Let’s start off by defining what I mean when I refer to ‘The internet of things’.

I think of it as putting computers into everyday objects that record data and then share it wirelessly with each other. They make big data on the fly. Their big data is fed into complex algorithms that then automate tasks for you.

Like a school’s smart fire alarm calling 999 for the fire brigade when it’s been going off for over a couple of minutes. Oh and also, to show it’s got a little computer inside of it just add the word ’smart’ in-front. e.g. Smart apple peeler or smart door stop, you get the picture.

Now we are all on the same page let’s start properly. My hypothesis is that with all these smart devices collecting data, the next step is to find a way of behavioural learning: analysing why they make silly decisions and trying to find a way of changing the settings to stop the algorithms from making such mistakes again.

We can see if this would work by looking at who is using big data at the moment. Data scientists look for insights through crunching big data sets. After their nifty big data reformat is pumped into a chart they can then do analysis by looking for patterns. They can then make decisions from these patterns in the data about where to save money, what new product to develop or which markets to expand into.

It’s then helpful – and perhaps more relevant – to think about proto-personas, and in turn how we can benefit from ‘smart’ devices that can learn. Most people live in a house or a flat, so lets look at how a smart house could save you money or even how we can use big data to save the environment.

User goal first draft:

“I want to save the environment, and I want my smart house to help me do it!”

This premise could lead you to a ‘smart’ house that automatically opens curtains and switches lights off when it’s sunny outside. All very clever. Maybe not so good if you like your privacy and have especially nosey neighbours. You signed up for eco-friendly, not Big Brother.

Rephrased user goal:

“I want to save the environment whilst preserving my privacy.”

Ok, we have our first UI component, a way to set your priorities through personalisation.

Learning from your mistakes is a poor way of working out how to get your personalised settings. What about looking at forecasts to see what behaviour your house will take in any given scenario? For example, smart cabinets and refrigerators that order food for you before you run out. However, automation of this may turn you into a chocoholic and feed unhealthy habits.

At this point, it makes sense to mix goals with personalisation settings: “I want my smart house to make a online shopping list so I can make meals that will help me lose weight and be healthy”.

Understanding what smart objects/furniture you need to buy to monitor your goals would be the next problem…