How to unlock the technical barriers to digitalization
Industry 4.0 represents a shift from ‘islands of automation’ to cyber-physical systems
Four industry experts came together at the 5G Manufacturing Forum to offer insight into how some of the biggest barriers effecting digitalization — like the changing role of data, legacy integration and workforce management — can be addressed.
As background, Ted Rozier, director of engineering at Festo, explained that Industry 4.0 represents a shift from “islands of automation” to cyber-physical systems that integrate computation, networking and physical processes. “All of sudden, we don’t have these blinders and we now have the ability to capture information, and when we put the right tools into place, we know what to do with that information. We can change the process of the complete manufacturing floor,” he said.
Data: Along for the ride
This integration of computation, networking and physical processes allows manufacturers to develop a robust data strategy for preventive and predictive maintenance, which lets them exist in a state of maintenance urgency rather than emergency, according to KORE Wireless’ Director of IoT Channel and Business Development Adam Cohen.
“Real-time data … helped shift the industry so much that there are new types of as a service [offerings] … this opens up a whole new area of services where predicative maintenance and monitoring … actually rule [the] whole business opportunity,” he added.
While data is king in a digitally transformed enterprise, Suresh Daniel, the data and architecture integration director at Coats Plc, highlighted the importance of ensuring that the data itself also undergoes a transformation to avoid bringing “data that’s not useful” along for the ride. “Organizations … move from era to era, but they take the old data with them so … you will not see good results if you don’t transform your data as well,” he claimed.
Change management: ‘Singing from the same sheet music’
Equally critical, said Daniel, is change management, which refers to how an enterprise approaches communicating and facilitating a company-wide move from the industrial era to the digital era. “You can’t pivot overnight,” he reasoned. “You have to take the whole organization on the journey.”
This means that the messaging around digitalization and the direction of the organization has to be clearly and consistently articulated to every employee. Without effective change management, Daniel cautioned, you can’t expect to materially change the organization. “If someone is still stuck in the industrial era, your adoption will drop. You will not get the ROI you need because the people are not transforming,” he said.
Rozier agreed, commenting: “Industry 4.0, digital transformation … in the end, there is a people side of it. When we talk about smart manufacturing, it touches everyone from the CEO all the way down to the operator,” he said, adding that, therefore, it’s important to break down the silos that exist in how different elements — data management, engineering, supply chain logistics, etc. — are taught. “A lot of these topics are siloed [and] taught in different classrooms,” he said. “Bring all of those disciplines together so that they can see what… a complete system look[s] like.”
Rozier also discussed the “pesky skills gap,” anguishing that it “keeps getting bigger and bigger.” The solution, he said, is in “trying to get everyone to sing from the same sheet music.”
“An engineer… loves to design amazing things, but no wants to document the ‘aha’ moment. If you can get several organizations to document [those] moments and bring that into the classroom, everyone is singing from the same sheet music and that is a pull up globally,” he continued. He said, too, that organizations must also support its incumbent workforce by allowing them to get a certification or credential around smart manufacturing so that they can keep up with the changing industry.
System integration: From dumb machines to smart factory
While questions around legacy system integration are some of the most common, Cohen was clear on the answer: It’s not really a barrier.
“When you ask, ‘can it really integrate into legacy machines,’ that’s I think the best part about IoT; [It] is that it’s the anti-tower of babel, it speaks every single language,” he said, explaining that an array of smart devices can act as “teeny translators” for every type of input or output device, even those that don’t “speak,” like a motor or an electrical line.
In practice, this might look like adding an IoT sensor to a large, “dumb” manufacturing machine built in the 80’s. Now, data about the machine’s condition — whether it’s on or off, its temperature, the amount of energy it’s consuming, etc. — can be gathered for use by the company. This increased visibility, Cohen said, will lead to an increase in product production, less down time and manufacturing line flexibility.
Daniel further shared that for Coats, integration came down to plumbing. Likening the process to building pipelines, he said that the company has a team dedicated to looking at each facility process and building connections between them. Agreeing with Cohen, he added that yes, legacy components do all have a way to connect to the new system, but they may do so differently. But what’s important it so make those connections as real time as possible. Once connected, the data from the legacy systems is funneled into the pipeline.
“Then the pipeline will then keep moving and ingesting data into different layers,” Daniel concluded.
While these transformations are undoubtably exciting because of what they can deliver and enable, Bhupesh Agrawal, director of private networks and edge computing in Intel’s Networks and Edge Group, was careful to point out that challenges remain.
“When we look at the use cases, especially in the context of industrial 4.0, we also have to align them with the TCO [total cost of ownership] and the ROI,” he argued. “Some of the use cases like factory automation where we talk about the closed-loop applications or robotics — Do you know how much it costs how much to put an ethernet LAN? It costs almost $100 to $200 on a per foot basis … it is costly, especially when you are talking about factories that are several acres.”
Further, while existing Wi-Fi solutions can provide the mobility needed for some smart manufacturing applications, like automated guided vehicles (AGVs), they fail to deliver the necessary reliability. And while advanced cellular technology is proving to be a good substitution here in the form of private networks, Agrawal questioned whether enterprise IT teams are ready to adopt an entirely new technology just yet.
“When we look at these use cases, we have to look at what is ready right now, what the TCO looks like and what is the ROI for it,” summarized Agrawal. “We, as an industry, are still in a market creation phase [of Industry 4.0] and it’s going to take some time to be really ready and really get the benefits.”