Visualising the IoTUK Nation Database
A few months ago we started a project with Bloom Agency and IoTUK/Digital Catapult to map the Internet of Things (IoT) sector across the UK.
The Internet of Things doesn't fit into the traditional ways of describing business (e.g. SIC codes). Internet of Things can cover everything from devices which count cars, to river-level sensor networks, to long-range wide area network infrastructure. It is a technology with the potential to cut across every traditional sector and to create new ones. That meant a new approach was needed to identify those companies and organisations that make up the sector. We trawled the web finding companies with a presence in the UK. We've published the dataset on Data Mill North with an open license and there is an API for the data too.
Visualising the data
With a nice, new, UK-wide, dataset it is possible to ask some interesting questions such as:
- How many businesses operate in the IoT Ecosystem?
- What do the businesses do?
- What vertical sector are the businesses in?
- When were these businesses created?
- How many IoT businesses were created in each of the last five years?
- Where are the businesses located, and which locations generate the most investment?
- Are there any significant geographic clusters of activity?
- How many staff are employed in IoT businesses?
- What is the size of the business/market?
- How much and what type of investments have been made into IoT businesses?
- How many IoT businesses exist that generate annual revenues of over £500,000?
To help answer these questions, and because we know most people won't look at raw data, we've created an interactive visualisation of the data driven by the same CSV file that we've published.

At the top of the visualisation we show the headline figures for the number of organisations, their total revenue, the total number of staff, and the total investment. Below these we show the breakdown per sector based on standard industrial classification of economic activities codes. Switching between the tabs changes the breakdowns to the chosen headline area. It is possible to deselect/select individual sectors so that you can examine whichever groups you are interested in (orange is used to signify things that are selectable/deselectable).
The power of hex
Below the sector summary we have two visualisations of the geographic distribution. One is a geographic heat-map and the other is a hexagonal heat-map of the UK's statistical regions. Why a hex-map? Firstly, statistical regions often have hugely different areas. Just compare North Yorkshire County Council (8,600 km²) to Westminster (21 km²). When these are drawn on a map they have very different visual impact even though, statistically, we are treating them as comparable entities. A hex-map gives each region the same visual weight by making them all the same area and tessellating them on a grid. Hexagons do a better job than squares as they have more neighbours which allows the resulting hex-map to feel closer to the real geography. But we still provide a geographic map for those that want it.

You can hopefully see that there are a couple of major clusters of IoT organisations in the UK. Perhaps unsurprisingly, these are in central London and around Cambridge (presumably related to Silicon Fen).
The colours of the heat-map scale to the range of the areas selected. That means that if you just select specific regions you'll see the local hotspots. For instance, in the north of England, IoT companies tend to be based in the larger cities.

Everything affects everything
The rest of the visualisation makes use of interactive barcharts created with a small barchart library. Each orange bar is selectable/deselectable and, as elsewhere, selections affect the rest of the visualisation. The organisations are pretty varied in character so we had to add lower/upper ranges to keep things manageable e.g. the University of Cambridge was founded in 1209 so initially broke our graph by making it cover 811 years.

As I've hinted at, everything from sectors, to geographic regions, to sizes and years are selectable/deselectable. And those selections propagate through the rest of the visualisation. That means you can ask "how many IoT businesses exist that generate annual revenues of over £500,000?" (246) but you could equally ask "how many people work in SMEs in the manufacturing sector?" with just a few clicks. So, the majority of the original questions can be answered but many more questions, that we haven't yet thought of, can be answered too.
Build your own
We tried to keep the visualisation as simple as possible whilst still allowing complicated questions to be asked. However, there will be questions that it can't answer. You may be able to think of better ways of visualising the data than we've have. You may have other datasets that could be combined in different ways. Thankfully, because you have access to the IoT Nation Database as CSV or via the api, you can create your own. We'd love to hear from you if you do.