Intuitive and flexible control of production systems is one of the key challenges facing modern manufacturing environments. Language- and text-based AI opens up new possibilities in this area by allowing complex processes to be described in natural language and automatically translated into standardized, machine-readable skills. This creates low-threshold access to the programming and adaptation of industrial processes.

Development of a semantic LCE database for resource optimisation in the MEX process within the PLCM. The aim is to link heterogeneous life-cycle data in a contextualised manner using an ontology-based structure, in order to support data-driven and resource-efficient decision-making in product development.

The aim of this work is to develop an integrated CAD–CAM approach in which the CAM component\nis more closely integrated into the modelling and optimisation processes. This is intended to enable a process-integrated\ndesign approach that allows for a more realistic prediction of mechanical properties and\nimproved component performance

The objective of this thesis is to develop a concept that enables the automatic instantiation of the existing ontology from suitable information sources, and to implement this concept as a prototype.

Deviations in manufactured components are ubiquitous and cannot be avoided in practice. Virtual tolerance analyses are carried out to investigate the effects of deviations on product functionality at an early stage of product development. Tolerance analyses using skin model shapes offer one way of accounting for deviations as realistically as possible. These represent deviated components via point clouds and surface meshes. However, the generation of skin model shapes results in gaps and self-intersections in the surface mesh. Addressing and resolving these issues form the focus of this thesis.

This thesis focuses on extending a computational framework for realistic tolerance analysis of mechanical assemblies by considering manufacturing-related deviations and deformation effects during operation. Based on Skin Model Shapes, the work addresses the coupling of structural and local contact deformation models within a unified framework to enable a more realistic evaluation of functional behavior at an early design stage. The core computational framework has already been developed and forms the foundation of this thesis.

The aim of this work is to develop and evaluate a methodology for the effective transfer of MEX data back into product development for design optimisation.

Rising product complexity, shorter development cycles, and sustainability demands are driving automation in product development. AI-based generative design, especially in topology optimization, offers efficient ways to explore complex design spaces and enable largely automated processes.