From theory to reality - powering up Industry 4.0

Chris Pope Profile picture for user Chris Pope November 28, 2017
Chris Pope, VP Innovation, ServiceNow, brings Industry 4.0 to light in practical, realistic terms - focusing on the five layers of intelligent automation.

automation intelligence
We have talked about the path to so-called Industry 4.0 here on diginomica. This term for a new age of industrial man and machines describes a seismic shift in infrastructure, operations and processes that is characterised by the arrival of cyber-physical systems and the layers of software-driven intelligent automation that they run on.

In more basic terms, our working environments have been penetrated by the arrival of technology at almost every level. So what does Industry 4.0 look like closer up? How do we break down the layers of intelligent automation inside it… and how do we flip the start switch to make it all happen?

Driving Industry 4.0 with intelligent automation

One of our first challenges is in accepting that robots and so-called Robotic Process Automation (RPA) have arrived for real. We are now learning the part that these machines and software systems can play in the way all organisations now refine and finesse their business models. The opportunities here are both massive and diverse, which means that this new realignment of industry will take some time to fully engineer and fully exploit for positive gain.

In the Industry 4.0 era we see that firms will now be structured around a deeper core of system interoperability and information transparency, which in itself will enable a new era of decentralised decision making. There is a liberation factor here and it works in a positive virtuous circle. That is, as we humans start to understand when and where we can apply RPA, we can further tune our workflows to be even more efficient.

In this new environment of information clarity and data analytics-empowered efficiency we can now free up our human workforce from repetitive tasks that can be specifically codified for machine brains to execute.

By connecting machines, workflows and systems, businesses are creating intelligent networks along their entire value chain and supply chain that can control and manage each other autonomously. But this is no plug-and-play quick fix because we must remember that data on its own is worth very little. It is only when we start to apply business context to our data and the services and applications that this data exists in that we can start to call ourselves an Industry 4.0 business.

Five layers of intelligent automation

Looking at precisely what our intelligently automated, machine-driven computer brains can actually do for us, it comes down to automation by degrees. If we have to distinguish five layers of intelligent automation today (and this is not an industry standard definition) then they could be considered as follows:

Self-optimisation - This is the ability for devices, applications and higher-level IT systems to adapt to changing conditions brought about by changes in data volume, velocity and (in some cases) variety. In software engineering terms this means we have a responsibility to build from ground zero with an appreciation for the need to architect upwards and outwards. If our IT stack is built with an inherent dynamism, then self-optimisation can be brought to bear at a wider and higher level for the long term.

Self-configuration - Devices can be powered up with enough knowledge of their surrounding network and wider environment to configure themselves automatically i.e. intelligently. This is a question of building software with the ability to a) know what it knows at the point of deployment and b) know it may not initially know everything it needs to know at a future point in time.

Self-management– Self-management is characterised by actions that are capable of addressing faults, errors or performance issues.

Cognition – This is not cognition in the sense of self-aware sentient intelligent machines, this is simply the point at which we start to give our machines a digital identity and reference so that they can learn from events (these being quite literally ‘data events’) that happen across the network. When our machines start to realise that there is a wider network of activity out there (because we define it and describe it and tell the machines that it is there) then they start to gain a degree of cognition about the universe around them.

Intelligent support – When all the above factors start to coalesce and function in seamless harmony (or at least in some form of occasionally syncopated rhythm), then we can start to build upwards and provide a level of support that is intelligently empowered by all the automation layers that lie beneath it.

Maintaining integrity

So then… Industry 4.0 becomes a reality and we just switch it on, right? Well, in a word, no. This is not how it works. We need to appreciate the fact that the business function, even in its pre-intelligent automation phase, has been working as hard as it can. People need reassurance that the integrity of production processes currently in place will not be disrupted.

Creating, developing, engineering and ultimately powering up an Industry 4.0 business is never an overnight process. But even if we accept this truth, getting an entire enterprise (or small- to medium-sized business) to start operating as a next-generation operation should be a Zen-like operation—one brick, one data stream, one workload, one service and one application at a time.

Industry 4.0 is here, but the electricity grid disseminating its power is being lit up one ‘neighbourhood’ at a time. Power surges only risk blackouts and burnouts, so let’s move forward carefully.

A grey colored placeholder image