Historically the automation that manages industrial plants, trains, and utilities was developed separately from the automation that manages bank accounts, civil records, and online purchases. The main driver for the former has always been reliability; for the latter it was accuracy. Arthur van Schendel and Frank Breedijk explain why these opposites are now coming together and what the benefits of connecting the physical and administrative worlds are.
The two separate environments come with completely different cultures. Undisrupted reliability and availability of physical processes requires a closed fortress, with no interference or intrusion from a world where data accuracy matters most. “The impact of a failure in the primary process is much larger than, for example, the payroll database being offline for a while,” explains Arthur van Schendel, Managing Director. “Moreover, business leaders in the physical world are wary of the IT phenomenon of gradual improvement, where they experience outages, security threats, and massive IT projects that promise an end to all problems, even though they never really do this.”
On the other hand, the people at the IT departments don’t feel appreciated, because the operators of all those factories, trains and energy plants won’t allow them to help. And at the same time, IT experiences a lot of C-level pressure to provide progression and relevance. As a result, IT has learned to become more agile and the pressures on uptime are just as high as in the physical world.
Thanks to new opportunities in the field of data processing and analysis, the cultural divide between the physical and digital realms can now be bridged. The roadmap towards a smart factory respects mutual interests and leads directly to added business value.
At the moment, it is not uncommon for operational systems to run on outdated software—and that’s a big problem, says Chief Information Security Officer (CISO) Frank Breedijk. “It’s often impossible or even undesirable to update these machines. It does sometimes happen, to minimize the risk of disruptions, but this often forfeits the warranties. Security is thus handled by simply not linking a system to the outside world.”
However, contradictions between environments start to diminish wherever business activities become more digital. After all, in the online world, availability is of crucial importance. A retailer will lose a lot of sales if the website is down. “IT’s attitude towards uptime, security and flexibility has changed,” says Frank. “In the digital services industry, you want to get a new functionality to the end user as soon as possible, and it must work at all times.”
“Also, factories, transporters and utilities are digitizing, partly because of competitive pressure, changing markets and pressures on the supply chain,” Arthur adds. “So data from physical machinery is becoming more and more important. Combined with data from the transactional systems, this can provide extremely valuable insights for planning, maintenance, uptime of machinery and plant output. In a smart factory, processes and production lines can be optimized, making it much easier to respond to market demands. This can generate significant additional earnings or savings.”
But this often comes with a challenge. Old-fashioned enterprise IT is not designed for fast data access and processing. This requires a novel solution that uses the available data from both worlds—without affecting the underlying systems.
We can start up a proof of concept in a matter of days.
IoT Edge Kit
With the IoT Edge Kit, Schuberg Philis is making it possible to close the gap, step by step. Data from the machinery becomes accessible in a virtual representation of the system: the so-called digital twin. From the IT side, planning data is added. Together, this provides information that you can analyze in great detail, without the mission-critical systems being connected to the internet.
“We can read every PLC, no matter the maker, version and output,” Arthur explains. “On the factory side, nothing needs to be changed. Everything is translated into a usable data format and is accessible for the people it is relevant to. IT makes sure that the data is unambiguous and consistent across different systems, plants and countries.”
“With this smart factory solution, we can address the increased pressure from the business to optimize the supply chain without any back-and-forth intrusion. As a first step, the only thing we ask from both sides is the relevant data. We can start up a proof of concept in a matter of days, to show what is possible.”
Bridging the divide
The physical-digital cultural divide can be bridged in four stages: descriptive, reactive, automated and eventually predictive.
Arthur van Schendel: “Whatever you do, it shouldn’t come at the expense of availability. Step one is the descriptive stage, which is a process of discovery: get the relevant data out of the PLCs that manage the machinery, translate this into the right format, link it with data from the IT systems, and uncover the value hidden inside. These insights can lead to better planning, more efficient stock management, faster throughput, less waste, et cetera.” The system is initially set up to guarantee that no outside influences on the factory automation are possible.
“Once shop floor trust has grown, you make the step to the reactive stage: communicating insights from both environments back to the people involved,” Frank Breedijk continues. “Thanks to timely information or an alert for the operator, you can avoid errors in the production process or make adjustments whenever necessary.”
Arthur adds: “The following step is to automate the reactive stage. In case of a disruption, for example, a signal from IT will be sent back to the factory: ‘Stop the machine’ or something similar, to alert operators.”
According to Frank, the final, predictive stage is slowly gaining traction in the market, but it’s still far from common in production environments. “Making predictions is complicated: it often requires a substantial investment in time and money. The question is whether this is cost-effective compared to, for example, periodic replacement.”
The smart and ultimately predictive factory is an attractive proposition, especially with regard to planning. Arthur van Schendel: “If you can predict a peak in sales, you ramp up work in the factory. It allows for optimum materials procurement, improved manufacturing and assembly, and delivery of exactly the right product. In that scenario, you’re operating in the highest maturity phase—but we don’t see that very often yet.”
“It’s better to work together with the people on the work floor, as they can directly see the benefits for themselves and their managers,” adds Frank Breedijk. “By starting in a descriptive way, you begin by understanding the process and establishing relationships. And this also opens up new possibilities in other fields.”
Arthur van Schendel