IoT architecture explained to a child

Benjamin Daix
3 min readMar 7, 2020


I wrote this article and customised the scheme, first for me and my own understanding of the IoT technical architecture, which always seems complex for a non-engineer profile.

So I tried to simplify it to the maximum, in order for the reader to understand the logic behind each stage of the architecture, consisting of the 5 steps that bring data from a source all the way up to its usability.

As it generates a lot of data, which are expensive to store online and not fully useful, we need several layers of technology and communication devices to filter those enormous data volumes, and according to that reduce the storage costs, improving the efficiency of an overall process.

Let’s have a look at the example of a water purification system.

STAGE 1: the sensors (one device per house)

Under each house, distributed in the water pipes, in the ground, manually are placed the sensors to control the water pressure, quality, leaks… They send all information through the wired or wireless channels to the aggregators located several meters away (no farther).

STAGE 2: the aggregators (one device per street)

Shared across the city buildings, the aggregators collect the sensors data in their immediate proximity. They play the role of batons to filter, control and send the processed (but not analysed) data to the next stage: the edge device.

They can also initialize the IoT they are linked to, in case of emergency. In our example, the aggregator can activate a pipe to shut down after a leak detection.

STAGE 3: the edge (one facility per city)

It would be too expensive and inefficient to send all data to the cloud without an initial analysis. Located some kilometers away from the aggregators, it will receive the data from the edge facilities stock, filter and then forward any relevant data to the cloud facility.

STAGE 4: the cloud (one office per state)

The edge facilities of all cities of a state send the most crucial and valuable data to the cloud, where an in-depth analysis will be run. Thanks to those filtered data, algorithms can be efficient and provide the most accurate information, in order to make decisions on controlling and improving the water management system.

STAGE 5: the application (one per user)

This stage is the reason why we’ve deployed all the previous stages!

Based on the analysis of the cloud data, we can now create and sell a service that consists of operational improvements to all or specific water treatment facilities, or offer a leak detection service for every household.

This description is voluntarily simplified.

I hope you find it useful.

About me // French Marketer who specialises in B2B tech and IoT. With the background of lobbying and over 10 years experience as creative director, I help IoT start-ups achieving stable growth, meaningful branding and long-lasting demand generation. More on



Benjamin Daix

French Marketer specialised in tech and IoT / Based in Warsaw (PL) /