Design of a Wearable Tactile Display

Tactile displays are a viable way for people to interact with wearable computers. Human tactile perception is robust. A variety of shrinking tactile stimulator (tactor) technologies are available. Tactile displays are uniquely appropriate for wearable applications because of their close proximity to our 20 square feet of touch receptors: our skin. Tactile displays can solve issues of intrusive computers and multiple demands on user visual and audio attention. They are discreet and seamlessly integrate with most human activity. Tactile displays will neither conflict with nor replace audio and visual display but rather support information on these other displays and fill in the gaps where necessary. This paper presents our work in optimizing the design of a tactile display and discusses some of the issues and opportunities surrounding tactile displays for wearable computers. Additionally, we hope to inspire more work in this area. INTRODUCTION The Wearable Group at Carnegie Mellon has been designing and testing wearable computers for industrial and military maintenance applications for 10 years. These applications find our users in warehouses, aircraft hangars in cockpits and vehicles in rain or bright sun. Diverse and extreme situations such as these open the user to uncontrollable variables such as lighting, ambient noise, weather and a plethora of distractions. As a result the interaction and interface with wearable and mobile computers continues to be a major area of research, development, and innovation [1,2,]. Issues of user interaction in a wearable context are often successfully addressed with multiple modalities for input and output [3,4]. Typically these interaction modalities are various combinations of audio and visual for both input and output. Often the design of interaction with wearable computers is driven by a need for users to have their hands free [5]. Our experience prompted the question, what about when a user must be eyes free or ears free? In addition to situations where users are deaf, blind or both, environmental factors might prohibit audio or visual input and output. At certain times, making noise or diverting the eyes to a computer display might be socially inappropriate or physically unsafe. The use of touch as an additional modality is one opportunity to address these issues. Some good work has already been done in considering the use of tactile displays for wearable and mobile computers [6,7,8]. Additionally, researchers in the Navy have been working with tactile displays to convey information to pilots for over 10 years [9] There is also considerable reference material from tactile work for Virtual Reality (Haptics) and (Sensory Assistive) Devices for Deaf-Blind people. (See below) We define a tactile display as a device which presents information to the wearer by stimulating the perceptual nerves of the skin. A familiar model for this is the ‘silent‘ function on most pagers and cellular phones. In this model, one signal is presented through vibration. Our hypothesis is that with multiple addressable tactile stimulators (or tactors) spread across an area of the body we will be able to convey more complex and coordinated information. Haptics and Sensory Assistance vs. Tactile Information Display Our work is in the area of Tactile Information Display. This work is different from the work of both the Haptics community and work in sensory assistance for the deaf blind communities. Haptics usually describes any computer peripheral that adds the tactile sense to a Virtual Reality or Tele-operations application. These devices provide a force feedback or vibration based on cues from a visual interface. For example, a haptic device would be a desktop mouse that provides some force feedback creating the sensation of an ‘edge’ when the mouse is rolled over the edge of a window [10]. Tactile stimulation takes its cues directly from the visual interface. Similarly, Sensory Assistive devices aim to aid the deaf or blind in their ability to perceive the world around them. Here tactile stimulation is used to translate visual or audio information. A sensory assistive device takes visual images and creates a texture map display [11] whereby there is direct correlation from the image to the texture map. Both Haptics devices and Sensory Assistive devices are concerned primarily with a direct translation of real or computerized visual or audio information into tactile stimulation. Whereas, a tactile information display is neither direct nor a translation – it will present coordinated tactile information – not directly based on visual or audio information. Tactile information displays employ a different channel for communication between humans and computers. Currently pagers and cell phones employ a tactile display of notification. This ‘vibrate mode’ does not take cues from a previously computerized interface experience. It takes cues from real life. Information design and learning have occurred [12]; this allows a user to understand that when this device vibrates, it is someone trying to get their attention. The metaphor here is that vibrations in your pocket can be akin to someone shaking your shoulder to get attention. This vibration is the tactile display of tactile information. The differentiation between Virtual Reality Haptics and a vibrating pager, is a subtle but important one. We foresee a need for more exploration and experimentation in the area of tactile information design. This will give us a better understanding of the difference between the different kinds of information that can be presented through touch and how that information can be presented. Ultimately so that we can design effective tactile displays. In this paper we present a tactile display design and discuss the application framework for using these displays for tactile information design experimentation. TACTILE DISPLAY DESIGN Our initial efforts in tactile display design included research into temperature, electric stimulation, compressed air, and vibro-tactile means for stimulating the skin. Other research will not be discussed in this paper. Vibro-tactile methods were found to be the most promising because of their small size, and weight, low mechanical and electrical requirements. Vibro-tactile motors are also best suited to meet our Tactile display requirements (see table 1). In addition to tactor selection, our early efforts have helped to identify other requirements: 1) tactors must be held tight to the body, 2) general locations for tactors, and 3) details that help create a flexible vest design that will support multiple experiments. These initial designs are customized for wayfinding or navigation applications. Directions on how to get from here to there are presented as tactile information, steering the user through speed distance and turns. Navigation using a tactile display has a simple information set to begin experimentation. Forward, Backward, Left and Right in addition to information about speeding up, slowing down is a great starting point for our research. Figure 1. Early Tactile vest designed during the ‘99-‘00 academic year. Figure 1 depicts our first tactile display vest design. Tactors are contained inside hemispheres on the harness. These tactors are modified piezo buzzers from Radio Shack and are wired (down the back) to a belt-worn infrared receiver. A customized remote control has buttons to turn each Table 1. Wearable Tactile Display requirements: Light weight Silent Tiny, very tiny Low power Tactors can be felt through clothing Tactors must be held tight on the body Physically discreet Support experiments in tactile information design, i.e. flexible! tactor on and off. The piezo buzzer is rated at 12 volts DC and 15 milli-amps operational specifications. In addition, the buzzer was rated at a sound level of 75 decibels and buzzer tone of 300 to 500 Hertz. Buzzer dimensions are .59 x .59 x .98 inches. This design was loud, bulky and consumed too much power. For these reasons, it was not a reasonable tactile display. We do like the overall harness styling. It is comfortable (for someone with a 32” chest), easy to get on and off, and the design highlights the embedded technology. The tactile display design in this paper represents an optimization and application of all that we learned in our initial efforts. For this newer design we were able to locate and customize a tactor that is smaller, silent, and uses less power. A wireless infrared kit allowed us to create a testable vest that did not require any programming. Additionally, we created a vest design that can be customized for