Forget About Apps, Forget About FitBit, Think DCRT

It's technology that can estimate what we eat without any visual prompts.

Think “acceleration sensor!” NTT, Nippon Telegraph and Telephone Corporation, the third largest telecommunications company in the world, and listed on the NY Stock Exchange, back in 2002 issued their “Vision for a new optical generation. 

This report, one of the first to do so, offered an overview of just how important full-scale Broadband would become.  

The report stated that “By "conquering time", our "disposable time" will be increased, and by "conquering distance", the existing "range of activities" of people and companies will undergo phenomenal expansion. The full report is a fascinating read even today, some 15 years later, and offers a great insight into what can become reality.  

Keeping with their ahead of the curve positioning, they have just previewed at their annual NTT R&D Forum a wrist-wearable device that could help us eat a little better and smarter. The technology is called “Dietary Content Recognition Technology” and can estimate what we eat without any visual prompts; allowing us to keep a diet log with writing the foods we eat on a piece of paper, or in an app, or taking its photo. It is fascinating. 

Our arm movements are arranged in a time series, and the amount of each extracted characteristic is considered as one movement. Basically a physical movement algorithm, which estimates what the foods are. The device contains sensors and gyroscopes. The company hasn’t released an official photo of it yet, but yahoo Beauty published one that looks a lot like the iWatch.  

Don’t get too excited yet as it’s only in development, But so far the company has demonstrated it’s accuracy with beef bowls, sushi, bread, curry and pasta. Next step according to the company’s press release is to “examine, for example, how much accuracy is required” and will be conducting market research.

Topline is that if the technology is a smart as they say, for the US consumer, it better be 99.999% accurate.