Haphone, LLC

Mopad Usability Theory

The design of Haphone Mopad began not with the question How do we make a better mobile phone (personal music player, PDA, etc.)?, but rather with the question, How do we make the ultimate personal information device (UPID)?

Others have asked this question before and produced various visions (e.g., wearable computers, and handheld UPIDs). However, Haphone’s developers produced a new vision – one based on the limitations imposed by user modes, and limits on user desires for availability.

To explain our vision, we’ll first examine the ideas of availability, occupation/encumbrance, and user modes. We’ll then explain our conclusions regarding requirements and design, and how these lead to and define our vision – a vision which is realized in our product, Mopad.

Availability

Availability of an application or device has two aspects: presence, and ease of access.

Presence seems simple, but has a practical side which is often under-appreciated by designers. It is not enough for an application to be theoretically present, it must also be present in practice. For example, the Apple Newton supported more applications than the original Palm PDA (the Palm Pilot). But because the Newton was so large, users found it too cumbersome to carry with them on a regular basis. Thus, though the Newton theoretically had more presence then the Palm Pilot, in practice it had less.

Ease of access means simply how quickly and easily a user can access an application or device. For example: wristwatches and pocket watches can be equally present, but a wristwatch is easier to access than a pocket watch.

One of the typifying characteristics of UPID visions is that it all applications should be available at all times (anytime, anywhere). Wearable computers, for example, try to achieve this by placing a complete computer on the user’s person.

We, however, believe that the goal of anytime, anywhere is overemphasized. While this a desirable goal for some applications, this is not a desirable for other applications. The differentiation between applications that should and should not be available “anytime, anywhere” depends largely on user modes, which are explained below.

Occupation

Applications (and devices) can be characterized by their level of occupation. There are several types of occupation: manual, visual and auditory. Degree of occupation depends on amount (number of hands, size of screen, percent of sound input) and duration (what percent of the user’s time is taken up by the activity).

User modes (see below), applications and devices can all be characterized by their levels of manual/visual (MV) occupation. Thus an understanding of levels of occupation becomes important in designing and evaluating UPIDs (and PIDs in general). The major levels of MV occupation are:

Hands-free: User moves about freely. E.g., conversing or listening to music.

Momentary wrist or palm sized device usage: User momentarily stops work on other tasks to use the control. E.g., initiating or receiving a phone call, checking the time, changing the television channel.

Continued palm sized device use: User sits or stand while holding and operating the device with one hand. E.g., continuous television channel surfing.

Continued hand sized device use: User sits or stands while holding and operating the device with both hands. E.g., PDA text messaging, hand-held game playing.

Continued laptop or desktop device use: User sits while operating the device with both hands and most of their visual attention. E.g., word processing.

Encumbrance is unnecessary or undesirable occupation. E.g., mobile phone handsets cause manual encumbrance because they occupy the user’s hand for a task (conversation) that does not by itself require hand-use.

Minimizing encumbrance is one of Mopad’s primary design goals. Virtually every aspect of Mopad (modularization, wearability, mount variations and remote control variations) acts to meet this goal.

Surprisingly, minimizing encumbrance, does not seem to be one of the goals of other UPID visions. Wearable computers are infamous for their encumbrance: hands (keyboards and mouse), eyes (heads-up visual display) and the body as a whole (multiple, bulky components worn on belts, in backpacks or in vests). Even today’s smart phones (junior versions of handheld UPIDs) encumber the user with excessively large displays and control pads, while not providing hands-free operation.

User Modes

A user mode is an information focussed physical activity. User modes differs substantially from each other in the level of MV occupation that they allow or encourage, and thus limit the form factors of personal information devices that can be used, and that are optimally used.

The major user modes are:

Sitting at Desk: User is sitting at a dedicated work area. Typified by large MV occupation. Desktop or laptop computer devices are allowed and preferred.

Sitting at Television: Typified by light manual occupation and large visual occupation. While it seems that large manual occupation would be allowed, in practice users seem not to like this. (Hence the failure of web surfing on televisions.) This user mode is sometimes called “lean back” to distinguish it from the “lean forward” of Sitting at Computer.

Sitting at Large: User is sitting, but without a computer or television handy. Typified by moderate MV occupation (e.g., a PDA type device).

Moving Hands-Free: User is moving about, but has their hands free. E.g., user is walking. Light MV occupation is acceptable, though visual occupation is likely to slow movement (in order to keep the screen steady enough to read). Off-person devices are usually not available.

Moving Hands-Busy: User is moving about with their hands occupied. E.g., user is technician moving about a shop, or a housewife engaged in grocery shopping. MV occupation must be kept low. Off-person devices are usually not available.

Moving Undressed: User is nude is dressed lightly (e.g. in pajamas). MV occupation may be light to moderate and must be provided entirely through off-person devices.

Driving: User is sitting can view moderately size displays and operate moderately complex controls, but must keep their hands on the wheel and eyes on the road as much as possible. MV occupation must be kept low: or more specifically, moderate in amount, but short in duration. Off-person devices can be used.

Running: User’s hands are free, but unusable, so this mode is equivalent to Moving Hands-Busy.

Sleeping: MV occupation must be kept light to moderate and provided entirely through nearby, off-person devices (e.g., clock-radio or phone by the bed).

Synthesis

With a full understanding of availability, occupation and user modes, several important conclusions can be drawn:

PIDs are characterized by the amount of MV occupation that they permit. Pure audio devices: zero; wrist devices: light; handsets: light to moderate; PDAs: moderate; and laptops and desktops: large.
Applications vary substantially in the amount of of MV occupation that they demand. Audio applications (telephony, radio, music playing) demand little MV occupation, while word processing demands large MV occupation.
Many applications are not only unworkable, but also undesirable for certain user modes. E.g., users aren’t interested in engaging in word processing or television watching while running or driving.
Devices that are optimized for one user mode are usually either unusable or else poorly usable in other user modes. E.g., a desk phone works well at a desk, but not at all when the user is away from the desk. Similarly, laptops and car stereos are unusable away from their natural user mode. Another important example of this rule is wearable devices, which are typically ill-suited to user-modes which don’t allow them to be worn (e.g., while the user is sleeping).
Conversely, devices that work across user modes generally do not provide optimum performance in any one user mode – e.g., mobile phone handsets.
Applications that can be used at different levels of MV occupation, usually have different optimal implementations for different user modes. E.g., while walking, a music player is best implemented as a wearable personal music player, but while in the living room, it is best implemented as a stereo with a remote control. E.g., while sitting at a computer, an address book is best implemented as full sized desktop application, but while walking, it is best implemented as a scrollable tree-type menu with relatively little information being presented.
Applications and application implementations may be divided into groups which correspond to their varying demands for MV occupation. These groups correspond to the different types of control/display devices that they most naturally run on.
Audio applications are very usable at all levels of MV occupation. Moreover, an audio device that is designed for minimal MV occupation user modes is still at near optimum performance for maximal MV occupation user modes.
While many applications vary in desirability with user mode, audio applications, especially telephony, are strongly desirable across all user modes. Moreover, audio applications are desirable across the whole population, while advanced information applications (e.g., expense tracking, word processing) are of interest only to the smaller population of information workers.
Audio applications integrate very naturally because audio tends to be a single-tasking function and because audio input devices tend to block each other. This is in contrast to video applications which can be run simultaneously without substantially blocking each other.

With this background information and analysis, further conclusions about the design of a UPID can be drawn:

It is impossible to build a single device that meets the diverse MV occupation requirements of all the applications that a user wishes to run. MV devices are by their nature fixed in size, and at any given size are MV encumbering for applications that run at smaller MV occupation levels, and too small for applications that run at larger MV occupation levels.
The only applications that virtually demand a single consistent physical device are user alerting communication applications: e.g., telephony, paging, text/instant messaging. All of these are either zero to light MV occupation applications or can be routed through zero to light MV occupation devices.
Aside from these communication applications, the application that most benefits from omnipresence to the user is synchronization. E.g., while the user does not need to have a web browser with them at all times, if they move between different web browsers, then the most useful thing to take between those browser is their favorites list. Similar desires exist for basic personal information applications: address book, calendar, etc.. Since synchronization is purely a data application, it has zero MV occupation requirements, and so has the same form factor requirements as audio applications.
So, some applications (telephony, paging, text/instant messaging, synchronization) demand a single device, while applications in general, and form factor limits dictate that multiple devices be used. Ideally, this paradox would be resolved by a single device which morphed into different form factors for different user modes. While true morphing is physically impossible, it can be closely emulated through a modular system in which a core module is connected to and disconnected from appropriate peripheral modules so as to achieve the optimum configuration for each user mode.
Surprisingly, the modular system described above is buildable now. While UPID visions are generally perceived as requiring advanced miniaturization technology, the usability considerations discussed above dictate form factors that are achievable with today’s semiconductor and wireless communication technologies.

Haphone Mopad

The analysis presented here, focussing on: availability, occupation and user modes, contradicts current UPID visions. In their place, the analysis promotes a modular system, with an audio application and synchronization focussed core, supporting wearable and non-wearable configurations, and multiple control/display modules each suited to different user modes.

Haphone Mopad is designed to meet this new vision, while also factoring in considerations of wearability and stylization. While the Mopad system is expected to improve in capability and performance with increasing semiconductor chip integration and miniaturization, Haphone Mopad is able achieve the desired and realizable goals of the UPID ideal today.