Integrating Haptic Feedback into Design
Create interfaces that give users greater functionality than ever before.
Touch-sensitive surfaces have become ubiquitous in our modern world. Although touchpads made their debut in the early 1980s as a mouse replacement for portable computers, it wasn’t until the advent of the iPhone that touchscreens captured the imagination of consumer electronics device manufacturers and the world at large. The reason for their appeal is simple: they remove the need for any special purpose peripherals (like mice or styli) and make commanding and controlling a computer or machine as direct as pointing and selecting. Because they are programmable, they also enable a single physical surface to act as a window onto a virtually infinite variety of applications.
Although the revolution enabled by touch-sensitive surfaces is most visible in our mobile devices, this technology is becoming commonplace in automotive interfaces (both as touchpads and touchscreens), wearable devices and game controllers. As component prices drop, this technology has the potential to replace nearly every control interface including those found in white and brown goods.
At Immersion Corp., my team and I have spent nearly 20 years thinking about, prototyping and experimenting with touch interfaces for nearly every application imaginable. We have learned some valuable insights that will help device designers transition to touch-sensitive interfaces, particularly the value of including tactile feedback in these interfaces. This article will highlight some of the key considerations that designers need to take into account when contemplating this transition.
Replacing a Button
Buttons and button arrays are both the most common control interface and the most common candidate to replace with a touchscreen or touchpad in new product design. Each button or switch has a design intent and control imperative that they govern. As a designer considering migrating or creating an interface based on a touch-sensing technology, it is important to understand the meaning and purpose of the button interface:
- Usability of layout – how many functions does a button array need to support?
- Naturalness – What are user expectations about this button interface? Have they seen similar interfaces?
- Cost of error – If the user wants to undo an action, is it possible? What is the cost associated with this?
- Context of use – Is the button in a wet environment? Is it in a loud or high-vibration environment?
Figure 1. A basic haptic system
When the user control imperative is very well defined and unchanging, button interfaces can satisfy these considerations successfully using static button layout and fixed functionality. This is a very common solution in existing white and brown goods markets. By using very low-cost mechanical solutions: snap domes, contact switches or micro-switches in fixed configurations, highly specific interfaces can be efficiently and effectively achieved. In this case, replacement with a touch surface is unlikely to provide any ROI.
When the user control imperative is variable, context dependent or needs to change over time, static button layouts can be replaced by dynamic touchpad or touchscreen interfaces. This transition happened more than a decade ago in mobile devices and is now well under way in automotive secondary controls, where touch interfaces are rapidly becoming mainstream. Using a touch surface for this type of control interface enables a dynamic, responsive user interface, but it also places responsibility on the designer to create a meaningful and confusion-free user interaction. In our experience, an important driver of a user’s response to a touch interface is providing effective tactile feedback (haptics) that confirms the interaction.
Many user interactions can result in system-level changes, such as an increase in volume on a TV but there are many interactions that do not have obvious or immediate system feedback. Consider, for example, a remote switch for a generator or a valve, or enabling a home automation program remotely. It’s the responsibility of the interface to provide users with confidence that their intention was received and enacted by the system or device as well as feedback about the resulting status of the system or machine. Because they are programmable, haptics can go beyond simple, confirmation-style mechanical feeling button feedback by also providing status and state information to users during and after the touch surface interaction.
Basics of a Haptic Button
There are a few important design considerations when given the power and flexibility of a haptic display in a button replacement context. Primarily, it is important to effectively simulate a button sensation so that all users can understand that the interface has registered their intent. Crafting the perfect confirmation effect is much harder than just pulsing a motor attached to the touch surface. It requires careful consideration of the fidelity, mechanical ground path and proper signal tuning. A basic haptic system is shown in figure 1.
Most mobile phones have high-quality haptic actuators capable of accurately simulating the sensation of a mechanical button. This is evidenced by the fact that many iPhone 6S users didn’t realize that the home button is not a real button, but a touch surface with haptic feedback (and many still don’t). In the world of haptics, this level of mechanical fidelity is challenging to achieve for mass-market products. This is why many devices with haptics feel buzzy or even somewhat irritating when providing confirmation type feedback. Figure 2 shows the difference between a mechanical feeling haptic effect and one that is buzzy and indistinct. Mechanical switches are able to provide this level of crispness by virtue of their mechanical design. This is the level of quality that designers should target to create confident user interactions with touch surfaces.
Figure 2 shows the difference between a mechanical feeling haptic effect and one that is buzzy and indistinct.
One of the main challenges for device designers integrating haptics into a touch surface is mechanical isolation. Mobile devices are handheld and typically have a low mass so that a haptic actuator can easily generate accelerations that feel mechanical. For white goods, automotive and other IoT applications, there is a large immovable mass, which makes generating switch-like sensations challenging. In practice, this issue is solved by isolating and suspending the touch surface from the frame (wall/car/chassis) of the device. The smaller the moving mass, the easier it is to create high quality, mechanical feeling sensations with low-cost actuators.
Finally, it is critical that any haptic system be driven by a control signal that is responsive to the natural manufacturing tolerances that occur during actuator production. High-quality haptic actuators such as those used in mobile and automotive applications are resonant mechanical systems. This means that there is a particular frequency at which these actuators work best. Unfortunately, this frequency varies depending on manufacturing tolerances, and this variance can result in dramatic differences in fidelity and strength. In order to adapt to these variances, it’s necessary to utilize a haptic driver that includes some type of feedback circuitry to enable automatic tracking of the motor frequency. Figure 3 shows the difference between motors driven at their nominal frequency and a circuit that adapts to manufacturing tolerances.
Beyond confirmation feedback to users, touch surfaces can provide feedback to users at the system level. This is a critically important distinction. Mechanical switches provide tactile feedback to users in a physical and unchanging way, whether or not the user imperative is possible or was executed (i.e. buttons on a washing machine feel the same regardless of the machine state). Moving from buttons to touch surfaces means that users can now be presented with direct feedback about their actions that has significantly more meaning than with mechanical buttons.
Figure 3 shows the difference between motors driven at their nominal frequency and when driven by a circuit that adapts to manufacturing tolerances.
Consider, for example, an IoT wall switch embodied as a touchpad with a haptic actuator and intended to be mounted in a standard wall box. In this system, basic button-like mechanical feedback could alert the user that the system has toggled from an “on state” to an “off state” or vice-versa. If the user long-presses or taps the touch surface, they might activate other system states. In that case, haptic feedback could provide varied feedback to confirm to the user that their gesture was recognized and enacted or an unambiguous error sensation that it was not enacted. This is not possible with traditional mechanical switches. The number and type of effects is truly staggering and, as long as the system can render high-fidelity sensations, this feedback will be meaningful and distinct for users.
As white and brown goods and other IoT interfaces become increasingly computerized and connected, controls will necessarily become more contextualized. The usability and cost effectiveness of touch interfaces will become greater than mechanical buttons in these markets. By subscribing to key design principles outlined above, device designers can create interfaces that give users confidence, delight and greater functionality than ever before.