Kortum P. (ed.) HCI Beyond the GUI. Design for Haptic, Speech, Olfactory, and Other Nontraditional Interfaces

Подождите немного. Документ загружается.

8.3.4 Applications of the Interface to Accessibility

Olfactory interfaces can be an intuitive means to communicate danger to others.

For example, people notice leaks of inflammable (but invisible) gas by smell.

Usually, methyl-mercaptan, to which humans are highly sensitive, is blended

FIGURE

8.16

Generating an olfactory field in free space with two air cannons.

The vortex rings collapse in front of the user and the high-speed air flow is

reduced.

Source:

From Nakaizumi et al. (2006); courtesy of Joseph “Jofish” Kay.

(a) (b)

Tube

Shutters

Fresh air

Air intake valve

Scented air

Push

FIGURE

8.15

Scent Projector, with nose-tracking and scent-switching functions.

(a) System overview. (b) Scent-switching mechanism. The system was developed

at the Advanced Telecommunications Research Institute International.

8 Olfactory Interfaces

282

with methane or propane gas. The olfactory communication function is effec-

tive for both sighted and visually impaired people. If computer -control led olfac-

tory interfaces become common, a wide variety of physiologically intuitive

signals can be provided, especially for communicating dangerous situations.

8.4

HUMAN FACTORS DESIGN OF THE INTERFACE

Although humans are said to be able to distinguish thousands of odors, there is

currently no method, as mentioned previously, to synthesize arbitrary smells by

mixing a small set of “primary” odors. In this section, various aspects of olfaction

are described.

8.4.1 Odor Intensity

The subjective intensity of a smell is related to the concentration of odor mole-

cules, but there is a wide variety in subjective intensity for various odor materials

at identical concentrations. When multiple odor molecules coexist, odors at higher

concentrations are not always subjectively dominant. Physical aspects that are

related to subjective odor intensity follow:

Concentration:

The most basic value, described in grams, molar concentration,

percentages, or parts per million (ppm).

Diffusivity:

This parameter refers to how fast the odor diffuses into the atmosphere,

and is defined as the distance that the odor molecule can reach within a specified

time or the time required for the odor molecule to travel a certain distance. The

higher the concentration and the lower the molecular weight, the larger the dif-

fusivity obtained. However, air flow has a significant effect on odor diffusion

when compared to static diffusion of odor material.

Boiling point:

The temperature at which vapor pressure becomes equal to environ-

mental pressure. The boiling point for most odor compounds is within the range

of 150



to 300



C. Odors with a significantly low boiling point tend to have an

intense smell, and odors with a higher boiling point have a weaker smell.

Volatility:

The amount of odor material that is consumed in a given time period.

Generally, Weber-Fechner’s law can be applied to the relationship between the

concentration of odor material and subjective intensity.

8.4.2 Threshold Value

The minimal concentration at which humans can “detect” a smell is called the

odor threshold value. There are multiple definitions of threshold value: detection,

recognition, and differential.

8.4 Human Factors Design of the Interface

283

Detection threshold value:

Smells cannot be detected at very low odor concentra-

tions. If the concentration is gradually increased, subjects can detect “some kind

of smell,” even though they cannot distinguish the type of smell. This concen-

tration is called the detection threshold value.

Recognition threshold value:

By increasing the odor concentration beyond the

detection threshold, the quality or impression of the smell at a certain con-

centration can be described. The minimal concentration at which the quality

or impression of a smell can be described is called the recognition threshold

value.

Differential threshold value:

The minimum difference in concentration that can

be detected is called the differential threshold value. Usually, it is described in

percentages as the ratio of the minimal detectable difference to the original

intensity (Weber ratio). This ratio varies depending on the type of odor. The dif-

ferential threshold value generally ranges from 10 to 30 percent for olfaction, as

opposed to 1 to 2 percent for differences in light intensity.

8.4.3 Adaptation

After continuous exposure to an odor, the subjective intensity of the odor gradu-

ally decreases; after a certain period (typically several minutes), the odor is no

longer consciously detected. This phenomenon is called adaptation, and is consid-

ered to be a mechanism of the sensing organ for automatically adjusting sensitiv-

ity to an odor. Through adaptation, smells that continuously exist, such as body

odor, are not detected, while odors that occasionally exist are clearly noticed.

The level and the temporal aspect of adaptation vary depending on the odor mate-

rial. The adaptation caused by a single odor material is called

self-adaptation

. The

recovery phase is generally faster than the adaptation phase.

Cross-adaptation

is the phenomenon of decreased sensitivity of one odor

through exposure to another. This adaptation occurs for some combination of

odors but not others.

Another mechanism is called

habitation

, which is often used in a similar con-

text to adaptation. Habitation is considered to occur within the brain, while adap-

tation is a function of the olfactory cells. However, the definition of habitation

appears to vary among researchers, and its definition has not been standardized.

8.4.4 Multiplier Effect and Masking Effect

When two different odors are mixed, the intensity is enhanced in some cases and

suppressed in others. Enhancement is called the multiplier effect and suppression

is called the masking effect. The masking effect in particular is used in daily life.

When an aroma is introduced into a space where an unpleasant smell exists, the

8 Olfactory Interfaces

284

unpleasant smell is perceived to decrease. This is a subjective phenomenon known

as sensory deodorization, as the unpleasant smell is not actually removed. Typical

examples of sensory deodorization are perfumes and deodorants in lavatories.

8.4.5 Nonlinearity

Changes in concentration sometimes result in qualitative rather than quantitative

changes, especially for very low concentrations. For example, a type of fecal odor

(indole) changes to a flowery smell at very low concentrations. In addition, a mix-

ture of smells sometimes results in a qualitative change of smell. This phenome-

non is called

modification

. Usually, surrounding odors consist of multiple

components, and these odors are modified. The blending of aromas is considered

to be a sophisticated technique in creating desired aromas via modification.

8.4.6 Spatial Perception

The question of how humans perceive the surrounding environment through

olfaction is an interesting topic for interactive human–computer interfaces, espe-

cially in virtual reality. Finding the location of a smell source is an important task

related to spatial perception by olfaction. By moving and comparing the trajectory

and the temporal profile of the intensity, the smell source can be located (Yamada

et al., 2006). This might be considered to be a counterpart to the motion parallax

cue in v isual perception. Another interesting question is whether t here are

olfactory counterparts to binocular cues through the use of two nostrils.

In his historic research, Be

sy (1964) reported that the time difference of the

olfactory stimuli for each nostril contributes to the directional localization of the

smell source, just as the direction of auditory sources can be determined by the dif-

ference in sounds perceived by each ear. More recently, Porter et al. (2007)

reported that scent tracking is aided by internostril comparisons. Considering that

the internostril distance is only 2 to 3 cm, which is less than the interpupillary (6

to 7 cm) and interaural (approximately 15 cm) distances, the “stereo effect” in olfac-

tion may not be as strong as the “motion effect.” However, it is very interesting to

note that humans make use of two nostrils to help understand the environment

by olfaction.

8.5

INTERFACE-TESTING TECHNIQUES

Historically, the evaluation of odor has long relied on sensory analysis by experi-

enced human testers. This means that it was difficult to evaluate a smell objec-

tively. Even though experienced testers were trained to provide objective results

rather than subjective impressions, these results are assumed to be affected by

environmental conditions, the tester’s physical condition, and so on.

8.5 Interface-Testing Techniques

285

However, recent progress in chemical sensors has resolved this problem. The

technical field related to smell-sensing machinery is called electronic nose or e-nose

(Pearce et al., 2003). The electronic nose is a system to identify the components

of odors. Multiple gas sensors (chemosensors) are used in the system to detect, ana-

lyze, and categorize multiple component odors. Among the various types of chemo-

sensors, metal oxide semiconductor sensors, conducting polymer sensors, and

quartz microbalance sensors have been frequently used for electronic nose

systems.

Metaloxide gas sensors have been used to detect flammable gases such as

methane and propane or poisonous gas such as hydrogen sulfide, and are

now used as odor-sensing elements as well. When odor molecules are absorbed

onto the surface of a semiconductor, the electric conductivity of the element is

changed according to the amount of absorbed molecules. Conducting polymer

sensors also use the principle of electric conductivity variation, while quartz

microbalance sensors have a resonance frequency output that varies when odor

molecules are absorbed in the sensitive membrane on the surface of a quartz

resonator.

One problem with odor-sensing equipment is that the temporal response is

still insufficient for interactive use. The response time for a typical gas sensor is

dozens of seconds. However, some recent sensors have improved response times;

if the response of gas sensors in particular becomes faster than human olfact-

ion, these sensors could be incorporated into the interaction loop of an olfactory

display system.

8.6

DESIGN GUIDELINES

Although the development of computer-controlled olfactory interfaces is still at an

early stage, some guidelines can be indicated when designing and implementing

them.

8.6.1 Expected Effect of Appending

an Olfactory Interface

First, the designer should take into account whether it is meaningful or effective

for the entire system to append an olfactory channel. If the olfactory interface is

inherent to the designer’s intention, such as creating odor as the main content

of the interface, this decision is reasonable. However, there may be issues when

the olfactory interface is not the primary mechanism of interaction. In this case,

designers should decide whether the olfactory effect is worth the additional cost

and complexity to the system.

8 Olfactory Interfaces

286

8.6.2 Number of Odors

As mentioned previously, there is no method for synthesizing arbitrary smells

from a small set of component odors. Hence, determining the number of odors

required for the application is important. Even though humans can distinguish

thousands of smells, the number of smells required for a single application may

be relatively small. For example, if we want to emit smells according to the user’s

behavior in an interactive game, several to a few dozen odors may be sufficient. In

this case, it is sufficient to blend a relatively small number of smells or select pre-

blended aromas; blending hundreds or thousands of odors is not necessary. The

number of simultaneously required odors directly affects the complexity of the

system; thus, careful selection of the odors is essential for creating cost-effective

systems.

8.6.3 Number of Users and Spatial Dimensions

The next point to consider is the number of target users for simultaneous smell

deployment. If the application is for a single user, a device of personal size, such

as a small scent generator located close to the user, can be applied. If the applica-

tion is for many simultaneous users, odors should be delivered to a larger space.

There are two approaches to achieve multiuser olfactory interfaces: providing a

personal device for each user or controlling a large-volume olfactory space at

any one time. The effects of the volume of space on the temporal aspects of odor

are discussed in the next subsection.

8.6.4 Temporal Aspects of Odor Control

When applying olfactory interfaces with audiovisual programs or interactive appli-

cations, the speed of changes or replacements in olfactory stimuli is an important

issue. If the odor is to be used as background or ambient media, the time constant

is relatively large with constant or slow/gradual changes in odor. This is an effec-

tive way of using odor to provide information in a subtle manner.

However, the u se of olfactory interfaces together with audiovisual programs

or applications wi th rapid interaction requires short-term changing or switching

of odors. In this case, designers face the problem that emitted smells cannot be

erased instantaneously. Once an odor is diffused in the a ir, it remains unless it

ismovedbyaircurrents.Possiblesolutionsare(1)designtheentirefacilityto

include ventilation systems and (2) emit as small an amount of scent as

possible.

The first approach is effective when the olfactory interfaces are used in dedi-

cated, specially designed facilities, such as amusement parks. The olfactory inter-

faces used at such locations can simply emit a large amount of odor materials,

8.6 Design Guidelines

287

which ensures the olfactory experience. The emitted odors are then removed by

the ventilation system and fresh air or another odor is provided. This approach

can also be applied to a closed-booth configuration. An example of this type is

the interactive smell system shown in Figure 8.17, which makes use of a slow,

undetectable air flow to produce spatiotemporal patterns of smell in a rectangular

parallel-piped booth.

If one cannot make dedicated facilities, such as for home use, then the sec-

ond approach would be effective. Because the amount of the emitted smell is

small, the odor emission device should be placed close to the nose. Although

the emitted smell is diffused into the surrounding atmosphere, the amount is

sufficiently small that the concentration quickly diffuses below the detection

threshold and the smell is unnoticed. There are several approaches of this type,

depending on the locati on of the odor-emitting dev ice. Nose-mounted and face-

mounted types are possible solutions (Yamada et al., 2006). Odor emission

points can include other part s of the body, such as the chest (Morie et al.,

2003) or arm (Mochizuki et al., 2004). For users who do not want to wear

devices, a local scent delivery technique through free space can be adopted

(Watkins, 2002; Yanagida et al ., 2004).

Sweat

High-pressure chamber

Fan array intake

Diffusion chambers

Interaction zone

Air movement

0.2 M/S

Low-pressure chamber

Extract duct

Smell wall

Coffee shop

Laundromat

Tires

Petrol

Fresh air

Rubbish

Cut

rass

FIGURE

8.17

Scents of Space.

An interactive smell system.

Source:

From Haque (2004), Haque Design and

Research/Josephine Pletts.

8 Olfactory Interfaces

288

8.7

CASE STUDIES

Two case studies exploring olfactory interfaces can be found at

www.beyond-

thegui.com

8.8

FUTURE TRENDS

As mentioned previously, olfactory interfaces are still in an early stage of develop-

ment compared to visual, auditory, and haptic interfaces. In recent years,

however, research papers and/or demonstrations have been presented in almost

every academic conference related to human–computer interaction, including

the field of virtual reality. This implies that many researchers in various countries

are now interested in olfactory interfaces and that researchers in this field are cur-

rently networking to establish a research community.

The future of olfactory interfaces depends on at least two factors (Wasburn &

Jones, 2004): the development of practical applications to make olfactory interfaces

more common, and fundamental improvements in olfactory interfaces, including

the elucidation of human olfaction at both the physiological and psychological levels.

From a technical viewpoint, it is very important to know how many types of odors are

necessary for a certain application and for producing arbitrary smells. If this number

is less than a hundred, constructing practical interfaces will not be difficult.

REFERENCES

Amoore, J. E. (1970).

Molecular Basis of Odor

. Springfield, IL: Charles C. Thomas.

Barfield, E., & Danas, E. (1996). Comments on the use of olfactory displays for virtual envir-

onments.

Presence

5(1):109–21.

sy, G. (1964). Olfactory analogue to directional hearing.

Journal of Applied Physiology

19(3):363–73.

Buck L., & Axel, R. (1991). A novel multigene family may encode odorant receptors: A

molecular basis for odor recognition.

Cell

65:175–87.

Cater, J. P. (1992). The nose have it!

Presence

1(4):493–94.

Cater, J. P. (1994). Smell/taste: Odors in virtual reality.

Proceedings of IEEE International

Conference on Systems, Man and Cybernetics

, San Antonio, TX.

Davide, F., Holmberg, M., & Lundstro

m, I. (2001). Virtual olfactory interfaces: Electronic

noses and olfactory displays. In Riva, G., & Davide, F., eds.,

Communications Through

Virtual Technologies: Identity, Community and Technology in the Communication Age

Amsterdam: IOS Press, 193–219.

Dinh, H. Q., Walker, N., Song, C., Kobayashi, A., & Hodges, L. (1999). Evaluating the import-

ance of multi-sensory input on memory and the sense of presence in virtual environments.

Proceedings of IEEE Virtual Reality

, Houston.

8.8 Future Trends

289

Haque, U. (2004). Scents of Space: An interactive smell system.

Proceedings of ACM SIG-

GRAPH 2004 Sketches

, Los Angeles.

Heilig, M. L. (1962). Sensorama Simulator, U.S. Patent 3,050,870.

Heilig, M. L. (1992). El cine del futuro: The cinema of the future.

Presence

1(3):279–94.

Kaye, J. N. (2001). Symbolic Olfactory Display. Master’s thesis, Massachusetts Institute of

Technology.

Kaye, J. N. (2004). Making scents.

ACM Interactions

11(1):49–61.

Lorig, T. S., Elmes, D. G., Zald, D. H., & Pardo, J. V. (1999). A computer-controlled olfac-

tometer for fMRI and electrophysiological studies of olfaction.

Behavior Research Meth-

ods, Instruments & Computers

31(2):370–75.

McCarthy, R. E. (1984). Scent-Emitting Systems, U. S. Patent 4,603,030.

Mochizuki, A, Amada, T., Sawa, S., Takeda, T., Motoyashiki, S., Kohyama, K., Imura, M., &

Chihara, K. (2004). Fragra: A visual-olfactory VR game.

Proceedings of ACM SIGGRAPH

2004 Sketches

, Los Angeles.

Morie, J. F., Iyer, K., Valanejad, K., Sadek, R., Miraglia, D., Milam, D., Williams, J., Luigi,

D–P., & Leshin, J. (2003). Sensory design for virtual environments.

Proceedings of ACM

SIGGRAPH 2003 Sketches

, San Diego.

Nakaizumi, F, Yanagida, Y, Noma, H., & Hosaka, K. (2006). SpotScents: A novel method of

natural scent delivery using multiple scent projectors.

Proceedings of IEEE Virtual Reality

2006

, Arlington, VA, 207–12.

Nakamoto, T., & Hiramatsu, H. (2002). Study of odor recorder for dynamical change of odor

using QCM sensors and neural network.

Sensors and Actuators B

85(3):263–69.

Nakamoto, T., & Minh, H.P.D. (2007). Improvement of olfactory display using solenoid

valves.

Proceedings of IEEE Virtual Reality

, Charlotte, NC.

Ooyama, A. (1994).

Handbook of Sensation and Perception

. Tokyo: Seishin Shobor.

Pearce, T. C., Schiffman, S. S., Nagle, H. T., & Gardner, J. W., eds. (2003).

Handbook of

Machine Olfaction

. Weinheim: Wiley-VCH.

Plattig, K. H. (1995).

Spu

rnasen unt Feinschmecker

. Berlin: Springer-Verlag.

Porter, J., Craven, B., Khan R. M., Chang S. J., Kang, I., Judkewitz, B., Volpe, J., Settles, G.,

& Sobel, N. (2007). Mechanism of scent-tracking in humans.

Nature Neuroscience

10(1):27–29.

Tonoike, M., ed. (2007). [

Information and Communication Technology of Olfaction

] (in Jap-

anese). Tokyo: Fragrance Journal Ltd.

TriSenx. (2007). Home page—available at

http://www.trisenx.com/.

Yamada, T., Yokoyama, S., Tanikawa, T., Hirota, K., & Hirose, M. (2006). Wearable olfac-

tory display: Using odor in outdoor environments.

Proceedings of IEEE Virtual Reality

2006

, Charlotte, NC, 199–206.

Yamanaka, T., Matsumoto, R., & Nakamoto, T. (2002). Study of odor blender using solenoid

valves controlled by delta–sigma modulation method.

Sensors and Actuators B

87:457–63.

Yanagida, Y., Kawato, S., Noma, H., Tomono, A., & Tetsutani, N. (2004). Projection-based

olfactory display with nose tracking.

Proceedings of IEEE Virtual Reality 2004

, Chicago.

Yatagai, M., ed. (2005).

Kaori no hyakka-jiten

[Encyclopedia of aromas]. Tokyo: Maruzen.

Washburn D. A., & Jones, L. M. (2004). Could olfactory displays improve data visualization?

Computing in Science & Engineering

6(6):80–83.

Watkins, C. J. (2002). Methods and apparatus for localized delivery of scented aerosols, U.S.

Patent 6,357,726.

8 Olfactory Interfaces

290

CHAPTER

Taste Interfaces

Hiroo Iwata

9.1

NATURE OF THE INTERFACE

Taste is an important sense that is seldom used in traditional displays. It is the

fundamental nature of the underlying sense and the difficulty of implementing

unobtrusive devices that have made taste interfaces so difficult.

9.1.1 The Sense of Taste

Humans detect chemical features of food with taste receptor cells. They are assem-

bled into taste buds, which are distributed across the tongue. Although humans can

taste a vast array of chemical entities, they evoke few distinct taste sensations:

sweet, bitter, sour, salty, and “umami.” Distinct taste receptor cells detect each of

the five basic tastes (Chandrashekar et al., 2006). Examples of these basic tastes

follow:

Sweet:

A common sweet substance is sucrose, known as table sugar.

Bitter:

A strong bitter taste prevents ingestion of toxic compounds. A small amount

of a bitter substance contributes to perceived interesting or good taste/flavor.

Caffeine in coffee is an example.

Sour:

Sour substances are mostly acids. Vinegar is an example.

Salty:

A common salty substance is NaCl, known as table salt.

Umami:

This is a Japanese word meaning “savory,” and thus applies to the sensa-

tion of savoriness. Typical umami substances are glutamates, which are espe-

cially common in meat, cheese, and other protein-rich foods (Kawamura &

Kare, 1987).