Explainer: the Internet of Things
Many, if not most, of us are now connected to the internet, and we have become familiar with it: we shop, we bank, we socialise online. The Internet of Things is not a different internet but refers to the connection of “things” that are also on the internet.
Imagine sitting at your desk in the office and controlling the appliances in your home ready for the end of the day. You could switch on the heating or find out what’s in the fridge for dinner. When you arrive home and sit down in your favourite chair while your partner is travelling on business, they might be able to give you a hug through the chair.
When you leave the house for a run, you’d be able to check that you really did lock the front door without having to return, and you can rely on your trainers to tell you how far you’ve run in a day or how many calories you burned doing it.
Your car might monitor its performance, notifying the manufacturer when it begins to behave outside its expected performance envelope. After consulting with the manufacturer, it might even go so far as to look at your online calendar and book itself in for a service at a convenient time.
Current estimates suggest that if there are around two billion people online, there are around ten billion things online, and this will grow to 50 billion things by 2020. The internet is the global network connecting computers in every country, whether through optical fibre, ethernet, broadband or wireless.
The Internet of Things might be better named Things on the Internet: it is not a precisely defined term, but usually refers to the ability to track and perhaps control things connected via the internet. It gives computers the ability to monitor and manipulate the physical world, blurring the distinction between the virtual and physical, and potentially moving the bulk of internet communication from human-human communication mediated by computers, to computer-computer communication mediated by humans.
How it works
To put a thing on the internet, you first need an identity - or address - for the thing in question. If it’s to be tracked, provide data, or be controlled, it needs to be distinguished from all the other things. Architecturally, a key question is what comprises this set of “other things”. If it’s every device on the planet, then the number of addresses required is very large; if it’s every other device in your home, the number of addresses required is very much smaller.
Next, you need to have some way of recognising the identified thing. In many cases this can be as simple as scanning a barcode attached to the thing from an Internet-connected reader - many stock control and related systems work like this, tracking objects using networked readers. More recently, Radio Frequency ID technology has allowed the “barcode” to be embedded in the object itself, and to be read without requiring a line-of-sight to the object. Even more recently there’s been interest in using visual codes that are aesthetically pleasing, and can be made to fit in with the many patterns around us.
Finally, the thing needs to be networked. This might be by proxy as in the barcode or RFID scanning examples above. Alternatively, many modern devices will now embed some sort of radio technology and be capable of direct connection to the internet using the standard internet protocols - in these cases, the thing may be identified by its network address.
As with many technology revolutions, it’s already happening in parts. Many of us in the UK will already have smart metering for water, allowing your water meter to be read wirelessly from a passing van.
Smart metering of electricity is currently receiving much attention, from both technical and social perspectives. Some proposals go so far as to suggest that the energy supplier will be able to control when and how household appliances are used. Home automation systems, from a number of providers, are growing in popularity and allow remote monitoring and control of, for example, lights, electrical sockets, doors, and windows.
However, big challenges remain before the Internet of Things becomes a feature of our every day lives. There are social implications and questions to be answered about our understanding of the data that would be produced by our things. Recent revelations about access to and storage of our human-human communications data have caused widespread consternation - but consider the implications of having all of your interactions with all objects monitored and recorded.
Given that many people find it hard to correctly configure and secure single devices like PCs, how will they fare when presented with the “smart home” - effectively a complex distributed system containing many tens or hundreds of networked devices, including critical domestic systems like electricity, water, central heating, refrigeration, washing machines and lighting? How will we collate, store and manage the rich, and often privacy infringing, datastreams from all these tracked and connected things?
This points to a need to enable meaningful Human Data Interaction with these datasets. Although we’re well on the way on the technology front, we can’t afford to sideline humans altogether.Comment on this article
Richard Mortier receives funding from Research Councils UK.
University of Nottingham provides funding as a founding partner of The Conversation UK.