No two tomatoes are necessarily alike
It's important to note that when you scan a tomato, for example, the sensor isn't identifying your target as a tomato, and looking up a typical tomato's average nutritional scores in a static database. No, SCiO susses out the specific fat, carb, protein and calorie content for the very tomato you've scanned.
The key, Sharon says, is for the Consumer Physics database to have molecular profiles of everything we might target. He says that when SCiO ships to Kickstarter backers, he expects the database to recognize about 80 percent of all the various foods we eat. Proteins are the hardest to detect, followed by carbs, and then fats are the easiest to profile. But in the end, Sharon says, all these materials are within SCiO's reach. Only transparent materials — like clear liquids — present serious trouble when it comes to identifying molecular structures.
When the SCiO takes a reading, its light penetrates a few millimeters into its target's surface, and covers a circle roughly 15 millimeters in diameter. So if if you intend to scan a fruit tart, for example, you'll get different readings if you first scan the topping, and then the custard underneath.
Still, the system is packed with potential, assuming all of Sharon's plans prove out by the time SCiO begins shipping. He says the technology is powerful enough to determine the difference between unripe, ripe and spoiled fruit — useful for casual produce shoppers. Likewise, with the right database information, SCiO could recognize milk that's been tainted with melomine, an additive that caused a scandal in China in 2008.
Forget food, let's scan some stuff
Consumer Physics is focusing on food for its initial roll-out, as the world is obsessed with nutritional data, and calorie tracking presents such an obvious consumer use case. But the sensor technology itself is capable of analyzing a wide range of substances. Only metals and other reflective materials present serious scanning challenges, Sharon says.
Luckily, pharmaceuticals are easy to scan, so when SCiO launches, its apps will also be able to determine the identity of various pills. During my product demo, we scanned a generic ibuprofen pill, and then a name-brand version. Both nondescript pills appeared in the Consumer Physics database, and the app was able to tell the difference between the two.
SCiO launches with an open API, so third-party developers will be able to extend the list of materials covered by the Consumer Physics database, and also create apps suited to specific use cases. Sharon that in same cases, only 10 samples of a particular material are necessary to establish a reliable database profile.
Sign up for CIO Asia eNewsletters.