X
Business

Your network will be your life coach: Intel

In the future, your network will help you decided between a career as a fashion model or a doctor, and tell you to lay off the sugary snacks, according to Intel.
Written by Jo Best, Contributor

In the future, your network will help you decided between a career as a fashion model or a doctor, and tell you to lay off the sugary snacks, according to Intel.

The chip-makers' researchers have been working on making technology able to understand the context in which it's being used and its environment, with a view to being able to make recommendations to its user on everything from their choice of TV channel to their future career path.

According to Andrew Chien, VP and director of Intel Research, corporate technology group, the company has been working on ways to make computers more aware of their surroundings, with a view to creating devices that can take in nebulous raw data from sensors to understand complex ideas like its user's mood.

"[Context aware computing] can figure out stuff like 'gee, are you happy right now', 'are you eating things that are appropriate for your dietary needs or goals?'," for example, by analysing sensor data provided by networked crockery, Chien said.

Such systems could also provide users with reminders from the immediate — 'your yoga class is in five minutes' — to the complex — 'the restaurant you were thinking of going to is getting busy these days, perhaps it's time to make a reservation' by analysing data from a range of sources and conflating them in order to make suggestions.

Intuitive computing could even one day be used to "nurture" or "coach" its user by answering questions like: "'Can you give some guidance? Should I become a fashion model or enrol in medicine?'," the Intel exec said. "These are very long-term goals."

While such sensor technology isn't new, said Chien, the biggest challenge is to vastly improve the accuracy. Intel's own sensor research is working on the premise of such analysis needing to be 90 percent accurate at least 90 percent of the time.

"It better be right almost all of the time else [devices] are going to be more annoying than the device we have today," Chien said.

Sensor-loaded technology will also need to keep power requirements low and protect the privacy of its user, he added.

However, such intuitive computing might not be realised in the immediate future: analysing real time video requires four teraflops of performance in a handheld device, according to Chien.

Jo Best travelled to Shanghai as a guest of Intel.

Editorial standards