MBA & Eng. Felix Wildemann has been with Beckhoff for just over ten years, nine of which he spent in product management for Industrial PCs. In 2017, as the responsible product manager, he founded the C60xx ultra-compact Industrial PC series, which is used for classic control tasks and edge applications in particular. In addition to hardware support, Wildemann also performs interface functions for software product management at Beckhoff in order to drive forward IoT and ML solutions on an interdisciplinary basis.

The Industrial Internet of Things is generating more and more data that needs to be managed. This data complexity needs to be mastered. What role does edge computing play in this?

Wildemann: Traditional automation is very cautious about opening up and networking to the outside world because the security and stability of the production plant could be called into question when networking with IT. Compromises are often necessary when networking both worlds directly.

Edge computing can be the key to mastering data complexity and uncompromising networking of IT and OT. This is because data is pre-selected, sorted and processed at the edge in order to filter the truly relevant process data. Although all edge functions can also be executed directly on the controlling industrial PC, the targeted decoupling and pre-processing of relevant process data at the edge brings decisive advantages. This is especially true in brownfield scenarios, i.e. when an edge device is integrated into an existing system concept: With the help of a dedicated, subsequently installed edge device, IoT functions can be brought into the existing application without the control unit having to be subsequently modified in terms of software or hardware. But even in the greenfield scenario, i.e. with completely new system concepts, embedding an additional, dedicated edge is worthwhile in many cases.

Sensor data is preselected at the edge, software applications are processed or even AI calculations are carried out. What should users consider with regard to the edge software landscape? Which technologies should/must be taken into account?

Wildemann: The basis for a holistic edge concept is a modular software kit that optimally combines IT and OT. Modularity can be understood here like an app concept - the functions actually required are selected and executed on the edge device from a large number of possible functions. No additional programming knowledge is required because users can access ready-made modules and integrate them quickly and easily.

It is essential that the edge device speaks the language of the machine, but also provides secure, standardized IT protocols and mechanisms that enable an indirect, secure connection of the machine to the cloud or local servers via the edge device. In this context, international standards for interfaces and protocols are important so that all the different parts of the system can be brought together for a holistic edge concept.

Beckhoff Automation realizes open automation systems on the basis of PC-based control technology.

© Beckhoff / Jürgen Rehmann

Beckhoff provides a wide variety of blocks and protocols for this with the TwinCAT control software. For example, data can be recorded on the machine side via the TwinCAT ADS protocol or via OPC UA in the Edge without the need to adapt the TwinCAT program code of the control computer. On the OT side, the protocols of common cloud providers and private cloud scenarios are supported in addition to standard IT protocols.

The supreme discipline in the edge is then the 'real coupling' of edge and controller: it is not enough to collect data from the controller and make it available. The software in the edge must be designed in such a way that it does not impair the availability and real-time capability of the control system. Nevertheless, relevant, processed process data must also be able to be fed back into the control system in order to retroactively optimize the system, achieve predictive maintenance and minimize machine downtimes.

Container technologies, such as Docker, are also suitable for accommodating all desired functions on a single device. Docker makes it possible to channel and optimize the execution of edge software in order to save resources, but also to integrate non-system apps and functions directly into the edge device. This means that customer-specific functions can also be mapped in the Edge and operated in parallel with TwinCAT, for example.

The hardware requirements grow with the tasks. What should users look out for when selecting the right edge computing hardware?

Wildemann: An edge solution only works if the hardware as well as the software is optimally aligned with the IoT and control approaches. This requires a scalable portfolio of compact, modular and versatile mountable Industrial PCs of different performance classes that can always be perfectly adapted to the respective edge scenario - depending on the required computing power, available installation space and required interfaces.

Beckhoff's approach here is that the Industrial PCs with TwinCAT software can be used both as control computers and as edge devices. This has the unbeatable advantage that the hardware has the important characteristics of a classic Industrial PC, such as long-term availability, robustness and durability. At the same time, industrial bus systems can be directly integrated so that various sensor data can be collected in the field directly on the Edge in addition to the data available in the controller. With Microsoft Windows and TwinCAT/BSD, two operating systems are available that support Beckhoff's own TwinCAT control software as well as all IT protocols and Docker functions.

This means that users can access the entire Beckhoff Industrial PC portfolio without being limited to the selection of specific hardware. Only the modularity, scalability and flexibility of the portfolio ultimately enables the realization of a perfectly fitting, holistic control, IoT and edge solution.