Kathy Gibson reports – As companies around the world shift their manufacturing to Industry 4.0, there are some important considerations they should keep in mind.

Tilman Taubert, EMEAR IoT manufacturing lead at Cisco, shares some of the lessons he has learnt from customers moving their IoT projects beyond the pilot phase.

An important consideration that companies have to resolve is the convergence of information technology (IT) and operational technology (OT), he says. “A lot of organisations are going through the process now of finding the right fit: who supports what; who is driving what; and who is operating what?”

Next-generation architecture is another important discussion, and new technologies could change some existing perspectives, Taubert adds.

Arguably, one of the most important considerations is security. “From a security point of view, the whole industry is moving to Zero Trust,” he says. “This is a quantum leap from the current security standpoint, but we are seeing that’s the direction leaders are moving in.”

However, technology considerations shouldn’t be what drives Industry 4.0, Taubert says. Organisations planning to leave the safety of their pilot projects and move to widespread deployment need to rather focus on value drivers.

“You need to be thinking value-backward rather than technology-forward – focus on industry-specific value drivers,” he advises. “Develop a compelling vision and inspire the organisation through lighthouse projects.”

This is the advice from McKinsey’s “Capturing value at scale in discrete manufacturing with Industry 4.0” study, which also advocates mobilising the organisation and innovating the infrastructure.

To mobilise the organisation, transformation should be driven from the top with clear business ownership. Projects should lock in bottom-line benefits to prove value early. Capabilities should be built throughout the organisation and a highly active culture cultivated.

In terms of innovating the infrastructure, it’s recommended that an integrated target technology stack based on a thorough analysis of the status quo be defined. Organisations can then choose a transition pathway which should be enabled locally before scaling globally.

“And this is where technology comes into play,” Taubert says. “The role of IT is expanding to manage and secure the industrial network and to ensure business success.”

The importance of the network cannot be over-emphasised, he adds. “If we are building industry 4.0, connectivity is an underlying imperative – without it, manufacturing would stop. It is critical and integral for every technology because everything relies on it.”

In the modern factory, where automation and control are mission-critical, network uptime and quality are key to increasing production, reducing costs and improving quality.

“The imperative is to be agile, at the speed of the business,” Taubert stresses. “The network has to be agile enough so as to not take too long for changes to happen; or to deploy new networks or devices. And to achieve this agility, the infrastructure will become more and more centralised.”

Customers are using Cisco’s networking and security in industrial automation environments solutions to connect various industrial and manufacturing systems, including distributed control systems (DCS); IACS devices with controller drives, sensors and I/Os; machines like package units, robots, human machine interfaces (HMIs), AGCs; supervisory control and data acquisition (Scada), manufacturing execution systems (MES), historians, asset management systems; and industrial IoT and artificial intelligence and machine learning (AI/ML) applications.

Typical use cases include industrial automation and control in the classic OT networking environment; as a foundation for IoT applications like predictive maintenance, digital twins or machine vision; for macro (industrial MDZ) and micro (cell/area zone) segmentation; to achieve visibility and continuous monitoring for IACS communication and devices; distributed edge computing infrastructure; and video surveillance.

The outcomes they want to achieve range from leveraging available knowledge, process and workflows; increasing operation efficiency through realtime processes visibility and simplified troubleshooting; reducing opex by making things easy to configure, upgrade, replace and maintain; making plant operations more reliable; and increasing security.

The networking technology needs to support these outcomes, so there are a number of technology requirements that are key, Taubert points out.

The first is a need for more network bandwidth. “All of the applications in the industrial space need more bandwidth. They are typically applications that use video, AGVs, thermal imagery, 3D sensors – and they all drive an increased need for bandwidth.”

At the same time, networks must offer low latency, resilient communications and rich data. “Once we have enough bandwidth, we need to consider the time we need to transport sensor data to central intelligence and how quickly it needs to get back. If we want realtime control, we have to think of latency considerations. We often rely on the network for safety purposes and realtime control, again speaking to latency.”

Cybersecurity is one of the most important considerations, considering the explosive growth in connected devices which increases the threat surface.

Simplified scale could be hard to achieve, especially combined with the need for more security, but Taubert says it’s vital for organisations to be agile as they deploy and manage more devices across more locations with the same resources.

Finally, they need to think about edge computing now even if they are not implanting it yet. “When compute is closer to the source, it’s possible to process it and act on it faster.”

There are three key pillars that will contribute to the success of an Industry 4.0 project, Taubert explains.

“The first is the IT and OT partnership. This doesn’t mean we are taking over one anothers’ responsibilities but, regardless of how it ends up on an organisational chart, we have to be open-minded as the partnership is crucial to success.”

Stakeholder buy-in from senior management is a must-have, he adds. “And value contribution is key. Without that, it’s a technology-forward approach and we seldom see projects that go past the initial proof of concept.”

From a security perspective, it’s vital to have visibility of all the assets and how they communicate. “We have to make all the sensors on the network visible as its crucial that we capture all the traffic.”

Partnering is vital, Taubert concludes. “Having experienced partners at our side is key. We are investing heavily to ensure that implementation partners are qualified and certified in IoT specialisation, that they are trained and experienced.”


The vital role of cybersecurity

Cybersecurity is one of the most important issues to consider when implementing Industry 4.0, but it is also the most common initiator of an organisation’s IoT transformation journey.

Taubert explains that often a security audit for insurance, or an exposure to an attack like ransomware, will kick-start an Industry 4.0 project.

“Security projects almost always come with some architecture development which drives further transformation.”

Security is often an issue in industrial environments because of antiquated, unpatched or legacy systems, he says. These systems are frequently not designed to be secure, for instance, lacking segmentation. There could also be gaps in the OT staff’s security knowledge, especially as IT and OT systems start to merge.

“In most cases, we currently do not have the right visibility into industrial networks,” Taubert points out. “What’s out there, who is talking to who, what are they saying? So, typically, this is the first thing any project needs to do. Without it you cannot continue with the next steps of access control and change control.”

The changing business needs mean that industrial systems are moving from automation islands to connected systems, so realtime information with no downtime and quick access is required.

Taubert believes that all cybersecurity programmes have some common key objectives: to mitigate risk from unpatchable systems; gain rapid visibility of IACS asset threats; get an up-to-date and complete inventory of IACS assets with details of each asset; protect the manufacturing platform from known and unknown threats; and limit network traffic when attacks happen.

A typical cybersecurity journey is a series of sprints, with each phase building on the one before, he adds. The phases are discover; baseline and segment; detect and enforce.