A few years ago, Matt Webster decided to dispatch with the annual birthday surprise dilemma and asked his wife whether she’d like as a gift an activity monitor. She was not impressed.
The problem wasn’t him asking, but the technology itself. “If it doesn’t automatically track the calories I eat, then I don’t want it,” she told him.
Although there is no such automated device on the market, Webster, who works as a senior scientist at GE’s labs in upstate New York, was not deterred. His specialty is diagnostics and biomedical research and he decided to build a device that could count calories in any food. “I thought that this was crazy and impossible, but I took it as a challenge,” he says.
Webster started by chomping through a massive food library compiled by the U.S. Department of Agriculture. The database holds nutritional information for 6,500 foods. “I wanted to boil it down to a simple recipe that determined calories from a small handful of data points,” Webster says. “Perhaps I could measure those data points with sensors and use them to calculate the calories in any food.”
Applying the elimination method, he started with a menu of fat-free foods, before moving onto more complex items. “I walked through it rationally,” he says. “I eliminated fat and accounted for water to figure out what the average calorie density was.”
The analysis allowed Webster and his team to write an equation that estimates calories in food with just three simple measurements: weight, fat content and water content. “The equation takes the fat, water content numbers and assumes values for the rest,” he says.
The rest is a combination of sugars, carbs, proteins and other ingredients. “You actually don’t need to know the details,” he says. “We just have to account for it. That’s the secret sauce.”
To gather that data, the team is developing advanced electronics and sensors that shower food with microwaves and look for fat and water signatures in the waves that pass through. “You can do this because water and fat interact with microwaves very differently,” Webster says.
The GE team together with researchers at Baylor University’s Department Electrical and Computer Engineering is now testing the system on simple mixtures of oil, water and sugar. They have built a prototype, but the prize is a push-button device that could be in every kitchen.
One day the team could link the device with a smartphone app or a workout wristband. Says Webster: “I am working on my wife’s dream present.”
Filed Under: M2M (machine to machine)