Conference Agenda

Abstract: With the emergence of large language models such as ChatGPT, the use of artificial intelligence (AI) has become firmly established in both private and business contexts. The use of AI applications is now considered a decisive factor for companies’ competitiveness. Logistics and supply chain management (SCM) offer promising areas of application in this context. Small and medium-sized enterprises (SMEs), in particular, face the challenge of identifying suitable fields of application and successfully integrating AI into their existing business processes due to limited resources.

This paper aims to develop an AI implementation model that supports SMEs in identifying suitable application areas in logistics and SCM and adopting them in a structured way.

First, expert interviews are conducted to identify requirements for an AI implementation model for SMEs in SCM and logistics with regard to technology, organization, and environment. Subsequently, an extensive literature analysis is undertaken to compare existing AI and digitalization implementation models against these requirements. The analysis shows that although many models exist, they often lack a specific focus on AI and do not sufficiently address the previously identified requirements.

Based on these findings, an AI implementation model is developed that systematically integrates organizational conditions and implementation measures considering the environment SMEs are operating in. The model structures the implementation process into six phases with several sub-steps, which are carried out iteratively. Finally, the AI implementation model is validated with practice partners.

The model presented in this paper supports SMEs in SCM and logistics by providing a practical, phase-based approach to AI implementation. It supports organisations in evaluating AI use cases, prioritizing them, and implementing them.

Cutting stock problems are still considered within the class of NP-hard problems that continue to be actively studied in the field of combinatorial optimization. These problems encompass processes such as material utilization and waste generation, and their effective resolution plays a crucial role in managing operations that directly influence efficiency. From a sustainability perspective, the reduction of unused waste directly contributes to lowering environmental impacts, while improvements in material utilization enhance supply chain resilience by reducing dependency on raw material inputs. Consequently, unnecessary waste, overproduction, excessive ordering, and redundant transportation processes can be effectively minimized. Therefore, these problems are of significant importance in the context of sustainable manufacturing and logistics.

This study addresses a variant of the two-dimensional cutting stock problem with the objective of improving material efficiency within supply chains. The proposed approach is developed based on benchmark instances in literature.

While existing benchmark studies predominantly focus on two-stage and three-stage guillotine cutting strategies, this study expands the solution space by investigating randomized non-guillotine cutting patterns that eliminate traditional guillotine constraints. A mixed-integer goal programming model is formulated with the dual objectives of minimizing total material usage (objective cost) and maximizing the recoverable leftover area (leftover value). The proposed methodology aims to generate cutting patterns at an optimal level by integrating an evaluation framework that simultaneously considers minimal waste and maximum recovery potential.

Computational experiments were conducted, and the obtained results were systematically compared with those reported in the literature. The findings indicate that the proposed non-guillotine approach achieves superior performance, yielding lower objective costs and higher reusable leftover values compared to prior studies. In particular, it is observed that in cases where guillotine constraints restrict pattern diversity, the non-guillotine strategy provides a significant advantage by enabling more flexible and efficient cutting configurations.

Overall, this study contributes to the literature by demonstrating that relaxing structural cutting constraints can lead to both economic and environmental benefits. Furthermore, it provides a robust decision-support framework for practitioners seeking to optimize cutting operations under sustainability and resilience considerations.

This research presents an analysis of electromagnetic interference (EMI) affecting Radio-based positioning readers deployed in logistic rail transport systems, where they serve as critical components of the train position determination subsystem. As modern rail operations continue to evolve toward higher levels of automation, reliable positioning becomes increasingly essential to ensure operational safety, traffic flow stability, and overall system efficiency. However, the electromagnetic environment within rail infrastructure is inherently complex due to dense electrical installations, extensive metallic structures, and the coexistence of multiple overlapping communication and control subsystems. These systems include signaling networks, wireless communications, and onboard electronics. As a result, the railway environment is highly susceptible to electromagnetic interference (EMI), where unwanted electromagnetic energy can disrupt the normal operation of equipment. This complex interaction of emissions increases the risk of performance degradation, signal corruption, or even system malfunction making effective electromagnetic compatibility (EMC) design and mitigation strategies essential for ensuring safe and reliable operations.

The research integrates full-wave electromagnetic simulation using advanced 3D numerical solvers to model the complete coupling pathway from interference sources to the positioning reader antenna. The modeling incorporates multilayer dielectric boundaries, conductive enclosures, and realistic rail vehicle underbody geometries, enabling accurate characterization of both near-field and far-field interference phenomena. Electromagnetic field (EM field) extraction is employed to quantify coupling intensities under various installation configurations and structural alignments. Furthermore, simulation results are systematically validated through parametric studies examining material conductivity, geometric separation, and relative positioning, allowing the identification of worst-case EMI scenarios likely to occur under real-world operating conditions.

By combining numerical modeling with EMI pathway identification, this study aims to significantly enhance the reliability of Radio-based positioning systems. The results establish a robust engineering framework that supports consistent system performance, reduces susceptibility to electromagnetic disturbances, and improves the stability of localization algorithms. Furthermore, unresolved EMI effects can introduce system-level disruptions, including repeated signal reprocessing, misread recovery cycles, and delayed decision-making, which collectively contribute to increased latency and reduced efficiency within overall logistics operations. Ultimately, this work contributes to safer and more efficient operations by ensuring accurate and reliable position detection even in electromagnetically challenging environments.