IOT (Internet of Things) is a buzzword, which has quite a lot of different meanings. The possibilities it offers combined with AI tools are the major factors behind the all the hype about the 4th (and 5th) industrial revolution. In this article, we share what we think this is all about, and how it might be important for business innovators.
There’s always a lot of confusion around buzzwords. This whole blog is dedicated to the practical, business and industrial applications of AI, and for that matter, we simply can’t avoid talking about IOT. There are some important opportunities hiding below this expression. Problem is, there is a lot of overlap between these buzzwords, especially because the best value is in the combination of them. These new expressions often make the impression of bringing on something new, something extra, although sometimes it already existed for decades, under a different guise.
Think ‘smart’. All this buzzy expression means is that ‘smart’ tools can communicate with other smart tools via Bluetooth or WiFi. Basically that they are wired, they can be accessed from a digital network.
What IoT really means
IoT basically means the ‘smartening up’ of a tool, which already a has a function beyond communication. The TV, the phone and the watch is already ‘smart’, now the termostat, the washing machine and the car is following their footsteps. These are appliances that already had clearly defined functions (a washing machine is there to wash clothes), whereas a TV has always been a media platform, it can just do more and better of it (digital and on-demand, instead of analogue and broadcast).
The Internet of Things mean nothing more that there is a growing number of devices have functions for receiving, sending and recording data. These things can thereby talk to each other.
‘Internet of Things’ has only a slightly different meaning from ‘smart’, as a ‘smart home’ is just a home that is filled with ‘IoT’ stuff. These images are sticky and viral, because mostly everyone can relate to them - the people who read tech news usually have homes, tv sets and kitchen appliances.
So, is there all to it? Everything that runs on electricity can be connected to the network and voilá: Internet of Things?
Basically yes, but the situation has become a little more complicated than that by now.
When IoT took the Internet by storm, it was mostly due to a single new product: the Nest.
The smart termostat company has been acquired by Google for 3.2 billion in 2014, creating widespread speculation, that Google will soon rewire all our homes with smart tools.
Nest was different in other ways as well, though: it was capable of learning. Its ‘smartness’ went beyond communication, it memorized its usage pattern and was able to automatically follow it.
Now, if you give Machine Learning capabilities to devices that can communicate to each other in realtime, you suddenly created a new level of complexity. These could lead to the utopic image of the true ‘smart home’ which actually adapts to the life of its human owners, optimizing everything towards their needs automatically.
It’s already pretty commonplace that you can set the heating, the dishwasher or the microwave to certain triggers (“when I start the engine of my car at my workplace, start heating up my dinner”). But all these require considerable efforts to set up, and you need to review them constantly, every time the pattern changes a bit.
This is where AI could step in, to make the whole experience smoother, without the need for micromanagement. IoT digitizes our physical reality, the AI can utilize the immense amounts of data and form it into dynamic action.
There are some real life examples out there already, like the ‘ET City Brain’ in the city of Huangzhou, China. Every traffic light has a built-in camera, so the central computer knows exactly what the traffic situation is: and dynamically adapts the traffic lights programming. Yes, this means that ambulances get ‘green waves’ so they can cut through the streets smoothly.
AI and IoT, or ‘AIOT’ can be a victim of overhype - developers, like us, know this all too well. There are numerous industrial and business use cases where the sensors gather data, and this data is the resource for the algorithms which process and react to it, we wrote about them here.
We will see many new examples in the near future. Coordinating drones, managing fleets, logistics in general is one area where the most interesting practical use cases might spring up in the near future, as there is a huge efficiency potential in optimizing networks of vehicles.
‘Smart’ doesn’t need to stay at ‘home’ either, because the office can benefit from dynamic resource sharing and other solutions that play around the humans doing work within their walls. And that is how these buzzwords sometimes carry more depth than it might seem at first: the richer and more refined the data is, the better solutions machine learning algorithms can come up with. We are only scratching the surface of what the ‘Internet of Things’ might actually mean for our lives, by the end of this decade.