Birgit Vogel-Heuser

Jonghan Lim, Birgit Vogel-Heuser, Ilya Kovalenko

Traditional manufacturing faces challenges adapting to dynamic environments and quickly responding to manufacturing changes. The use of multi-agent systems has improved adaptability and coordination but requires further advancements in rapid human instruction comprehension, operational adaptability, and coordination through natural language integration. Large language models like GPT-3.5 and GPT-4 enhance multi-agent manufacturing systems by enabling agents to communicate in natural language and interpret human instructions for decision-making. This research introduces a novel framework where large language models enhance the capabilities of agents in manufacturing, making them more adaptable, and capable of processing context-specific instructions. A case study demonstrates the practical application of this framework, showing how agents can effectively communicate, understand tasks, and execute manufacturing processes, including precise G-code allocation among agents. The findings highlight the importance of continuous large language model integration into multi-agent manufacturing systems and the development of sophisticated agent communication protocols for a more flexible manufacturing system.

Kamil Rosiak, Alexander Schlie, Lukas Linsbauer et al.

Automated production systems (aPS) are highly customized systems that consist of hardware and software. Such aPS are controlled by a programmable logic controller (PLC), often in accordance with the IEC 61131-3 standard that divides system implementation into so-called program organization units (POUs) as the smallest software unit and is comprised of multiple textual and graphical programming languages that can be arbitrarily nested. A common practice during the development of such systems is reusing implementation artifacts by copying, pasting, and then modifying code. This approach is referred to as code cloning. It is used on a fine-granular level where a POU is cloned within a system variant. It is also applied on the coarse-granular system level, where the entire system is cloned and adapted to create a system variant, for example for another customer. This ad hoc practice for the development of variants is commonly referred to as clone-and-own. It allows the fast development of variants to meet varying customer requirements or altered regulatory guidelines. However, clone-and-own is a non-sustainable approach and does not scale with an increasing number of variants. It has a detrimental effect on the overall quality of a software system, such as the propagation of bugs to other variants, which harms maintenance. In order to support the effective development and maintenance of such systems, a detailed code clone analysis is required. On the one hand, an analysis of code clones within a variant (i.e., clone detection in the classical sense) supports experts in refactoring respective code into library components. On the other hand, an analysis of commonalities and differences between cloned variants (i.e., variability analysis) supports the maintenance and further reuse and facilitates the migration of variants into a software product line (SPL).

Fandi Bi, Birgit Vogel-Heuser, Litong Xu

Complexity of products, volatility in global markets, and the increasingly rapid pace of innovations may make it difficult to know how to approach challenging situations in mechatronic design and production. Technical Debt (TD) is a metaphor that describes the practical bargain of exchanging short-term benefits for long-term negative consequences. Oftentimes, the scope and impact of TD, as well as the cost of corrective measures, are underestimated. Especially for mechatronic teams in the mechanical, electrical, and software disciplines, the adverse interdisciplinary ripple effects of TD incidents are passed on throughout the life cycle. The analysis of the first comprehensive survey showed that not only do the TD types differ in cross-disciplinary comparisons, but different characteristics can also be observed depending on whether a discipline is studied in isolation or in combination with others. To validate the study results and to report on a general consciousness of TD in the disciplines, this follow-up study involves 15 of the 50 experts of the predecessor study and reflects the frequency and impact of technical debt in industrial experts' daily work using a questionnaire. These experts rate 14 TD types, 47 TD causes, and 33 TD symptoms in terms of their frequency and impact. Detailed analyses reveal consistent results for the most frequent TD types and causes, yet they show divergent characteristics in a profound exploration of discipline-specific phenomena. Thus, this study has the potential to set the foundations for future automated TD identification analyses in mechatronics.

Alexander Weigl, Mattias Ulbrich, Suhyun Cha et al.

A wide range of interesting program properties are intrinsically relational, i.e., they relate two or more program traces. Two prominent relational properties are secure information flow and conditional program equivalence. By showing the absence of illegal information flow, confidentiality and integrity properties can be proved. Equivalence proofs allow using an existing (trusted) software release as specification for new revisions. Currently, the verification of relational properties is hardly accessible to practitioners due to the lack of appropriate relational specification languages. In previous work, we introduced the concept of generalised test tables: a table-based specification language for functional (non-relational) properties of reactive systems. In this paper, we present relational test tables -- a canonical extension of generalised test tables for the specification of relational properties, which refer to two or more program runs or traces. Regression test tables support asynchronous program runs via stuttering. We show the applicability of relational test tables, using them for the specification and verification of two examples from the domain of automated product systems.

Lorenzo Sabattini, Valeria Villani, Julia N. Czerniak et al.

With the increasing complexity of modern industrial automatic and robotic systems, an increasing burden is put on the operators, who are requested to supervise and interact with such complex systems, typically under challenging and stressful conditions. To overcome this issue, it is necessary to adopt a responsible approach based on the anthropocentric design methodology, such that machines adapt to the humans capabilities. Moving along these lines, a methodological approach called MATE was introduced in [1], which consists in devising complex automatic or robotic solutions that measure current operator's status, adapting the interaction accordingly, and providing her/him with proper training to improve the interaction and learn lacking skills and expertise. In this paper we propose an evaluation and validation procedure to guarantee the achievement of the requirements of a MATE system.

Valeria Villani, Lorenzo Sabattini, Julia N. Czerniak et al.

Modern manufacturing systems typically require high degrees of flexibility, in terms of ability to customize the production lines to the constantly changing market requests. For this purpose, manufacturing systems are required to be able to cope with changes in the types of products, and in the size of the production batches. As a consequence, the human-machine interfaces (HMIs) are typically very complex, and include a wide range of possible operational modes and commands. This generally implies an unsustainable cognitive workload for the human operators, in addition to a non-negligible training effort. To overcome this issue, in this paper we present a methodology for the design of adaptive human-centred HMIs for industrial machines and robots. The proposed approach relies on three pillars: measurement of user's capabilities, adaptation of the information presented in the HMI, and training of the user. The results expected from the application of the proposed methodology are investigated in terms of increased customization and productivity of manufacturing processes, and wider acceptance of automation technologies. The proposed approach has been devised in the framework of the European project INCLUSIVE.