Micro- and nanostructures influence the optical, mechanical, haptic and biological properties of surfaces. For example, plastic surfaces in cars have a better grip and higher quality if they are provided with microstructures, and the right surface treatment can significantly reduce air resistance in engine components.

In this context, laser structuring is of particular interest as a comparatively new and promising approach to producing structures on free-form surfaces. In this process, a focused, pulsed laser beam is guided quickly and precisely over the component surface. Compared to conventional processing methods such as coating or etching, laser structuring proves to be more environmentally friendly and precise and also offers an extended range of design options.

However, the creation of large-scale micro- and nanostructures poses a considerable challenge in computer-aided manufacturing (CAM) systems. This is because massive volumes of data are generated, especially if traditional path planning algorithms are to be used. For large components - for example, a turbine blade 5 m long and 0.4 m wide - such a network would contain around 8 x 1016 surfaces and require 8 terabytes of memory in STL format. This represents an almost insurmountable hurdle for modern computing systems.

Efficient path planning

The procedural representation of a functional structure avoids the need for such an extensive triangular network and enables far more efficient path planning. Here, the structural data is defined using mathematical formulas, equations and algorithms. These descriptions allow a precise and systematic design of surface structures without having to rely on image or pixel-based representations. Procedural surface structures also generate more compact data sets and prevent any distortions that could occur when using grayscale images due to scaling or transformations.

Another advantage of procedural structures is that they allow the surface structures to be calculated in real time during the process. This is particularly valuable in applications where quick adjustments are required.

Robot-assisted work

Although existing laser structuring systems work with high precision and deliver good results, they are restricted in terms of the size of the components to be processed due to their limited working area. In order to provide large surfaces with the procedural structures described, large, high-precision systems are required, which is associated with correspondingly high investment costs that not every company can or wants to bear.

The Fraunhofer IPT has therefore developed a process in which an industrial robot is used to structure large-area components with the laser. The system-related inaccuracies of the robot are compensated for by a newly developed intelligent laser structuring head that automatically detects and corrects positional deviations. This tool has a two-dimensional positioning device with high-precision motors to ensure the required accuracy for micro- and nanostructures.

In an industrial context, the laser scribing head is just one of many tools that is replaced as required. Currently, the tool change is done manually, with the configuration of the software requiring considerable time and effort. This includes ensuring the compatibility of different control software, integrating the laser head controller into the robot controller and setting up additional interfaces.

Microservices as a basis

Architecture of microservice-based laser structuring using procedural structures.

© Fraunhofer IPT

To overcome the challenge of calculating procedural structures and controlling robot-assisted laser structuring processes, the Fraunhofer IPT opted for an approach based on microservices. Microservices are mainly used in large server systems to operate and orchestrate applications such as websites, web services, email servers or databases quickly and flexibly. The services are divided into the smallest components and executed in separate containers, which enables a high degree of modularity of the services.

This approach is interesting for future production technologies that require an increasing number of individualized products: These are accompanied by the need to flexibly adapt production machines and make production processes adaptable. However, automation technology has strict requirements for industrial applications, both in terms of data processing platforms and communication. Both require real-time capable, fail-safe, highly available and low-latency technologies in order to meet the required quality parameters.

For this reason, the Fraunhofer IPT has developed a real-time microservice architecture tailored to the requirements of the industry, which enables the prioritization of containerized applications and thus the separation of critical and non-critical workloads. This allows real-time critical applications to be run on the same platform as non-critical applications - leading to considerable resource efficiency, as the computing resources are only used when they are actually needed.

Kubernetes is the foundation

The architecture is based on Kubernetes, an established technology in the IT world for orchestrating a large number of applications. Kubernetes enables the automated deployment, scaling and management of containers in a dynamic environment. This enables the efficient execution and coordination of microservices and applications in a cluster. In order to use Kubernetes in an industrial environment, specific adaptations are required to ensure greater reliability and real-time capability. The Kubernetes cluster at the Fraunhofer IPT was therefore configured in such a way that real-time-critical applications are initially isolated from non-critical applications through CPU pinning and prioritization at process level.

Real-time adaptation of the process

Microservice-based laser structuring using procedural structures opens up new possibilities in photonic production, as the use of procedural structures allows large-area micro- and nanostructures to be generated efficiently and at process runtime. This enables real-time customization of the laser structuring process. The development of a customized real-time microservice architecture based on Kubernetes shows how future production technologies can benefit from flexible and highly available solutions.