Calorie-Counting ‘Magic Wand’ Raises Controversy

Founder Isabel Hoffmann “counts” the calories in a pack of corn chips with TellSpec’s technology

A Toronto entrepreneur has crowdfunded a device that claims it can count the calories in your food. But can she deliver?

There are two ways to see Toronto entrepreneur Isabel Hoffmann: She has either revolutionized the field of food science, or she has crowd-funded nearly $400,000 for a product she can never deliver. According to one B.C. scientist and businessman, it’s probably the latter.
Hoffmann is the co-founder of TellSpec, an Ontario startup developing a hand-held device the company says can tell you what is in the food you eat. According to the promotional video accompanying an Indiegogo campaign, the idea is simple: a user points the TellSpec device, which looks like a computer mouse, at any food. The device shines a laser at the target and relays the rebounding light signature to TellSpec’s servers. The company’s sophisticated algorithms then determine whether the food contains such ingredients as soy, gluten or pesticides.
Hoffmann and her co-founder, York University mathematics professor Stephen Watson, raised $386,392 in two months and promise to ship a product as early as next August. The crowdfunding campaign gained steam from positive press articles in the National Post, Canadian Business, Mashable, Metro, and othersnone of which include scientific assessments of TellSpec’s claims.
UBC spectroscopy researcher Ed Grant is more skeptical. He is not only the head of a UBC research group in spectroscopy, but is also a businessman, having founded and run spectroscopy research company SpectraCode for over a decade before selling it in 2006. BCBusiness asked Grant’s lab to evaluate TellSpec’s claims.
Grant says the TellSpec product will probably work, but not how the company promises. At best, he says, the device will be able to roughly identify what food a user is eating, and then reference a pre-made database of ingredients, not much different than just flipping over the package to read the label.
That’s a far cry from TellSpec’s demonstration video, in which Hoffmann waves the device over a corn tostada and determines that it contains tartrazine. According to Grant, that kind of analysis is difficult in a laboratory under controlled conditions with tens of thousands of dollars worth of equipment. With a device like TellSpec’s, “That’s virtually impossible,” Grant says. “And I don’t feel uncomfortable saying that. With the platform that they are proposing, if they were even able to quantify a major component of any arbitrary substance, I would be really impressed.”
Hoffmann maintains that her critics just don’t understand what she is doing. She says her invention will work because of groundbreaking mathematical techniques that she and her co-founder have developed. She compares her work to solving Fermat’s last theorem, a legendary difficult mathematical problem in the field of number theory that was solved by a mathematician in a completely different area of study.
“That’s why there are skeptics out there,” she says. “Because they’re not professors of mathematics.” Hoffmann says the company will not publish any clearer information about how TellSpec works until the company has secured a patent, but she maintains that it is fully functional. She points to a video demonstrating a blocky-looking prototype of the device, showing it identifying ingredients in a cracker and fruit. And even though TellSpec does not employ any experts in spectrometry, according to Hoffman, she says all the engineering will be dealt with by other companiesprobably in China.
Jason Jiang, the CEO of a Chinese micro-spectrometry company, is also a TellSpec sceptic. His company, Inno Optics, builds the kind of tiny spectrometers that Hoffmann hopes to use. Inno aims to roll out a new miniaturized device in the next year that will do only some of what TellSpec can, at a price of $10,000.
A $320 food scanning laser, he says, is a “fairy tale.” The small size and power of a miniature spectrometer are just not enough to collect the data for complex food analysis, with or without a sophisticated database, says Jiang.
“I like to believe that’s the future of this technology,” he says. “But we’re still many years away from that.”
Hoffmann says she hopes to raise another $5 million to fund TellSpec’s production in the next year. If she succeeds, she says a working model will be in her funder’s hands next August.