Tactile-Feedback Solutions for an Enhanced User Experience

Tactile feedback has been shown to greatly improve the touch-screen user experience, and the variety of actuation technologies available makes it possible in virtually any electronic device.

by Michael Levin and Alfred Woo

THE PROLIFERATION of touch-screen interfaces in consumer-electronics devices such as mobile phones is irrefutable, and it is easy to identify key attributes that make touch screens a compelling user-interface (UI) mechanism. Improved industrial design and mechanical durability, optimization of UI real estate, and the capability of leveraging a rich input-gesture vocabulary are a few that come to mind. However, the loss of tactile or haptic feedback, which users expect from conventional mechanical-input mechanisms, creates problems including higher error rates and user frustration. The solution is to add tactile feedback to the touch-screen interface to secure the best features of both touch screens and conventional controls. In fact, many device manufacturers have already done this; touch-screen mobile phones, cameras, and media players from companies such as Samsung, LG, Nokia, and RIM feature touch feedback to enhance the user experience. Haptic feedback systems have also been deployed in many non-consumer markets, including aerospace, automotive, casino gaming, medical, military, and industrial equipment. Other applications of haptics range from joysticks, steering wheels, and control knobs to touch pads, touch-sensitive switches, and keyboards.

Both quantitative task performance and qualitative user satisfaction have been shown to be significantly improved when haptic feedback is provided to the user. There have been many academic studies that confirm that tactile feedback bridges the usability gap between touch screens and mechanical controls. One of the most recent studies¹ concluded that the incidence of errors in touch-screen text entry was drastically reduced when tactile feedback was introduced. Moreover, subjects expressed a much lower frustration level with the enhancedinterface and perceived text-entry tasks to require less mental and physical effort.

Research by Motorola² evaluated user pref-erence for tactile and audio feedback on a mobile phone. Users were asked to compare audio-only feedback against synchronized audio and haptic feedback. More than 80% of the test subjects preferred the phone with both audio and haptic feedback over audio alone. The results also suggested that the haptic component enhanced the perception of audio quality.

In addition to augmenting the touch-screen interaction experience, haptics can also add a sense of touch to gaming for a more immersive experience, provide a rich non-audio channel for communicating alerts, and enhance the media consumption experience by adding a virtual subwoofer. The most satisfying UI feedback is multimodal, combining audio, visual, and tactile cues. In situations where audio or visual feedback is inadequate, such as distractive environments or in the case of vision- or hearing-impaired users, the tactile sense may be the primary mode of interaction.

Touch feedback has become a high-demand feature in touch-screen devices, and haptic actuation technologies are rapidly evolving as a result. The technology options can be confus-ing, but, in the end, the intended haptics usages and form-factor constraints largely determine the optimal actuation technology for the application. In the following sections, we outline the different actuation technologies and discuss the related hardware implementations.

Inertial Actuation

Inertial actuation is the most common and proven haptic actuation method. Virtually every mobile phone incorporates inertial haptic actuation; the vibrational alert is an example of simple tactile feedback using inertial actuators. What separates the typical vibra-tional alert from high-fidelity tactile feedback is the manner in which the actuators are driven.

Inertial actuators in mobile phones are typically one of two types: a rotational motor with an off-center mass [eccentric rotating- mass (ERM) actuator] or a spring-mass system that vibrates in a linear motion [linear resonant actuator (LRA)] (Fig. 1). ERM actuators are available from a wide range of vendors, including Sanyo, Mabuchi, Johnson Electric, and others. LRAs are a more specialized component, and the majority of the LRA actuators in use today are manufactured by Samsung Electro-Mechanics Co. (SEMCO).

Even when used as a simple vibration motor, the two types of actuators are driven differently. The ERM is typically a DC motor and is driven with a simple DC on–off signal. The LRA, however, is driven at its resonance, so the input signal will be sinusoidal at the nominal resonant frequency of the spring-mass system, typically about 175 Hz, which is close to the 200–300-Hz peak-frequency sensitivity of the Pacinian corpuscle cells.³ Pacinian corpuscles are nerve endings in the skin that respond to deep pressure and high-frequency vibration.

Creating high-fidelity haptics with inertial actuators is a matter of modulating the input waveforms to produce the desired mechanical response in the actuator. The dynamic response of ERMs can be drastically improved by using complex waveforms that may include momentarily overdriving the actuator to minimize rise time and actively braking to minimize spin-down time. A wide range of haptic effects can be created by using techniques that modulate waveshape, frequency, duration, and amplitude. However, to achieve the most compelling results, the waveforms must be tuned for the response characteristics of the specific actuator and integrated system. This step is particularly exacting and critical when trying to replicate the sharp feel of conventional mechanical snap domes or tact switches, which need to be very short and precise – the complete effect may last only 20 msec and the ERM shaft may rotate less than one full revolution. This precise control of actuator response distinguishes Immersion Corp.'s rich haptic feedback from the generic buzz of ordinary mobile-phone vibrational alerts.

Mechanically, the inertial motor is typically mounted to the device housing, and the inertial forces produced when the actuator is energized vibrate the entire device. This implementation is appropriate for small hand-held devices, such as mobile phones. However, it is also possible to incorporate inertial actuation into "fixed" or larger devices where only the touch screen is actuated. This requires a frame or holder that couples the actuator to the touch screen and some mechanical compliance in the system, typically a foam gasket, which allows the touch screen to vibrate independently of the housing. Immersion Corp. provides tools and services that support a range of inertial actuators for both handheld and fixed device applications.

The impact of the haptic system on battery life is largely dependent on the usage model for touch feedback. If the primary usage is to augment touch-screen interactions with tactile clicks and other short-duration effects, the power draw of the haptics is fairly small and may impact overall battery life by only a few percentage points. However, if long-duration effects are used for alerts or gaming enhancement, this impact will become more significant. Inertial actuators, in particular, can draw in excess of 100 mA of current during transients, but this is usually offset by the very short duty cycle.

Inertial actuation is well-suited to handheld devices in which vibrating the entire device is desirable, and for larger or fixed devices if the requisite additional mechanical components can be accommodated. It is a proven technology that can be usually implemented with off-the-shelf electromechanical components.

Piezoelectric Actuation

While inertial actuation has been commonplace for as long as mobile phones have existed, there is currently a technology trend toward piezoelectric actuation of touch-screen haptics. Piezoelectric actuation offers several potential benefits: thinner form factor, faster response time, and high bandwidth. It also holds the promise of higher-fidelity haptics, a larger "vocabulary" of haptics effects, and the ability to integrate touch feedback into more demanding form factors – thinner devices with relatively large screens. However, there are also challenges associated with designing and integrating products with piezoelectric actuators.

Piezoelectric actuators are typically ceramic materials that deform when a voltage is applied. They are manufactured in a variety of shapes, but the types most commonly used for touch feedback are beams and disks. Piezoelectric elements have long been used as speaker transducers and sensors, but the use of piezoelctric elements for haptics actuation is still new. Major suppliers of piezoelectric transducers include Hokuriku, Kyocera, Murata, NEC-Tokin, and TDK.

Regardless of the shape, the behavior of the piezoelectric element is similar – voltage applied across the device results in deformation. In a beam, the deformation is in the form of bending. In a disk, the actuator flexes to bow up or down in the center (Fig. 2). When mounted appropriately, either behind or at the perimeter of a touch screen or liquid-crystal display (LCD), this deformation causes displacement of the touch surface as it is pushed away from a fixed surface.

Fig. 1: Shown above are typical rotational and linear actuators used to create haptic feedback in mobile phones.

The choice of beam vs. disk is largely a mechanical-packaging consideration. The thickness of the piezoelectric elements is typically between 0.5 and 3 mm, and the footprint will vary depending on the size of the touch screen and the haptic performance required. The disk is best utilized behind the LCD because its opacity and relatively large footprint do not lend itself to integration above or alongside the viewable touch area. Beams can be used in either location, but minimizing sprung mass is important, so locating the piezoelectric beams outboard of the viewable area enables decoupling of the LCD from the sprung mass, improving performance considerably.

It is important to note that, unlike inertial actuation, piezoelectric-actuation systems can be designed to create haptic effects by either bending the touch screen or pushing the touch surface against another surface, resulting in translational motion. As a result, piezoelectric actuation is particularly well-suited for appli-cations in which only motion of the touch screen is desired and not vibration of the entire device. Provided that the ratio of the unsprung mass to the sprung mass (e.g., fixed portion of the device to the actuated portion of the device) is sufficiently high, adequate mechanical isolation between the two elements can be achieved to yield an experience that approximates a "touch-screen-only" feedback experience.

The high bandwidth of the piezoelectric actuator makes it capable of reproducing high-frequency transients, which are essential for very high-fidelity touch feedback. This property of piezoelectric materials can simplify the process of creating haptic drive waveforms because there is less need to compensate for the inertial actuators' mechanical performance limitations with clever waveform design. In fact, the piezoelectric output bandwidth is far higher than needed for haptics, and components of the drive signal above approximately 400 Hz should be filtered because they may create objectionable audio artifacts without improving haptic quality. Filtering high-frequency components also reduces power consumption because the piezoelectric element presents essentially a capacitive load – a high-frequency output tends to result in high current draw.

The fast response time of piezoelectric actuators can lead to a reduction in power consumption. While there are multiple factors that contribute to the quality of the perceived haptic effect, the magnitude of the acceleration of the touch surface is a primary determinant. Because, relative to inertial actuators, the piezoelectric system can reach a given acceleration more quickly, it allows the duration of the haptic event to be shortened while maintaining the perceived quality of the effect. This can reduce the duty cycle and help preserve battery life.

One of the challenges in designing piezo-electric-based touch feedback is supplying the voltages that are required to actuate the piezoelectric elements. A range of piezo-electric architectures are available, and they can require input voltages from the tens of volts to hundreds of volts. The lower-voltage piezoelectric elements are costlier, so the design process should take into account the tradeoff between piezoelectric-element cost and drive-electronics cost.

Mechanically, piezoelectric actuators can also be more problematic to integrate. Because they are ceramic, and in the case of beams, very thin, they can be prone to breakage during handling as well as in use. The risk can be reduced by laminating the beam to a thin brass or steel shim. This process, however, increases the component cost.

Piezoelectric-based haptic touch screens are still new to the market. SMK, in conjunction with Immersion Corp., has developed and marketed touch screens with integrated bending-beam piezoelectric haptics. Immersion Corp. has also publicly demonstrated piezo-electric-actuated touch-screen technology in a prototype tablet PC.

Surface Actuation

Surface actuation is a new technology invented by Pacinian Corp. that is based on electro-attractive forces to enable direct activation of the touch surface. The principle behind the system is basic physics: an attractive force is generated between two surfaces that have a charge differential. By optimizing the surface materials, spacing, and drive signal, high-fidelity feedback can be generated on a variety of transparent and opaque materials. Though simple in principle, each embodiment is unique, and a deep understanding of the system dynam-ics is required to develop an optimized system.

Unlike piezoelectric-actuation systems, in which discrete actuators provide the motive forces, the surface-actuation mechanism is integrated within the touch screen or touch surface. The surface-actuation topology presents some interesting advantages over the other methods. First, it is easier to achieve a consistent response profile over the entire touch surface. Second, the actuation force scales with the touch surface area, which makes surface actuation more compelling as the touch-screen area increases. Inertial and piezoelectric-actuation methods require larger or additional actuators as size increases; with surface actuation, this is not an issue.

As with piezoelectric actuation, the touch surface moves relative to the housing, providing feedback directly to the finger touching the screen and not vibrating the entire device. Other than the motion of the surface, the only moving part is a spring element that compresses during activation and restores the surface to its neutral position when the charge is reduced (Fig. 3).

Because of the mechanical simplicity, the reliability of this haptic subsystem is very high; it has been tested to more than 200 million actuations, with no degradation in performance. As with piezoelectric systems, the system response can be very fast, with frequency response from DC to more than 500 Hz possible for small surfaces. Drive waveforms can be generated digitally or directly driven from standard audio signals. Through optimization of the travel, spring, and drive signal, the system can replicate the characteristics and perception of high-quality mechanical switches and keyboards.

Fig. 2: Piezoelectric beams and disks flex when energized.

The additional space required for the active layers is only slightly more than the desired travel. In a well-integrated system with a typical touch-surface displacement of 0.2 mm, the increase in thickness can be less than 0.5 mm over a standard touch screen without tactile feedback. Other than two power leads and the drive electronics, which can be located remotely, there are no actuator components that extend outside of the active surface, thereby simplifying system packaging.

Power consumption for surface-actuated systems is very low, less than 10 mW for many configurations. However, although drive voltages vary depending on design requirements, they are higher than that for other solutions. Pacinian Corp. has partnered with semiconductor manufacturers to develop a family of appropriate drive components with a small form factor and accessible cost.

Summary

In the end, there is no single haptics technology that is optimal for every application. The best choice will depend on the space, performance, usage, and cost constraints of the target device. Each technology option presents a set of trade-offs that product designers should consider:

Ease of Integration: What is the size and shape of the actuators and drive components, and how flexible are the options to physically integrate the components within the device?

Control Flexibility: How simple is it to create the haptic drive waveforms and can the waveforms be created using a variety of methods to achieve a range of haptic results?

Cost: What is the projected development and manufacturing cost for the haptic system?

Performance: How faithfully can the system recreate complex haptic effects and what is the range and perceptual resolution of the haptic vocabulary?

Power Consumption: What is the projected power consumption of the haptic system, assuming an equivalent haptic effect strength?

Technology Maturity: How market-tested is the technology?

As can be seen from Table 1, inertial actuators excel in actuating the entire device, as is common in handheld applications. However, the performance and control flexibility of newer haptic actuation technologies dominate in "touch-screen-only" haptics implementations.

Touch feedback is increasingly being implemented to improve the quality of user-device interactions and to enable a more intuitive, engaging, and satisfying user experience. In 2009, touch feedback has become a core feature in many tens of millions of consumer-electronics devices. As touch interfaces become more commonplace and device manufacturers continue to focus on the user experience as a differentiating competitive advantage, touch feedback is expected to play a leading role. There is no better time than now to evaluate how high-fidelity touch feedback can differentiate a product or service. With the multitude of solutions available, there are implementation options to meet most system requirements.

References

¹S. Brewster, F. Chohan, and L. Brown, "Tactile Feedback for Mobile Interactions," Proc. CHI '07, 159–162 (2007).

²A. Chang and C. O'Sullivan, "Audio-Haptic Feedback in Mobile Phones," Proc CHI '05 Extended Abstracts on Human Factors in Computing Systems, 1264–1267 (2005).

³J. Patrick Reilly, Applied Bioelectricity: From Electrical Stimulation to Electropathology (Springer-Verlag, 1998), pp. 94-97. •