Why The Internet Of Things Involves More Than Just Sensors

When the phrase “Internet of things” gets bandied about, it’s usually taken to refer to collecting data from sensors, whether those are on consumer devices like fridges or workplace devices like temperature controls. But is that too limited a view to adopt?

Car picture from Shutterstock

There’s no doubt that the potential to collect and analyse information from devices that previously were not tracked in any meaningful way is a major part of the hype around Internet of things concept. (It’s also the reason why the notion of capturing metadata from every device for law enforcement purposes is likely to be increasingly problematic.)

But we have to remember that capturing data in itself is not a useful activity when it comes to running a business. Analysing that data and using it to drive decisions about future directions is what actually proves to be useful.

From that perspective, it would make sense to ensure that the data we are already capturing as a result of regular business activities is being analysed — and the brutal reality is that in many cases it isn’t. While most businesses will perform what we might call compulsory analytics — financial reporting being the most prominent example — the willingness to engage in more detailed analysis of other data is often lacking.

In this context, a server running accounting software is just as much a “thing” as a temperature sensor, even if the latter seems far more sexy and modern. If that data isn’t being integrated and used elsewhere, then it’s just as potentially useful (but actually unused) as standalone factory monitors or other sensor-based data. “It’s the handling of that data that’s going to become critical for businesses going forward,” Telsyte analyst Foad Fadaghi points out.

This topic came up for debate during the press launch of a Telsyte study on the adoption of the Internet of things concept in Australia. The report itself wasn’t the source of the controversy. It was the fact that Microsoft, which had sponsored the study, presented a case study in the form of how Dental Corporation (part of Bupa) had adopted an “Internet of things” approach to collecting data from the 200 dental practices which is manages in Australia.

Across those sites, there were more than 15 different practice management solutions being used, which meant that consolidating data and identifying broader trends was virtually impossible. Indeed, the need to collect basic operational data via fax or email often meant that end of month processes took six weeks.

Dental Corporation hired integration specialist Breeze to create a solution that would effectively collate data from all those individual systems. Describing that as an “Internet of things” solution proved controversial to many of the attending journalists, but it seems to me a reasonable enough description, albeit not the first case study that might come to mind.

No doubt if that collection process had also included data from actual dental instruments, no-one would have quibbled. But in any IT process, you need to walk before you can run. If that core data about what treatments patients are receiving isn’t being collated, additional sensor data from the instruments is not going to be any use at all. It makes more sense to solve that first issue before moving on to more advanced cases. “You need to prove the relevance to management anyway,” Fadaghi noted.

Ultimately, there’s no point in getting too hung up on terminology. Cisco tends to favour the term “Internet of everything” rather than “Internet of things”, and there’s sense in that — the label makes it clear that it’s about integration of all your data, not just the newly-emerging category of sensor-driven devices. Saying “I helped solve a business problem and increase revenues” is a lot more impressive on a resume than writing “I worked on an Internet of things solution”.

Evolve is a regular column at Lifehacker looking at trends and technologies IT workers need to know about to stay employed and improve their careers.