PROFINET works smoothly in many applications every day. Nevertheless, we have to start thinking today about how future technologies can be integrated into this network in a practical way, for example, edge applications, 5G, or innovative operating concepts from augmented reality.
The classic automation pyramid with its strict hierarchical structures is already beginning to open up. Data not only flows between levels but is also collected directly from field devices, controllers, or control system components. Here, the boundaries between IT and OT are disappearing. Incidentally, this task is not only technological but also interpersonal because, by the same token, employees from IT and automation technology have to learn each other’s challenges and technical lingo. In the process, however, opportunities are also arising: If a functional OT network infrastructure with continuity to IT is available, this infrastructure can be an enabler for interesting business ideas.
Digitalization Means New Data Networks
The conditions, however, must be right. What is needed for such a convergent network? Above all, the data must be networked from sensors, devices, machines, and production cells. Simultaneously, higher-value information needs to be generated from this data by enriching it with semantics.
However – and it cannot be stressed enough – a benefit can only be derived from this data if such networking happens quickly, flexibly, and easily. Concretely, this means: Additional engineering steps should be avoided, and a mix of wired and wireless communications must be possible.
In the process, the basic requirements and structures of the industry for industrial communication properties, such as defined latency (real-time), speed, synchronization, freedom from loss and redundancy, remain, as well as the fact that change, expansion, or reduction may not affect the existing concepts. Users do not want to deviate from proven solutions, and they certainly do not want to redesign. What remains is a difficult balancing act to master in the future.
What needs to be done?
Here, PI has been expanding its specifications for some time, for example, in regard to OPC UA, APL, wireless applications and security, or MQTT for cloud communication and 5G.
In the process, it has become clear again and again how much detailed work is required for this task. Network functions are needed to guarantee a secure transfer of data when object modeling, and so new functions and protocols are integrated on the one hand.
On the other hand, data modeling takes on a new meaning in this context. Data must be properly prepared in order to be processed using machine-readable information models to assign semantics to data. Only when these two areas are meaningfully combined and standardized are consistency and the greatest possible benefit possible.
Practical Integration
Here, a successful example is the combination of OPC UA and PROFINET. With OPC UA as middleware, data and information models can be accessed in an object-oriented manner. With its application profiles, PROFINET virtually assumes the supply of the necessary information.
The fact that this approach works in practice has already been shown via proof-of-concept. A variety of different use cases, from controller-to-controller mapping to energy management to the NOA (NAMUR Open Automation) connection of actuators, show that a broad range of applications are covered. PROFINET’s openness thus makes it possible to also add more complex sensors or devices with an OPC UA interface that send their values directly to corresponding cloud services or edge gateways without needing to tediously reconstruct the automation solution.
Process automation is also breaking new ground. In addition to NOA (NAMUR Open Architecture), the definition of information models (e.g., PA DIM or PA Profile 4.0), and integration technologies such as FDI, only the communication part of the industry’s path to digitalization was missing in principle. That time has now come with Ethernet-APL. In contrast with the hierarchically structured system environments of today’s processing systems, Ethernet-APL allows unhindered access to all process and device data down to each individual field device – even in explosion-proof areas. And it works, as shown in test systems operated by BASF and Bilfinger Maintenance. With the combination of our PA Profile 4.0, Ethernet-APL, and OPC UA, many building blocks are coming together to form an overall picture – the path to digitalization in the process industry is paved.
Robust Networks Required
The topic of TSN (Time-Sensitive Networking) will also play a decisive role in the future. Among other things, TSN enables a convergent and robust use of a common Ethernet network for IT and OT applications. Its advantages include higher bandwidths, determinism, flexible network configuration, and large chip variety. Meanwhile, many additional components can be integrated into a system, such as measuring devices for quality data acquisition. Such actions, or even a virus scan or firmware update, stress the network, as standard Ethernet (TCP/IP) contains very large data packets and is not real-time capable. Should larger TCP/IP data packets be sent, they can block a communication path, making it impossible for real-time data packets to pass them. This not only delays the delivery of data packets in each switch but results in a downright traffic jam in the device. PI is therefore working on not only integrating TSN but also making this new path as simple as possible for users while remaining flexible for future Industry 4.0 use-cases.
Higher Requirements on Security
New thinking is needed when it comes to security. Previous measures, which relied primarily on isolating production plants, must be supplemented with new concepts that make provision for the protection of components and communication. For this reason, the previous measures are being safeguarded with further-reaching protective measures, including the use of security certificates, inherent security in PROFINET components, and end-to-end security for PROFINET communication as a configuration option.
Security, however, is a topic that must be continuously adapted to current developments and, as a result, is never finished. This applies in particular with respect to the increasing networking of production plants. In the future, questions as to whether a driverless transport system or a mobile robot may join the network or which read/modify rights a communication partner may have will therefore have to be clarified.
Feedback Desired
The examples of OPC UA, Ethernet APL, and TSN clearly show how proven PI technologies can meet new communication requirements, including the diverse profiles for application-specific tasks and the simplest diagnostic options. What is perhaps even more important in everyday life, however, is that the infrastructure and the associated ecosystem are also right, such as the design of connectors or cables, guidelines for setting up secure networks, or dealing with security requirements.
Without a doubt: Digitalization in automation requires networks to eliminate rigid structures, make connections more flexible, provide more data and be more open. In the future, however, just one solution will no longer cover all eventualities. Instead, agile action and small steps will be required, including trying things out or “just doing it!” In the process, proofs-of-concept must be used again and again to reassure users and manufacturers about what is actually important, necessary, and feasible. In this context, it is beneficial to not only remain open in regard to technological interfaces but also to look beyond a company’s own horizons and actively participate in cooperations and committees.
Xaver Schmidt