Are we swimming or drowning in IoT data?

  • September 10, 2021
  • Steve Rogerson

Steve Rogerson picks through the advice he was given at this month’s IoT Tech Expo in London.

Does the IoT have an image problem? I found myself asking that question this week while listening to a talk by Harry Tayler and Joe Savidge from Palantir at IoT Tech Expo in London.

Palantir is a Colorado-based software company that specialises in big data analytics.

So, what about this image problem? Well Tayler thinks we do because when people start talking about IoT the vision drifts into a science-fiction type view of people looking ridiculous wearing AI visors, drones hovering around scaffolding, or people dressed in exoskeletons that look like something from an Iron Man movie.

The reality is more down to earth, he said. Well, yes it is, for most but there are people wearing those AI visors; I saw one on a stand at the show. There are drones hovering round buildings, not fixing them, admittedly, but inspecting them. And exoskeletons exist, though you are unlikely to bump into one while doing your weekly shop.

Joe Savidge (left) and Harry Tayler from Palantir.

So while Tayler says these are a long way from where we are today, I think he is wrong. We are there today, but he is right when he says for most people the reality is more basic. As he said, companies do have more immediate issues with real processes and real shareholders.

“They don’t want the future, they want something for today,” he said.

A key part of this, and the common question when the person who holds the purse strings is asked to fund an IoT project, is will it give a competitive advantage. I would argue having drones inspecting buildings probably does, but Tayler’s message was more basic: There is no need to reinvent the wheel.

In the same way that cloud software has become commoditised over the years, the same is now true for much of the IoT. People do not build their own clouds, and now they don’t really need to write their own IoT software.

“There’s loads of it out there,” said Tayler. “An in-house platform can take years to build. With a SaaS platform, you could do this work in days. That is why building your own IoT structure does not make sense. It takes a long time and costs more than the equivalent SaaS platform.”

And that is only part of the problem. In the time it takes to build an IoT platform, the technology will have moved on and when it is ready, it will already be out of date. Bang goes the competitive advantage.

Marcelo Lecocq, a research director at IDC.

The other issue is, as always, security. As Marcelo Lecocq, a research director at IDC, told the conference, the trend to work from home caused by the pandemic was causing companies multiple security headaches. They are having to live with IoT devices connected to home networks, and that does not make them happy.

“Home workers have become a target for cyber criminals because of really-easy-to-hack home routers,” he said.

And Tayler said with nearly half of executives listing security as their main worry, nobody should be trying to build this in house.

Savidge added his advice was to start small.

“I have seen projects where thousands of sensors have been installed and the engineers are drowning in sensor data,” he said. “They are looking for needles in haystacks.”

In a way, that quote from Savidge summed up a lot of the conference, not in what it said itself but in the fact that it was a snippet of advice, and such soundbites were all over the place:

Karen Welsh, IoT technical project manager at Facility Performance Consulting.
Peter Brogan (left) and Thomas Therkildsen.
Sophie Scannel, modelling and analytics product lead at Palantir.
  • “You have to start with standards. You can’t get fruit without roots,” said Karen Welsh, IoT technical project manager at Facility Performance Consulting.
  • “Not everyone needs to be a technical expert, but you need to ask the right questions,” said Peter Brogan, head of research at the Institute of Workplace & Facilities Management.
  • “Intelligent networks become very important in the way they interact with the environment. For many applications, the network will have to support multiple use cases,” said Thomas Therkildsen, senior business development director at Syniverse.
  • “Be realistic about what you are starting,” said Sophie Scannel, modelling and analytics product lead at Palantir.

I could go on, but you get the picture, and that is without mentioning Welsh saying you need a data lake, not a data swamp. Oops, I have said it now.

And that brings me back to the start, and the image problem. Yes, the IoT does have an image problem, but it is not the futuristic drones and exoskeletons, they just add to the excitement, it is the fact that, as well as drowning in a data swamp, we are also drowning in advice about what to do. So, do I have any advice to add to this torrent? Well, yes I do, and that is jump into that swamp and start swimming. Yes, you will make mistakes, but nobody ever achieved anything without being willing to make mistakes.

Good luck on your IoT journey.

The show was held at Business Design Centre in London.