Factory automation

As an international basic standard, IEC 61508 provides the framework for the functional safety of electrical, electronic and programmable systems in factory automation. It covers the entire safety lifecycle, from risk analysis through development to operation and maintenance, and thus supports the design of sensors, controls and actuators for defined Safety Integrity Levels (SIL). Several trends are emerging in semiconductor solutions: microcontrollers and systems-on-chip integrate more capable diagnostic functions, redundancy and safe monitoring units. At the same time, specialised safety microcontrollers with lockstep cores, error correction and integrated self-tests are gaining importance. The increasing connectivity of machines is also driving secure industrial Ethernet interfaces and security functions, as cybersecurity and functional safety move closer together. In parallel, certified software libraries, operating systems and development tools make standard-compliant implementation easier. Artificial intelligence and edge computing are being used increasingly, but they pose new challenges for safety evidence. The announced third edition of IEC 61508 therefore addresses IoT, AI and complex semiconductor architectures more strongly, among other things.

Cybersecurity in factory automation has long since ceased to be a purely IT topic; it is a basic prerequisite for stable processes, secure remote maintenance and the protection of connected production plants against manipulation, failures and sabotage. Security functions are increasingly being considered directly in architecture, development and integration; this includes security by design, segmented OT/IT networks, continuous monitoring and AI-supported anomaly detection. Hardware-based protection concepts that secure authenticity and integrity at component level are also particularly relevant. This means that secure element chips are gaining importance: they serve as a trusted root for identities, store key material in a tamper-protected way and support secure authentication of machines, sensors and gateways. The trend here is towards smaller, energy-efficient and more easily integrable devices that fit seamlessly into industrial IoT and edge environments. Cybersecurity in factory automation is thus shifting increasingly from retrofitted protection to an embedded system property.

The use of AI in process control and predictive maintenance is increasingly becoming the link between real-time data, stable processes and falling downtime costs. The current trend is towards semiconductor solutions that bring AI functions closer to the machine: edge TPUs accelerate specialised inference directly at the edge of the network, NPUs handle more complex models in industrial PCs, gateways and embedded systems, and modern MCUs integrate AI elements for energy-efficient condition monitoring in sensors and actuators.

Particularly important here is the shift of analysis and decision-making into production itself, in order to reduce latency and avoid having to send data streams to the cloud permanently. At the same time, requirements are growing for robust connectivity, interoperability and cyber-resilient architectures, so that AI-supported control becomes not only faster but also more reliable. In practice, this means more predictive maintenance, more precise process management and higher plant availability with lower energy and service costs at the same time.

Rising energy costs, stricter climate targets and pressure for higher productivity are bringing the topic of energy efficiency ever more into focus in industrial manufacturing. WBG semiconductors such as SiC and GaN are therefore gaining importance, as they make drives, power supplies and converters more efficient, more compact and thermally more robust than classic silicon solutions. In practice they mainly enable energy-efficient motor drives, higher switching frequencies, smaller cooling and passive components, and better power density. Important technology trends currently include the increased use of SiC in powerful, high-voltage drives, the growing use of GaN in compact, high-frequency part-load and auxiliary systems, and integration into modular, scalable drive concepts. In addition, there is a clear trend towards greater system efficiency along the entire automation chain, from the power supply to the servo drive.

Sensors are the sensory organs of factory automation: they capture physical quantities such as position, pressure or temperature and make this data available to control systems in real time. Without them, precise process management, quality control and predictive maintenance would not be possible. At present, development is shaped above all by intelligent sensors with integrated signal pre-processing and communication. Modern smart sensors combine sensor element, analogue-to-digital converter and microcontroller in one housing and handle tasks such as filtering, plausibility checking or self-diagnosis directly at the measuring point. This reduces data volumes, increases robustness against interference and enables decentralised decision logic close to the process. In parallel, IO-Link is establishing itself as an open, manufacturer-independent point-to-point standard that transmits both process data and parameter and diagnostic information bidirectionally over a simple three-wire connection. IO-Link allows end-to-end parameterisation, fast device replacement scenarios and the integration of sensors into higher-level IIoT and condition-monitoring concepts. Together, intelligent sensor technology and digital fieldbuses such as IO-Link lead to a more connected, data-driven factory in which sensors not only measure, but actively contribute to the efficiency, flexibility and transparency of production.

Chiplets and heterogeneous integration are seen in factory automation as the key to more flexible, more compact and more powerful control systems. Chiplets are small, specialised semiconductor building blocks that are assembled modularly into an overall system, instead of integrating everything into one large monolith. Heterogeneous integration means combining different, separately manufactured devices such as logic, memory, sensors or power electronics into a shared assembly. This is attractive for automation because controls can be configured more specifically and adapted more quickly depending on the machine task. Central trends include advanced packaging, 2.5D and 3D integration, energy-efficient edge processing, and closer coupling of AI, IIoT and digital twins in industrial control architectures. The development is thus aimed at modular systems that combine higher computing power, better scalability and shorter development cycles.

Machine vision is increasingly becoming a central building block of modern production processes: systems with cameras, lighting and evaluation software inspect components in real time, support robot guidance and assure quality directly on the line. The trend is towards more intelligent applications in which AI-supported evaluation, edge processing and high-resolution image sensors work together. This makes it possible to handle complex surfaces, changing products and difficult lighting conditions more reliably. At the same time, 3D vision, multispectral methods and the connection to IIoT and automation platforms are gaining importance, because image data is then not only captured but can also be used across production.

At the heart of machine vision solutions are CMOS-based image sensors, which today characterise the majority of industrial cameras, as well as powerful processors and specialised AI accelerators for evaluation at the edge. Added to this are memory devices and fast interface ICs that transport and buffer large volumes of image data reliably.