We all now carry a powerful computer in our pockets everywhere we go. It’s also the HMI to much bigger computing in the cloud. We have entered the age of Everywhere Compute, so what does this mean for the traditional server in the industrial environment? And what does it mean for the people who install, operate, maintain, and use them?
It does mean change – especially for SIs, and I’d like to consider the implications – the different levels for computing and how SIs can help end users to build the right platforms for the Always-On, everywhere-compute world of digital transformation (DX)
First, let’s start with the Edge, and let’s start by understanding the macro-trends that are putting Edge Computing platforms at the heart of the best DX strategies across industry. Many will recall that before we started talking about DX – and perhaps the driving force of DX – was the Industrial Internet of Things IIoT. It’s a phrase I read much less nowadays, but the reason for this is not that it was a passing trend or fad. In fact, the reason you don’t hear people talking about it is because it has been superseded by the more-encompassing idea of digital transformation that IIoT connectivity enables. Internet enabled devices were the latest thing for a while – now they are the normal thing. And where devices and machines at the application edge don’t have in-built internet capabilities, an explosion in smart sensor technologies has offered bolt-on IIoT capability for almost any machine, line, or process.
The huge proliferation in data generated by this ongoing trend has led to the expansion of computing – but how and where computing happens needs to adapt with the proliferation of data now available at the application Edge. For a while, the question was – would all this data be processed in the cloud? Is this the end of servers as we know them? Then, when security and latency issues were considered, and with the rise of Edge Computing platforms, the question switched to – will all data be handled at the Edge? Is this the end of servers as we know them?
The reality, as SIs know very well, is that the future is already taking shape, and it is not one platform or the other, but a combination of the two – largely to the detriment of the traditional server room. So abundant and powerful is the contextualized data from all industrial operations that the only place to effectively compute it is in the Cloud. But not all data is created equal, and not all requires historical or enterprise-level context to be useful. Real time decision making, for example, can’t use the Cloud for processing because the latency is just too big, even if enterprises were happy to send so much critical data offsite and be so beholden to a decent web connection. No, some computing must remain at a site level. But what does this mean for end users?
In short – this all adds up to an Edge-Cloud hybrid built on a simple, autonomous and secure Edge Computing platform. Let’s take each of those points in turn.
The need for a simple Edge Computing platform comes down to a couple of key drivers. Most importantly, by moving compute from a serviced server room, we are necessarily taking it away from IT professionals and placing it in the charge of OT specialists. It would be a mistake to assume that OT professionals don’t have any IT competency, and as a new era of digital natives take over, there is not only more familiarity with digital systems, but also the expectation that computing is simple to manage. After all, we all carry an Edge device in our pockets, with local, real-time sensors (such as accelerators and gyroscopes), as well as access to cloud services (email, photo albums etc). We expect to be able to use the capabilities of our smart phones without specific IT training, and we expect them to be secure out of the box, or at most by adding a security app.
The same goes for industrial Edge Computing – though the requirements to ensure similar levels of usability are different. The right Edge Computing platform must have resilience, even fault tolerance for many applications. Achieving this requires running nodes in lockstep to automatically remove downtime associated with hardware failure by switching seamlessly between nodes. Replacing hardware in such an occurrence must be simple exchange of a hot-swappable click-out, click-in part. Similarly, the Edge-Computing environment must be easy to set up and have the capability to work with existing enterprise security approaches.
While it is too soon to say that the traditional server and server room are dead – not least because many companies will be running enterprise level computing from local servers with real-time Edge Computing at the application level – it is certainly not the best way to manage data in the era of digital transformation.
What this movement to the Edge means for SIs is a topic I covered in a recent blog, (SIs – Now is The Time to Think Differently). In a sentence, it means that SIs face a squeeze on their traditional models that requires them to turn their vital skills and capabilities away from an IT network management and integration model towards a digital business outcome model. By offering Edge Computing expertise and building simple, autonomous, secure, and vendor neutral platforms that offer the benefits of cutting-edge software solutions.
End users are increasingly demanding that their digital transformation strategy offers them the kind of functionality they expect from their mobile devices. SIs offering the best Edge Computing platform can deliver this and protect their own long-term sustainability at the same time