No longer selling machines, but renting out production capacities - with the digital business models of the future, traditional mechanical engineering companies will become full-service providers for their customers. But even on a smaller scale, digital processes on the shop floor ensure more efficient workflows and higher levels of automation, which can become a decisive factor in holding one's own against the competition in times of economic turbulence. However, these advantages can only be achieved if digitalization is approached with planning and foresight. This also includes keeping a critical eye out for potential problems and risks - above all the classic stumbling blocks that repeatedly occur in mechanical engineering digitalization projects.

The use case as a decorative accessory

The early bird catches the worm and a lot helps a lot: those who think like this may already have their first data reservoirs waiting to be used and evaluated. But then comes the disillusionment: the data required for the intended purpose is either not available at all or not in the required form. Such a situation is not uncommon when concrete application scenarios are not seen as the starting point for digitization measures, but merely as a nice add-on that will emerge by itself at some point.

For successful projects, concrete and potential use cases must be the linchpin from the outset on which everything else is based. Which analyses and findings would provide a real advantage for business processes? What data is required for this and in what format must it be transmitted so that subsequent systems can also work with it? If questions like these are considered before a machine is retrofitted with additional sensors or a PLC developer gets to work, the likelihood of having a digital scenario at the end that can actually be used to continue working immediately can be significantly increased.

All or nothing

However, a use case that is too ambitious can also be a hindrance. One example: Anyone looking into the "digital twin" will quickly come across the media-effective scenarios of large industrial groups in which, for example, the optimum configuration for commissioning a machine is simulated in advance by a digital twin that takes all physically relevant factors into account and determines ideal temperatures or torques in this way.

In fact, launching such a complex scenario in a "big bang" is not very realistic for a medium-sized company. Instead, it is advisable to proceed in small steps: Instead of a complex simulation, a digital twin can initially be used in the form of a digital type plate for machines, for example. With the help of a QR code on the machine, employees on the shop floor have all the important information for commissioning or productive use at their fingertips, from training videos and data sheets to manuals, which significantly reduces the amount of research required. Scenarios like these quickly offer companies immediate practical benefits, while at the same time the creation of digital twins for all machines gradually creates the basis for implementing increasingly complex scenarios in practice.

In this case, however, it is crucial to rely on standards right from the start - in the context of the digital twin, for example, the so-called asset administration shell (AAS). This generally applies to all digitalization measures, including the gradual digitalization of individual process steps that are later to be combined into a large whole. Standards are the only way to ensure that the semantics used are understood in other contexts at a later date and can therefore be seamlessly integrated into new scenarios.

Lost in Translation

Speaking of understanding: the interface between classic IT development and machine programming can also lead to problems in practice, as it is not uncommon to find completely different approaches to development in both areas. While IT specialists work in their development environment, versioning and testing in it, PLC experts often continue to work without automated tests or version management and implement their changes directly on the machine. The language itself can also contain differences: for example, PLC developers usually start their counting in arrays with 1, while classic software developers start with 0. If deviations such as these are not taken into account accordingly, the experts in both areas may end up talking about completely different things without realizing it.

Both worlds can be brought together by companies promoting exchange and sensitizing their employees to possible differences. For example, if the IT developer looks over the shoulder of the PLC expert when programming the machine on site, this creates a better understanding on both sides: IT can better understand how a machine was specifically programmed, while at the same time there can be an exchange about which data streams from the machine should be passed on to which downstream systems and in what way.

Act now, secure later

It is certainly no secret that security must be the top priority in all these networking scenarios. Such concerns are often among the main obstacles that stand in the way of implementing the planned digitalization strategy. What if sensitive company data or valuable intellectual property falls into the wrong hands? What if cyber criminals paralyze production processes or steal critical data? In the worst-case scenario, incidents like these threaten the existence of the company.

So is it best to avoid digitalization altogether to avoid taking any risks? If you think this way, you risk being left behind by the competition, and thus also the success of the company. A solid security strategy therefore makes much more sense. If the necessary expert know-how is not available in-house, it is advisable to rely on external security specialists to help develop a solid security plan and avoid potential gateways for cyber criminals. This due diligence also applies in particular to supposedly "small" digitization steps that can be implemented quickly - because those responsible often do not even consider possible risks here.

Conclusion: digitization with care and foresight

Think from the use case, rely on standards, ensure exchange, consult experts - with these four central maxims, some of the most common stumbling blocks on the path to digitalization in mechanical engineering can be avoided. Nevertheless, specific pitfalls always depend heavily on the individual company and its circumstances, so measures should always be thought through critically and additional experts for the respective area should be consulted if necessary. Those who approach their digitalization efforts thoroughly and with foresight are doing important groundwork to achieve the results that are needed in mechanical engineering today: solidly networked processes and production assets that not only ensure greater efficiency, transparency and automation in day-to-day work, but also create the basis for the new, digital business models of the future.

Christian Heinrich, Digital Twin expert at Xitaso