IoT reflections

Graphic image of a bright light shining amongst data

We were recently asked by a leading drinks publication to advise on IoT use cases and security for an article they were writing. It was only after that I remembered one of the first ever IoT projects was by Coca-Cola with their IoT enabled vending machines. Epic fail not mentioning that! However, it got me thinking about just how far IoT has come in recent years, not to mention the past 40 years when coke connected their first machine.

22 years ago, when I started my career in software development, the first project I worked on could today be considered an IoT project. I worked on enabling robotic soldering machines to be remotely controlled and diagnosed from anywhere in the world. Back then Kevin Ashton had just coined the term IoT so it wasn’t widely adopted or known about but that didn’t mean that toward thinking companies weren’t working on early “IoT” projects.

Over the course of the past 22 years at various points I’ve tried to be an IoT innovator but the sticking point for me was always the hardware. I remember at one point being excited to get a rabbit semiconductors development board and a thick book on how to program it (Stack Overflow wasn’t a thing yet) but quickly realised my embedded C programming skills and lack of knowledge about low level memory management etc was going to turn the project into a long arduous task.

Fast forward a few short years and the likes of Arduino, the Raspberry PI Foundation and even ARM have helped democratise hardware development and pave the way for other semiconductor innovators. The hardware eco system is such now that the array of sensors and dev boards means it’s easy to prototype new products, a lot of the code you need already exists (Stack Overflow copy and paste programmers rejoice) so it’s not a stretch to move these designs to custom PCB’s and manufacture in some volume.

I think it’s fair at this point to talk about software progression. A couple of years ago I gave a talk about how you can build an end-to-end IoT solution in less than 60 minutes. With the evolution of the cloud, it’s possible to almost build a no code IoT solution, configuring services to move and transform data to be displayed in tooling such as PowerBi. This would have been unthinkable even less than 10 years ago. Of course, in 60 minutes these solutions are of limited commercial value but do aid in rapid prototyping and with technologies such as Row and Column Level Security and Azure Active Directory integration it’s possible to start to build a more secure and resilient solution that can scale and be multi tenanted.

Originally, I was going to pen this article about the evolution of IoT security and I’ve somewhat digressed from the intended topic, but perhaps the greatest benefit of building IoT solutions using cloud services is not the speed at which you can build but the security frameworks, features, and tooling that the cloud can provide to help us secure our solutions. I’ll talk about security in more detail at a later date but if IoT is designed with security at the centre and from the start then your already on a winning path.