More Baloney about AI

 

Some years ago, in What the Nose Knows, I wrote about e-noses and their value to us in the future. I thought it was a rather mixed bag: 

At some point in the development of these fusions of silicon and biology, the question becomes not whether the e-nose can replace the human nose, but whether we want it to. Would I let an e-nose sniff-scan me for lung cancer? Sure. Would I use a robotic odor sentinel? Maybe, especially if I had a B.O. problem. But do I really want my refrigerator to tell me, “I’m sorry, Avery, I can’t let you eat those cold cuts”?

(For full effect you have to imagine the fridge speaking in the voice of HAL, the computer in 2001: A Space Odyssey.)

I remain a fridge-primitivist: I recently bought one that not only isn’t “smart” but has no automatic ice-cube maker. Still, the dream lives on for some people including Ashok Prabhu Masilamani, founder of Canadian tech company Stratuscent which makes a chip that can detect various volatile molecules in the air. He’s quoted in today’s WSJ in an article by Benoit Morenne headlined “A new frontier of AI-enabled gadgets.”

E-noses could also be integrated in smart fridges to detect early signs of expiring food and guide users to items that will expire next, says Dr. Masilamani. Scientists at the Nanyang Technological University of Singapore have developed a colored bar code that reacts to gasses from decaying food and a bar code reader that uses AI models to predict food freshness, according to a study published in the Advanced Materials journal in October.

The dream never dies . . .

The WSJ piece also discusses AI-enabled toilets that can analyze your stool sample at the time of delivery. Again, imagine the AI toilet speaking with the voice of HAL:

“Avery, you need to cut back on the kimchi. I have made you an appointment with a gastroenterologist.”

On second thought it would be much better if the AI spoke with the accent of an 18^th Century royal physician in England:

“Good news. A fetid and a stinking stool.”
[Assembled colleagues nod in agreement.]
“The colour good, well shaped, and a prodigious quantity.”

A Gloomy Day on Social Media

 

It snowed a couple of inches yesterday across the Front Range of the Rockies and this morning the lilacs and the aspen saplings at FirstNerve Manor were bent under the icy weight. It’s been dark, overcast, and raining all day long. One stares out the window waiting for something better to happen. Then, just when one begins to reorganize the pencil drawer comes a flash of insight—what a perfect afternoon to weed out the LinkedIn connections!

Some calls are easy—those people you interacted with years ago but whose careers have gone off in entirely different directions from your own. It’s unlikely you will ever have a professional interest in common with them again. So—“Remove Connection.”

Then there are the connections that bring you up short—the old friend who died unexpectedly nearly a year ago. It seems callous to delete the link, but even a candle lit in his memory would eventually burn out. (“Remove Connection.”) Then the likeable guy you pitched some business with before he dropped dead after a squash game a decade ago. You went to his funeral and memorialized him here. Does a persistent online link express anything further? (“Remove Connection.”) Ditto the former client’s late husband who you liked and respected.

Even more depressing is to find the connections who, once in senior positions, are now retired. One thinks, uncharitably, that they are now just taking up virtual space in your LinkedIn connections. (“Remove Connection,” “Remove Connection,” etc.)

The same goes for academics who have levitated themselves into “emeritus” status. (“Remove Connection.”) And then, a long-ignored discontent bubbles up—why link to any academics at all? Professors are easily found via their campus web pages. Basta. (Three more “Remove Connections.”)

The late afternoon sun has broken through and the birds are singing once again. The virtual thicket has been thinned out and is ready to welcome new tendrils of professional linkage. Time for a drink.

The Shape of Smells to Come

There is a certain charm to the stiffly drawn illustrations found in United States patent documents. The clear, static, highly conventionalized style evokes comic books and instruction manuals. Yet, in contrast to the ponderous, formal, lawyerly text of the patent, the hand of the artist can give the images a certain style of their own.

Here is Figure 1 from U.S. Patent 3,902,851, “Method of detecting odors and apparatus therefore,” by Andrew Dravnieks. It was granted on September 2, 1975. The illustrator did not sign his work, but I think it qualifies as a type of olfactory art.

Scent of Mystery Comes to Video

The New York Post’s Lou Lumenick has the story: “Elizabeth Taylor’s forgotten Smell-O-Vision adventure arrives on video.” He’s talking about Scent of Mystery, the first (and only) film to use the revolutionary Smell-O-Vision process promoted by Mike Todd, Jr. and his father. Elizabeth Taylor (then married to Todd, Sr.) was an investor in the technology and makes a cameo appearance in the movie. (Your Hollywood trivia of the day: Mike Todd, Sr. coined the word cameo.)

I researched the story of Smell-O-Vision—and its rival, AromaRama—for my book. Along the way I interviewed inventor Hans Laube’s wife and daughter, and saw a prototype of his scent-generating device, along with all the promotional material for the film’s premiere. It’s a great story in the history of technology.

Until now, Scent of Mystery was only available in a badly butchered VHS of a television mashup version. I’m looking forward to seeing the newly restored re-release. Sure, it was slightly corny even for 1960, and unspools at the leisurely pace of that time. But I look forward to viewing it end-to-end while imagining all the odors (there are lots of them).

A Late Sixties Version of The Feelies

Morton Heilig is considered by some to be one of the technological pioneers of virtual reality. Jon Turi wrote yesterday about Heilig’s 1962 3-D Sensorama simulator in Engadget’s Time Machine column. Turi mentions Heilig’s concept of a cinema-sized, multisensory Experience Theater that included a smell component. Given my fascination with the previous technologies of Smell-O-Vision and AromaRama, documented in What the Nose Knows, I took a look at Heilig’s U.S. Patent 3,469,837, issued on September 30, 1969. Heilig begins his description of the smell component with this observation:

Aroma systems without 3-D have been attempted in the past but they have always failed because of the psychological paradox of smelling a flat two-dimensional image. Odor implies a physical reality—a palpable presence—thus, the object providing the aroma should preferentially be three-dimensional, if a satisfactory natural impression is to be made on the spectator.

This is a questionable assertion, to say the least. I have never heard of the “psychological paradox” of the smelly flat image, and in all my research on the topic I have never seen it cited as a reason for the failure of Smell-O-Vision or AromaRama. I also disagree that smell always implies a palpable presence. Garlic on someone’s breath does not imply that garlic is still present and available for you to eat. The remembrance of someone long dead triggered by the smell of an old shirt does not imply his physical presence. And so on.

In the patent, Heilig describes two ways of delivering smell to a person seated in the Experience Theater. The first involved tubes that deliver a scented air stream to the person’s face. This is an invidualized version of Smell-O-Vision which delivered scents through tubes on the back of theater seats. Heilig’s concept also includes a suction/exhaust feature that carries the scent away from the viewer and out of the theater. Here’s what he says about this feature:

One of the serious problems with other attempts to add aromas to films has been the contamination of the theater air with one odor and the inability to clear the air of this aroma before the next one arrived.

Here Heilig overgeneralizes. Scent built-up was a problem for the shoddily designed AromaRama; it was not an issue for the more finely-tuned Smell-O-Vision system. People tend to exaggerate the residual odor problem because they overestimate how much scent is required to provide create a successful impression—it doesn’t really take a lot.

Heilig’s second idea for scent delivery is a cartridge that fits into the theater seat. It contains a dozen chambers with scent-saturated pellets; a solenoid system triggered moves the correct scent into a position where it intercepts the air stream to the view. Yes, it would have been kludgy; and imagine the clickety-clacking as a hundred solenoids simultaneously move their aroma cartridges into position

Still in all, I have to admit that Heilig wasn’t afraid to think big.

How Do I Smell, Kaori-chan?

There’s a lot about Japan that I really like. Among the common elements are a slightly off-kilter (from my perspective) aesthetic and an unbridled enthusiasm for technology and gadgetry.

The former has fascinated me since the badly dubbed sci-fi films I watched on TV on Saturday afternoons. (Who were those tiny women in the rowboat singing to Mothra?) And I fondly remember listening to baseball broadcasts of the San Francisco Giants on my Panasonic transistor radio—my first-ever technology purchase.

Old habits die hard, and since acquiring a Roku box I’ve been catching up on Japanese sci-fi movies. Judging by some of them, the scene has gotten even weirder. Don’t take my word for it, go watch the disturbing imagery in Meatball Machine (2005) and get back to me.

Or check out the claustrophic and bloody Hellevator (2004), written and directed by Hiroki Yamaguchi who after this effort does not appear to have been allowed near a camera for about five years. In fairness, the guy who played the cannibal-rapist convict was awesome. [And the elevator operator chick was hot.—Ed.] [Yes. Yes she was.]

After my late-night sci-fi viewing, this news story seemed almost charmingly droll:

On 24 August, Kanagawa Prefectural Police picked up Joji Kondo for stealing three seats from women’s electric bicycles in a housing complex at around 4:00 in the morning. After searching Kondo’s home they uncovered a further 200 seats.

According to police he admitted to the crimes saying “I wanted to smell the lingering scent of a woman.” MSN News Japan reported the 35-year-old as saying “I like the texture of the leather and the smell it has, I would lick it and sniff it.”

Alright, then.


Via Saga-S.co.jp

Meanwhile, the twin Japanese obsessions of olfactory technology and robotics have merged to produce a Sniff-Me Bot:

The female robot, Kaori-chan, has brown hair and blue eyes. When a person breathes in her face, she analyzes and quantifies components in their breath, and evaluates the smell on a scale of one to four.

I think I get it. Sort of.

UPDATE August 28, 2013
“Hey, wait a minute. Did that robot just fart?”

HuffPo’s Betsy Isaacson Plays Softball

How much effort does it take to be “a Technology writer” for the Huffington Post? Evidently not much. In a piece on Amy Radcliffe’s “smell camera,” Betsy Isaacson (a 2012 graduate of Harvard University) puts her by-line on a piece that quotes copiously from the Guardian and Wired.

Whatevs. But if you are going to re-package everyone else’s work, why not spice it up with a contrarian view (i.e., mine)? I wrinkled my nose at Radcliffe’s stunt ten days before Ms. Isaacson got around to posting on the HuffPo. She could have googled “Amy Radcliffe” and clicked to the second page. Too much effort, I guess.

Or is there some “Women + Technology” narrative that needs to be preserved here?

Exit question: when did skepticism cease to be a job requirement for journalists?

Call Off the Dogs

The space between the lost clinical art of olfactory diagnosis and Dr. McCoy’s tricorder of the future is filled with canine scent detection studies. These prove, ad nauseum, that a disease may have a distinctive scent signature. But to put this principle to work in a way that doesn’t involve teams of trained dogs we need a precise, chemically defined profile of each disease’s scent. Some new studies suggest we’re finally getting some traction on the problem.

My former Monell Center colleague George Preti, along with others at Monell and UPenn, has found a small set of volatile molecules that distinguish melanoma cells from normal melanocytes. The usual cautions apply—e.g., the volatiles were collected over cell cultures and not over the skin of actual patients—but the work is a significant step toward a device that can “smell” skin cancer. Preti et al. used sophisticated scent capture (solid-phase micro-extraction for GC-MS) supplemented by nanotubes coated with single-stranded DNA. Strictly speaking this isn’t an e-nose but an e-nose with nanosensors. The team then did a lot of heavy lifting to screen out volatiles not endogenous to the melanoma cells. (Earlier studies picked up traces of antiseptics and anesthetics from the hospital environs.)

The result: key differentiating compounds included isovaleric acid, 2-methylbutyric acid, and isoamyl alcohol, plus the sulfur containing compounds dimethylsulfone, dimethyldi- and trisulfide. None are exotic but the pattern allows discrimination of healthy and cancerous cells. Heightened production of these molecules is presumably to metabolic differences between the two cell types.

Meanwhile, researchers in the UK reported pilot data on detection of bladder cancer using a GC-sensor device that samples urine headspace. Using statistical algorithms the device correctly separates urine from cancer patients and that of healthy matched controls. Nice.

A leading member of the team is Professor Norman Ratcliffe, pictured below.



With necktie and fresh, unwrinkled lab coat he’s quite well dressed for an engineering professor. [Especially a British one!—Ed.] He appears to be holding a urine sample. Honestly, dude, glove up!

Prof. Ratcliffe presumably knows better than to bare-hand biofluids. He and his colleagues have previously done a lot of shitty work [Phrasing!—Ed.] using fecal volatiles to diagnose diarrhea, cholera, necrotizing enterocolitis, and irritable bowel syndrome.

From the Competing Interests statement of the PLoS ONE paper, it appears that team Ratcliffe has been granted one patent and has applied for another. To my inexpert eye, the granted patent seems rather broad—it covers the idea of a headspace collector and analyzer that can diagnose disease via odors. I don’t see how it would hold up in litigation, but then I’m not a patent attorney. [Or a patent troll.—Ed.] By publishing in PLoS ONE they certainly achieved plenty of free publicity for their patented device.

While human cancer studies make the headlines, olfactory diagnosis is advancing down on the farm. In Germany, a group of veterinarians and vet students was matched against an e-nose. The task: sniff post-partum vaginal discharge from a cow and judge whether or not the animal suffers from acute puerperal metritis. The humans performed adequately but they were completely dusted by the DiagNose device. The researchers note, however, that the e-nose is yet suitable for use in the field, er, barn.

The studies discussed here are “Volatile biomarkers from human melanoma cells,” by Jae Kwak, Michelle Gallagher, Mehmet H. Ozdener, Charles J. Wysocki, Brett R. Goldsmith, Amaka Isamah, Adam Faranda, Steven S. Fakharzadeh, Meenhard Herlyn, A.T. Charlie Johnson, and George Preti, published in Journal of Chromatography B 15:90-6, 2013,

“A pilot study combining a GC-sensor device with a statistical model for the identification of bladder cancer from urine headspace,” by Tanzeela Khalid, Paul White, Ben De Lacy Costello, Raj Persad, Richard Ewen, Emmanuel Johnson, Chris S. Probert, & Norman Ratcliffe, published in PLoS One, 8(7):e69602, 2013,

“An investigation of fecal volatile organic metabolites in irritable bowel syndrome,” by Iftikhar Ahmed, Rosemary Greenwood, Ben De Lacy Costello, Norman M. Ratcliffe, & Chris S. Probert, published in PLoS One, 8(3):e58204, 2013, and

“Evaluation of odor from vaginal discharge of cows in the first 10 days after calving by olfactory cognition and an electronic device,” by I. Sannmann, O. Burfeind, V. Suthar, A. Bos, M. Bruins, & W. Heuwieser, published online in Journal of Dairy Science, June 27, 2013.

Land of the Rising Scent

Japan is a bit of an enigma when it comes to smell. The Japanese avoid wearing strong Western-style perfume yet their long history of incense creation and appreciation speaks to an ongoing fascination with the olfactory.

The country is home to Smell Club and is a hotbed of innovation in smell technology. Japan keeps alive the dream of Smell-O-Vision: telecom giant NTT created wifi odor globes to provide a scent-track for showings of Terrence Malick’s movie The New World. NTT also produced the i-Aroma, a PC-linked device loaded with six scents. Researchers in Tokyo recently perfected a colliding airstream arrangement that give the illusion of smells emerging from specific places on a computer screen. And Nissan has developed a logoscent for its dealership showrooms and auto show events.

Still, watching items on Japanese smell technology jump off the newsfeed I sometimes experience a little déjà vu. It happened again today and I decided to investigate.

The item was about a gizmo, called the ChatPerf, that plugs into your iPhone and releases a puff of (a single) scent on software command. The link was to a YouTube of a young lady demonstrating the device. The video was uploaded today, but a little googling shows that the device has been kicking since last fall—an eternity in tech dog-years. It debuted in October, followed a few days later by a post on Slashgear, and a day later Elena Vosnaki at PerfumeShrine wrote about it. [I got your déjà vu right here.—Ed.]



The ChatPerf device has a website (still dated 2012) which pictures a different, presumably earlier version of the plug-in. The inventors seem not to know what to do with their new baby—the website implores people to suggest applications.



The questions linger: Is it vaporware? Why keep pushing it? Who is behind it? Can I get one? Does it synch with Fart Piano?

UPDATE May 14, 2013

ChatPerf, the company—or perhaps just some guy in a bathrobe on the 18th floor of the Akasaka East Tower in Tokyo’s Minato-ku precinct—behind the smelly iPhone plug-in gets more free publicity as ExtremeTech and Mashable treat the story as new.

I tweeted my own suggestions for smell-iPhone apps:

Can you change me now? Instant full diaper alert app from Pampers.
Instagas: It’s a fast, beautiful and fun way to share your bean burrito with friends and family.
Belch:  Eat, drink and share the post-ingestive effects with the Belch community of restaurant raters.

Virtual Smellscapes: The Dead Man’s Nose

Via Dead Men’s Eyes

Stuart Eve, a grad student in archaeology at University College London, is the guy behind the Dead Men’s Eyes blog/project (“A mixture of Augmented Reality and Heritage. Bringing past-worlds to life.”). He’s just unveiled the Dead Man’s Nose (DMN), a virtual system that coordinates smells and sights with actual physical locations.

Using my embodied GIS [Geographic Information System] of the roundhouses on Leskernick Hill, Bodmin Moor, I set the DMN to fire off a smell of lovely Barbeque whenever I got within 20m of a roundhouse. I set the fan to run slowly at first and get faster as I got closer to the ‘source’ of the smell. The DMN performed admirably, as I walked within range of the houses I heard the tell-tale whirr of the fan and the next moment I had the lovely scent of cooking ribs. Future models will allow for more than one smell at a time (I just need a couple more computer fans) and also a better housing, a bit of 3D printing is in order! 

Now I can use the iPad to view the roundhouses overlaid onto the video feed, plug in my headphones and hear 3D sounds that get louder or quieter depending on where I am in the settlement and also I can augment different smells as I walk around. Not only can I walk around the modern day Bronze Age landscape and see the augmented roundhouses, hear the Bronze Age sheep in the distance, I can also smell the fires burning and the dinner cooking as I get closer to the village….

Readily available technology + clever design = an immersive olfacto- spatial- temporal experience. Neato!

Obama’s Incoherent BRAIN

We have a good understanding of the sense of smell at the biological periphery. We know how odor receptors work and how the sensory information they convey is organized in the nose and olfactory bulb. Beyond that, however, things get murky. We know the higher brain areas of olfaction only in broad brush strokes. We have barely an inkling how odor intensity and pleasantness are modulated by brain areas dealing with memory and emotion. Wouldn’t it be great to have it all mapped out, from the physical molecule of trimethylamine to the behavioral yuck response?

That is the allure of President Obama’s $100 million brain research initiative. But it isn’t easy to get a handle on what exactly the project consists of.

Susan Young’s piece yesterday on the MIT Technology Review website is subtitled “Obama calls for $100 million to develop new technologies to understand the brain.” She makes it sound like a technology development initiative.

But today’s WSJ op-ed by Paul Allen and Francis Collins calls it “a federally coordinated effort to unlock the secrets of the brain,” to “gain powerful insights into neurological diseases and mental-health disorders.” That sounds like basic research, not technology development. So which is it?

Allen and Collins note that neuroscientists have a host of research techniques at their disposal. They observe that we can study small, isolated neural circuits but we don’t yet understand how the brain works as a whole. To understand “this piece of highly excitable matter” we need a massively scaled up effort, broadened to include “a range of disciplines, from physics and biology to nanoscience, computer science and engineering.” Or, as Allen and Collins say, “All hands must be on deck.”

But here’s what Young sees: “The BRAIN initiative proposes to develop new technologies that can record the activity from thousands, if not millions or billions, of neurons simultaneously at timescales matching behavior and mental activities.” Among the new technologies she lists are nanochips, nano particles, nanoprobes with wireless data transfer, novel optical techniques, voltage-sensitive fluorescent molecules, synthetic biology, improved calcium-imaging methods and enzymes that build ion concentration-sensitive errors into DNA.

So would new techniques get us closer to the Big Picture in odor perception? Perhaps. But it’s not as if we’re hurting for technology. Among the techniques used to study smell are calcium imaging, single cell recording, patch clamp recording, cell ensemble recording, fluorescent imaging of gene expression, optical tracing of trans-synaptic connections, autoradiography, fMRI, EEG, recording of chemosensory evoked potentials, magnetoencephalography, transgenic expression of receptors (a.k.a. “synthetic biology”), genomics, and computer modeling of receptor activation. There is no one path to understanding the Secret of Odor Perception—it is being attacked with a diverse arsenal of technology by a lot of really smart people focused obsessively on smell.

The BRAIN project, in contrast, sounds like a plan invented by the Underpants Gnomes: (1) create technology, (2) ?, (3) unlock the secrets of the brain. Like the Underpants Gnomes, BRAIN fans are adamant about wanting one thing: more. More technology, more researchers, more money. More underpants.

To be fair, Allen and Collins say “progress will also hinge on the cooperation of the public and private sectors, a welcome aspect of the president’s “BRAIN” initiative. We’ll need creative, nimble management to ensure the best work out of both sides.” It’s a bit ironic to hear Francis Collins call for nimble management. He led the federal government’s enormous Human Genome Project which, you’ll remember, was nearly beaten to its goal by Celera Genomics, J. Craig Venter’s private company which started later but moved faster.

Creating a Virtual Odor Source: Fluid Dynamics Meets the Fun House

Smell-O-Vision is the hardy perennial of sensory technology. Engineers and artists keep returning to the idea, tweaking it here and there: a virtual reality helmet, a small device for beneath movie theater seats, and so on. Smell-o-vision is an idea as intrinsically cool as it is goofy. It gets more than its share of giggles: for example, yesterday’s April Fools debut of the Google Nose beta. (OK, the “safe search” feature was mildly amusing . . .)

For some reason, most advances in smell-o-vision technology originate in Japan. At the recent IEEE Virtual Reality conference in Orlando, researchers Haruka Matsukura, Tatsuhiro Yoneda, and Hiroshi Ishida from Tokyo University of Agriculture and Technology demonstrated a smell-capable video display. People have jury-rigged such devices before. The new wrinkle here is that the scent emerges from a specific area of the screen.

How do they do it? Via “computational fluid dynamics.” They aim multiple converging airstreams so as to create an outward flow from a specific location.

The proposed system has four fans on the four corners of the screen. The airflows that are generated by these fans collide multiple times to create an airflow that is directed towards the user from a certain position on the screen. By introducing odor vapor into the airflows, the odor distribution is as if an odor source had been placed onto the screen. The generated odor distribution leads the user to perceive the odor as emanating from a specific region of the screen.

That’s sorta cool, even if it doesn’t live up to the ridiculously overblown headline at Extremetech: “Japanese smell-o-vision TV releases scents with per-pixel accuracy.”

As a proof of principle, this technology opens up all sorts of interesting applications quite apart from video screens. I could see colliding airstreams being used to create pop-up smells in a walk-through environment. Imagine an olfactory fun house . . .

Smartphones with a Sense of Smell?

That’s the claim advanced in a Slate piece about Sam Khamis and his Vinod Khosla-funded startup called Adamant Technologies, in Redwood City, California. The story, by Will Oremus, is rather breathless:

an array of sensors that together could potentially detect anything from halitosis to blood glucose levels to an impending electrical fire. The goal: a device that hooks up to your smartphone, turning it into a personal health monitor. [snip] At first, these apps would require a plug-in device that might cost on the order of $100. But Khamis told me his long-term goal is bigger: in short, a digitized, superhuman sense of smell that could someday be integrated into the phone itself.

Khamis got his PhD in experimental physics at Penn, where he discovered new ways to manufacture carbon nanotubes. Along with some researchers at the Monell Chemical Senses Center, he found a way to couple olfactory receptor proteins to carbon nanotube transistors. Gas phase odor molecules activate the receptors, and the transistors signal the event to the attached circuitry. Thus, a biological detector is, quite literally, wired into an electronic nose/brain. It’s cool, but not unprecedented, technology.

Adamant Technologies, with series A funding to the tune of $2.5 million, is so far a rudimentary web page. Although the Slate piece doesn’t say so directly, the underlying technology may be the receptor-transistor setup, now the subject of a U.S. patent application by Khamis and Paul Rhodes.

I would love to see this technology succeed. Do I think it will? Not soon. As I pointed out in What the Nose Knows, “The usefulness of an e-nose depends on its software as much as its sensors.” Detecting the different molecules found in bad breath is one thing; recognizing that pattern as bad breath is another and it requires big-time statistical processing and adaptive algorithms. Solving the cocktail party problem (tracking a target signal against a noisy background) is still another hurdle. When it comes to e-noses, the math is harder than the biophysics.

But suppose Khamis’s dream comes true and we each have an e-nose embedded in our smart phone. Then what?

It’s January 2017, and you’re riding a crowded elevator to work. An SBD occurs. The odor app on someone’s iPhone 8 says, “Hey, who cut the cheese?” Everyone stares at the doors. Awkward silence. 

After lunch there’s a ruckus. The chick from Accounts Payable is in tears and the HR lady is walking her out of the office. The screen on the HR lady’s iPhone is flashing “Alcohol Violation: Appletini.” 

You go the doctor about a sore knee, but first you have to breathe onto his iPhone as part of the mandatory Obamacare protocol. It detects the Marlboro you sneaked three days ago and your health insurance premium automatically goes up 20%. 

You come home late one night and your girlfriend waves her iPhone around your neck. It registers the new Lady Gaga fragrance. Your girlfriend doesn’t own the new Lady Gaga fragrance. You sleep on the couch. 

You dream about the old days, when smell was still the mysterious sense.

Blowing Smoke, Steampunk Style

The basic principle—acoustically driven smoke/scent rings—has been around for a while, but British architectural student Hee Park has scaled it up impressively. More about him here. I can think of some fun theatrical applications right off the bat.

On Second Thought Who Needs Lassie?

Cal Tech’s Heather McCraig

Is it time to send your diabetes scent detection dog back to the animal shelter? In a blog post on MIT’s Technology Review, David Zax takes a look at the possibility of diagnostic e-noses embedded in smart phones.

Bottom line: a neat idea that needs a lot of work to reduce to practice.

Zax embeds what he oxymoronically calls a “cool video from the American Chemical Society.” The video is about chemical vapor sensors and features a Cal Tech nerdette named Heather McCraig—a graduate student in Nate Lewis’s lab—and some really silly 1950’s-style background music. [Don’t be so square. It’s ironic post-modern nerd music.—Ed.] [Okay, okay.]

One of Ms. Craig’s talking points needs work: the canary in the coal mine is a good analogy for chemical vapor sensors but a bad one for e-noses. (Canaries detect odorless methane gas.) Still in all, she’s good on camera and is working in probably the premier chemical sensor lab in the country. [And her safety glasses are way more fashionable than yours.—Ed.] [Bite me.]

Hat tip to Instapundit.