1. Human haptic perception
II.2.2. Sensory motor control
III. Technology of Haptic System
III.1. Actuation
III.1.1. Classical actuators
III.1.2. Novel actuators for haptic devices
III.1.3. Comparison of the different technologies
III.2. Concepts and examples of haptic devices
III.2.1. Non-portable force feedback
III.2.2. Portable force feedback
IV. Haptic Feedback Applications
V. Conclusions
I. Introduction
II. Teleoperation System Building Blocks
II.1. Human operator model
II.2. Environment model
II.3. Controller model
II.4. Master/Slave model
II.5. Transmission line
II.6. Formulation using a two-port presentation
III. Basic Control Architectures
III.1. Four-channel (4CH) diagram
III.2. Two-channel (2CH) diagrams
III.2.1. Position error based (PEB)
III.2.2. Force error based (FEB)
III.2.3. Direct force reflection (DFR)
III.3. Three-channel (3CH) diagrams
IV. Performance
IV.1. Tracking errors
IV.2. Bandwidths
IV.3. Scaling product
IV.4. Transmitted impedance
IV.5. Transparency
V. Stability
V.1. Stability analysis with known operator and environment models
V.2. Stability analysis with unknown operator and environment models
VI. Applications
VI.1. Handling hazardous material
VI.2. Underwater vehicle
VI.3. Space robots
VI.4. Micro-surgery
VI.5. Mobile robots
VI.6. Future trends
VII. Conclusion
I. Introduction
II. Performance Specifications
III. Actuation
IV. Sensing
IV.1. Position sensor
IV.2. Force sensor
IV.3. Pressure sensor
V. Pneumatic Valves Technologies
V.1. Servovalves
V.1.1. Operating principle of jet-pipe pneumatic type servovalve
V.1.2. Characterization of the Atchley 200PN-176
V.2. Solenoid valves
V.2.1. Overview of Matrix – high speed valve technology
V.2.2. Speed-up control techniques
V.2.3. Characterization of the Matrix valves – series 820
VI. Presentation of prototype
VII. Conclusion
I. Introduction
II. Model of Pneumatic System
II.1. Model of the actuator
II.2. Model of the mass flow rate
III. PWM Control Design
III.1. Basics
III.2. Force tracking controller design
III.3. PWM-based teleoperation control
IV. Hybrid Control Design
IV.1. Model-based control design for a single pneumatic manipulator
IV.1.1. Hybrid control principle
IV.1.2. Application to a pneumatic system
IV.1.3. Simulations
IV.1.4. Experiments
IV.2. Hybrid bilateral control for a pneumatic teleoperation system
IV.2.1. Implementation of the force inner loop in the 4CH architecture
IV.2.2. Simulations
IV.2.3. Experiments
IV.2.4. Stability discussion
V. Sliding Control Design
V.1. Teleoperation based on three-mode control scheme (3MCS)
V.1.1. Open-loop model of the master and slave devices
V.1.2. Closed-loop teleoperation system
V.1.3. Experiments
V.2. Extension to a five-mode control scheme (5MCS)
V.2.1. Controller mode selection
V.2.2. Comparison between the 5MCS and the 3MCS
VI. Conclusion
I. Introduction
II. Model of Pneumatic Actuator and Servovalves
II.1. Theoretical model
II.2. Comparison of simulation and experimental models
III. Bilateral Control Based Master-Slave Telemanipulator
III.1. Tangent linear model of pneumatic actuator
III.1.1. Equilibrium set
III.1.2. Linear model setup
III.1.3. Reduced tangent linear model
III.2. 4CH teleoperation controller design
III.3. Experiment and discussions
III.3.1. Experimental setup
III.3.2. Experimental results
IV. Bilateral Control Based on Force Observer
IV.1. Implementation of the HOB and EOB schemes
IV.2. Experiment results
V. Conclusion


Related Posts