MagnifiSense: the wearable tech that learns your energy habits

14 September 2015 (Last Updated September 14th, 2015 18:30)

Researchers from the University of Washington have developed a wearable technology which picks up the electromagnetic radiation from a range of machines. By giving the user the means to monitor personal energy use, could we be persuaded to change our consumption habits?

MagnifiSense: the wearable tech that learns your energy habits

Magnifisense

We live in an increasingly energy conscious world, and new technologies are being developed to keep up with the desire to be greener.

Researchers from the University of Washington have developed a new wearable technology which picks up the electromagnetic radiation from a range of machines from food processors to cars. It’s called MagnifiSense and the developers believe that it could have a direct affect on how we use the things that they say we are "pretty dependent" on.

The device determines which piece of equipment is switched on by the level of electromagnetic radiation it gives off. When there is electric current consumption, a magnetic field is created; Ampère’s law is what determines the magnetic field associated with each different current.

The researchers needed a wide range of devices to gather their data, so they recorded the magnetic radiation of 120 appliances like blenders, microwaves, fridges, hairdryers, vacuum cleaners and laptops from 16 homes across 14 cities.

"We tried to have the machine learn the differences in magnetic radiation of these different devices," explained Edward Wang, one of the developers of MagnifiSense and UW electrical engineering doctoral student. "Basically what a device looks like in the magnetic spectrum."

The MagnifiSense prototype is currently a bracelet with three sensors on the top which is connected to an external sound card. It’s quite big and clumpy, but the goal is to incorporate the technology into smart devices as software rather than being its own piece of wearable tech.

The dawn of ‘persuasive’ tech

The device could be influential in changing how much energy people use, Wang believes, because it learns the habits of the person that is wearing it. It knows who turned on the stove, and whether you’re brushing your teeth or blow-drying your hair.

This wouldn’t necessarily just be a means of exposing the ones who are flicking the most switches, but could be a contributor to the growing design of persuasive technology.



Are smart meters really as ‘smart’ as they’re cracked up to be, or could they become unhealthy, dangerous, personal data sinks?


"What a persuasive technology looks at is how can we make people live a certain way," says Wang. "A hot topic right now is how can we make people more eco-friendly."

He added that while he is aware of the depleting reserves of fossil fuels, it doesn’t often cross his mind when he turns on a light switch, which he believes to be a common attitude. One way he thinks that MagnifiSense could play a part in this is by giving motivation to people to live a more eco-friendly life.

For example, if there was an organic food store that was particularly keen on sustainable living, it could offer incentives for customers that arrived there on foot or by bus, rather than driving. It could encourage shoppers to be greener by offering points that add up to rewards or small discounts for eco-friendly regulars.

Finessing the system

The ultimate plan for the MagnifiSense technology is to integrate it into smart phones and watches. The team envision that MagnifiSense could be linked up with energy tracking apps and smart monitoring technology such as remote thermostats. Right now though, the magnetic sensor inside is a "little too slow" according to Wang.

"It’s about ten times too slow, [which] is really not that far to be honest," he says. "If we can get even just one sensor a little faster, this technology is literally a software download away."

Wang pointed out that something else to smooth out is the final algorithm. Right now, the system figures out the appliance you’re using by detecting which one you’re closest to. This means that if you’re using both the stove and the microwave, it will switch continuously between the two. What would be more intelligent, he says, is for the system to figure out that you’re next to both.

He added that this is not necessarily a drawback though, as if you’re on a low power system which you want to have good battery life, you don’t want it to do too much.

"As you add on complexities to this system, the cost is computation," Wang says. "More complex computation takes more power."

Functionality at a cost of privacy, but who cares?

One thing that wearable technology, particularly smart watches, have been criticised for is breaching users’ privacy and tracking them inconspicuously. Wang says that the advantage of technology being wearable is that if you don’t want your energy usage to be recorded, you can just take it off.

"But really do you care if some system knows what your car’s magnetic radiation looks like?."

Also, the way the MagnifiSense algorithm works is a bit like a textbook, Wang explained. It has condensed all the data from the 120 devices to create a model which it can refer to every time something else comes in. It doesn’t see or hear things, because it responds only to electromagnetic radiation, rather than light or sound.

"Our system can’t hear anything that is not electronic," he says. "It can’t see [you] because [you’re] just not even in the right spectrum."

Knowing your activities throughout the day, or whether you took the bus to work is a privacy breach for some people, which Wang understands.

"If you want better functionality, there is a sacrifice of privacy," he says. "But really do you care if some system knows what your car’s magnetic radiation looks like?"