My TV Stinks!

smellovision

As an interaction designer, I spend much of my day observing people and trying to figure out how and why they use technology. About a year ago, I participated in a group activity that involved identifying ten substances by smell alone.

Even though the smells were extremely strong and distinct (mint, shoe polish, truffle oil, lemon), none of us successfully identified

more than half of them. We did, however, spend almost an hour talking excitedly about the memories certain smells evoked and how frustrating it was to recognize a smell but not be able to put a name to it.

For me, it was a revelation. I realized that in years of designing hardware and software interfaces, I’d never once considered smell, even though without any technological assistance it was powerful enough to make a roomful of adults squeal like excited children. Thus began my obsession with designing compelling smell interactions, an obsession that eventually led to the creation of Scratch ’n’ Sniff, a scented digital interface.

My project brings the scratch-and-sniff sticker into the digital age. When a user scratches a display monitor, it briefly smells of whatever image the user scratched, often to the user’s incredulous delight (visit vimeo.com/15881329 for a movie of people’s reactions at Maker Faire New York).

How It Works

Scratch ’n’ Sniff relies on fluid dynamics and some clever sleight-of-nose. Five hacked electric air fresheners loaded with custom scents are hidden just below a computer touch-screen monitor.

When the user rubs his finger on the screen, across an image of a grapefruit for instance, a Processing sketch communicates the coordinates to an Arduino, which fires the appropriate air freshener.

A fine mist of grapefruit-scented particles flows mostly along the surface of the screen, thanks to a property of fluid dynamics called the no-slip condition, which makes rigid surfaces especially “sticky” to fluids such as the grapefruit aerosol (there’s a great video demonstration of this phenomenon at makezine.com/go/noslip). Diffusion into the surrounding air dissipates the smell within 5 seconds. Presto! A scratch-and-sniff TV!

There were three main design challenges:

1. Designing the interaction. I chose to mimic the workings of the scratch-and-sniff sticker not only because it’s familiar to almost everyone, but also because it takes advantage of the nature of smell: the images visually reinforce people’s scent perception, while scratching the screen provides a clear rationale for the strength of the smell (the longer the scratch, the stronger the smell).

2. Emitting the smell. All manner of contraptions exist for emitting and dispersing smell, from scented candles and heated oils to complex industrial vaporizers. I used Glade Wisp Flameless Candle air fresheners because they’re cheap, readily available, and easily hacked to produce an almost invisible spray. (See MAKE Volume 16, page 161 for instructions on how to hack similar fresheners.)

3. Choosing scents. Light can be broken down into three primary colors, and all sound can be analyzed into waves of distinct frequencies, but scientists can’t seem to agree on a basic unit of smell. They’ve identified more than 700 chemical compounds that are the basis for thousands of smells we can detect. Rather than trying to combine scents, I chose five distinct and easily identifiable smells. There are companies devoted to producing scents of everything you can possibly imagine (see Resources).

scent dispensers
An Arduino Diecimila priming six scent dispensers (not for the faint of nose).
tools
Tools of the trade: scent canister, Glade Flameless Candle, and Demeter Fragrance Hershey’s chocolate scent.
scents
Under the hood: five scent dispensers mounted just beneath and parallel to a touchscreen.

Standing in the Nostrils of Giants

I am not the first person to become obsessed with smelly media. People have been creating scent-emitting devices for years. A quick search of the U.S. Patent Office’s database turns up all sorts of odoriferous inventions. There’s the “olfactory special effects system,” the “multimedia and scent storage medium and playback apparatus,” the “combined scent and audio point of sale display unit,” and Hans Laube’s original 1954 patent for Smell-O-Vision: “motion pictures with synchronized odor emission.”

The idea of augmenting cinematic experiences with smells is nearly as old as cinema itself. A mischievous projectionist is said to have used an electric fan to disperse a rose scent during newsreel footage of the Rose Parade in 1906. Fifty years later, Smell-O-Vision and rival AromaRama battled to win over the noses of American moviegoers, an ultimately futile episode that inspired filmmaker John Waters to create his tongue-in-cheek Odorama scratch-and-sniff cards. These were distributed at screenings of his 1981 comedy Polyester, where viewers were invited to enhance selected scenes in the movie with famously unpleasant scents, including glue, grass, and feces.

The dawn of the internet age heralded the arrival of digital scent technology. A 1999 Wired article profiled DigiScents, the company behind the iSmell scent generator. The unfortunately named device’s creators envisioned a fragrant future in which scent-emitting hardware attached to computers and other media devices would allow content creators and advertisers to add an olfactory dimension to the web. The company sank in 2001 when the internet bubble burst, but the idea persists. It resurfaced most recently in Japan as the i-Aroma, a USB device that emits astrologically appropriate aromatherapy.

Not all smell-related innovations have failed, however. Scratch-and-sniff stickers remain as popular today as they were 30 years ago, and it’s hard to open a glossy fashion magazine that doesn’t contain at least one perfume-drenched scent strip. If you’ve walked around a supermarket recently, you’ve probably noticed the growing number of air freshening products, from low-tech spray odor maskers such as Febreze to motion-sensing electronic devices that dispense scent whenever you enter a room.

Many hotels, casinos, and retailers have adopted “logoscents,” smells intended to enhance the emotional effect of their brands. A whole industry has arisen around the idea of ambient scent marketing. In 2005, ScentAir, a leading “scent delivery solutions” provider, partnered with long-time elevator music

company Muzak — soon you’ll find yourself wondering not only why you’re “Singin’ in the Rain” but also why you smell like it!

Designing Smelltech

arduino
Arduino microcontroller drives five Glade Flameless Candles (transistor version).

What distinguishes smell technologies that succeed from those that fail is thoughtful design — design that takes into account both the science of smells and how people interact with them. Before we consider what makes smelltech compelling, let’s first consider how smell differs from our other senses.

Take, for instance, the new iPhone, a prodigious piece of multisensory interactive technology. It has a high-resolution display that delights our eyes with crisp text and sharp images, sound output that floods our ears with the full dynamic range of flawlessly reproduced audio, and an interface that engages our sense of touch in increasingly novel ways. It’s not edible, so Apple’s engineers haven’t spent too much time worrying about how it tastes. But why have they entirely ignored our noses?

Try for a second to imagine how the message “See you tomorrow at 4” smells, and it becomes clear that smell is not a particularly effective method of communication. Smell moves through the air slowly and unpredictably, and unlike sight, sound, and touch, it can’t easily be turned on and off.

Because scents are invisible and diffuse, it’s hard for people to agree on them in the absence of visual stimuli. While you and I might agree that the orange I’m holding in my hand smells like an orange, if instead I were holding a glass of green liquid that smelled the same, we may no longer agree that orange is the predominant scent. Smell expert Avery Gilbert is quick to point out that “wide person-to-person variability is a hallmark of odor perception.”

In other words, smell is more subjective than the other senses (it is responsible for the majority of our sense of taste, which is

why tuna may taste horrible to one person and like heaven to the next). This also means that our sense of smell is more susceptible to suggestion.

A successful smell interaction will take these peculiarities into account. Here’s a list of further considerations: Give people a clear indication of when they should smell something. Don’t expect them to react to ambient smells. Smell takes time to diffuse and disappear. Make sure you account for this time in your design.

Choose distinct smells. Everyone knows Play-Doh, but few will identify marjoram. Test your setup in a variety of conditions. Scents behave differently in large, crowded spaces than they do in small, empty rooms. Don’t overwhelm users with smell. Olfactory fatigue sets in quickly in the presence of overpowering smells. Be prepared for some people to not “get” the smell. Remember, smell is subjective, so have at least a couple of different types of smell on hand.

Resources

Demeter Fragrance Library: demeterfragrance.com

Novelty fragrance oils: saveonscents.com

DIY Home Perfumery: MAKE Volume 22, page 135

A listing of hundreds of chemical compounds that make up the smells we recognize: flavornet.org

Avery Gilbert’s book What the Nose Knows (Crown Publishers, 2008) and his amazing smell blog: firstnerve.com

Smell marketers index: scentmarketing.org

Discuss this article with the rest of the community on our Discord server!
Tagged
Alex Kauffmann

Alex Kauffmann is a New York-based writer/artist/designer. His work, like his life, is beautiful, simple, and absurd. You can see a bit of both at chinaalbino.com

View more articles by Alex Kauffmann

ADVERTISEMENT

Maker Faire Bay Area 2023 - Mare Island, CA

Escape to an island of imagination + innovation as Maker Faire Bay Area returns for its 15th iteration!

Buy Tickets today! SAVE 15% and lock-in your preferred date(s).

FEEDBACK