This article is more than 1 year old

The Internet of things is great until it blows up your house

How to stop hackers letting the gas flow in your connected oven? Bitcoin has the answer

A few months ago I had a chat about the Internet of Things with the design head of a well-known home appliance manufacturer. Gartner had just published 2014’s hype chart, and with the Internet of Things sitting at the very peak of the hype cycle, he reckoned it might be an interesting way to differentiate his firm’s products in a market filled with cheap Chinese appliances.

After our chat, I had a thought: I could teach him about the Internet of Things with some broad-brush product designs. After reviewing the line of products manufactured by his firm, I found two that I could reimagine, using the pixie dust of intelligence and connectivity.

Nearly every home in Australia has a clothes iron. The major difference between models is how much steam they can put out on demand. Every iron has a dial to set its temperature - and if you don’t set it just right, you can damage your clothes.

The solution to that problem seems obvious. Design an iron equipped with Bluetooth LE, linked to a smartphone, running an app that uses its camera to scan a QR code printed on a fabric care tags. This QR code contains all of the care information for that article of clothing, so every time that dress or dress shirt goes under the iron, the app adjusts the iron to the ideal temperature.

Extending this idea, it’s easy to imagine setting the temperature and cycle for both washing machine and dryer based on that fabric care tag.

All of that additional capability would add about $5 to the cost of goods, or around $20 at retail. Not much to pay for something that could prevent a fair bit of damage to clothing.

Next, I took one of the absolute necessities of cold winter nights - the electric blanket - and made it smart. Today, electric blankets feature a dial and thermostat, none of which compensates for the fact that our body temperature, and need for heat, changes as we pass through each evening’s sleep cycle.

Again, the solution seems obvious. The electric blanket should sense the ambient room temperature, the temperature setting of the blanket, and the activity of the person beneath the blanket, using these to generate a heuristic for when and how much heat is needed across an evening’s sleep.

Once that heuristic had been learned, you’d have a smart electric blanket - one that would never need adjustment. You wouldn’t even need to turn it on. After it had learned the particular needs of its owner, the electric blanket would turn itself on at the right time, and be at just the right temperature - relative to the ambient temperature - to keep them comfortably warm. It would adjust its temperature to accommodate a falling body temperature in the hours toward dawn. It would do all of this invisibly and automatically.

An electric blanket would be hard pressed to do all of this computationally expensive work on its own. Those readings would be uploaded to the cloud, where sophisticated analysis algorithms could be run over the data, checking it against thousands of other samples from other electric blankets, producing a best guess at a temperature regulation model for the blanket.

This ‘internet of electric blankets’ would also tie into Fitbits and Jawbones and the Apple Watch, and whatever else we have on hand to sense our activity when we’re sleeping. The electric blanket listens to all its sensors, communicates with the cloud, and tunes its behaviour to always deliver the best possible night’s sleep.

An internet of electric blankets can be a platform for ubiquitous computing. Wherever an electric blanket owner goes, their heuristic travels along, so whether they used a smart electric blanket in Melbourne or Moscow, they would get the very same great night’s sleep, because their sleep heuristic lives not in any one blanket, but in the cloud, and can be downloaded to any smart electric blankets, anywhere.

To do all of this would add something less than $12 to the cost of goods for an electric blanket, around $50 at retail. That’s a lot for a blanket that might only cost $150, but when the manufacturer offers ‘Electric blanket-as-a-service™’ - free for the first year, and at a modest annual fee thereafter - it becomes very appealing. The internet of things means appliance manufacturers realise new profits as service providers.

More about


Send us news

Other stories you might like