Combining system knowledge with deep learning for fault diagnosis in industrial applications offers the potential to reduce the dependency of deep learning algorithms on extensive labeled datasets. However, existing methods often rely on highly specialized, problem-specific knowledge or demand detailed physical insights into the system, which limits their generalizability. Additionally, inconsistencies in knowledge representation hinder the ability to compare and build upon prior approaches. In this work, we address these challenges by leveraging commonly available knowledge about the phase structure of systems and the hierarchical organization of condition spaces. This information is systematically represented using knowledge graphs (KGs) based on the Resource Description Framework (RDF). To integrate this knowledge into deep learning, we transform the input data and the corresponding labels based on the KGs, and employ a graph neural network (GNN) trained with a semantic loss function informed by the knowledge about the condition space. The proposed approach is evaluated on three diverse datasets with varying characteristics under the two scenarios of domain generalization and novel fault detection.

The navigation of AMRs (Autonomous Mobile Robots) within factories, especially when facing moving and changing obstacles, still proves to be a challenge. ROS2 offers promising open-source behaviors that facilitate easier programming. When trying to implement reinforcement learning algorithms that aim at providing more robust policys, one is still faced with a difficult stack of technological layers. We provide an exemplary setup that shows the interplay between ROS2 (Robot Operating System 2) Humble and Reinforcement Learning. This setup works with different robot simulations, and we investigate how well a policy trained on one robot can be transferred to another.

In the research area of capabilities and skills, the standardization of terms and models is progressing in working groups of Plattform Industrie 4.0 and Industrial Digital Twin Association. At the same time, there is an increasing amount of publications that use capability models for purposes such as automated selection of capabilities or planning of sequences. However, there is currently no concept and formal description of effects generated by resources when acting on products. Having such a formal effect definition is needed to express both required effects as well as effects achieved by capabilities in order to compare requirements with possible solutions to a desired production step in an automated manner. This article introduces a formalism that can be used to define effects in the context of capabilities. An effect is seen as the change between an initial and a target state, which can both be specified using properties. The formalism is used to express effects in three different manufacturing application examples. Furthermore, an implementation of the formalism using the Web Ontology Language is provided. This ontology is used to model effects in a machine-interpretable way and to automatically compare effects for compatibility.

Machinery fault diagnosis is increasingly reliant on data-driven algorithms, yet struggles with adapting to unseen operating conditions. To address this, we propose integrating phase information into heterogeneous graphs for fault diagnostics with Graph Neural Networks (GNNs). Our method involves identifying the distinct phases that a machine undergoes within a cycle and segmenting the signals accordingly. These segmented signals are then represented in a graph of multiple connected sensor networks with diverse node and edge types. Prior to graph classification with a GNN, individual Convolutional Neural Networks (CNNs) preprocess the node attributes to account for their unique characteristics. The evaluation in a domain adaptation setting demonstrates the effectiveness of our approach, offering insights into improving the robustness and domain adaptability of fault diagnosis models.

This paper presents an approach to describe abilities of manufacturing resources by a formal description of capabilities using Semantic Web technologies. A hierarchical ontology architecture is proposed to represent, publish, and extend knowledge on capabilities for different application domains and use cases. Furthermore, the paper describes patterns of how the underlying formal logic can be used in taxonomy modeling and the inference of implicit capability facts. The usability and performance of the approach was validated by formalizing capability knowledge of related work and evaluated in benchmarking a prototypical implemented tool for managing and querying catalogs of resources and their capabilities. The proposed concept is intended to be used as a foundation for a future multi-layered feasibility checking, which evaluates the compatibility of resources and their offered skills with the requirements of manufacturing tasks at symbolic and subsymbolic levels. Extended evaluations might be based on parameters, analytics, simulation, and other means.

We show how to use the Operational Space Control framework (OSC) under joint and Cartesian constraints for reinforcement learning in Cartesian space. Our method is able to learn fast and with adjustable degrees of freedom, while we are able to transfer policies without additional dynamics randomizations on a KUKA LBR iiwa peg-in-hole task. Before learning in simulation starts, we perform a system identification for aligning the simulation environment as far as possible with the dynamics of a real robot. Adding constraints to the OSC controller allows us to learn in a safe way on the real robot or to learn a flexible, goal conditioned policy that can be easily transferred from simulation to the real robot.

We present a way of modelling device functionality in an OPC VA information model, leveraging the OPC UA programs specification. This allows for tool-independent orchestration of distributed and heterogeneous devices, while legacy devices can easily be included into the system by creating an OPC VA wrapper according to our specification1.

Skill-based engineering is gaining attention as a means to increase flexibility and changeability in engineering industrial automation systems. This paper proposes the Skill-based Engineering Model (SEM), which formally describes the core entities that play a role in skill-based engineering. Accordingly, we propose a four dimensional classification scheme for skills, and assess the suitability of a property model and OWL ontologies to describe and match skills. For each case, we identify a list of challenges that must be addressed to make skill-based engineering a reality in industrial automation systems. We believe that this paper can guide researchers to study various open aspects of skill-based engineering to make it feasible in complex industrial automation systems.