Research Associate (f/m/d) Research focus: AI-based design of engine architecture, HAP5

Modern // regional // industry-oriented: The Brandenburg University of Applied Sciences (Technische Hochschule Brandenburg, THB) is a young and dynamic university with approximately 3,200 students enrolled in 26 study programmes across the Departments of Computer Science & Media, Engineering, and Business Administration, located in the city of Brandenburg an der Havel — directly at the doorstep of Potsdam and Berlin. As a family-friendly university on a green campus, we offer attractive working conditions.

Within the Department of Engineering, as part of the research project DaHKIZ (Data-based Holistic AI-supported Civil Engine Development), the following position with the opportunity to pursue a doctoral degree is to be filled at the earliest possible date, initially limited until 30 April 2029: in cooperation with Rolls-Royce Dtl.

Research Associate (f/m/d) Research focus: AI-based design of engine architecture, HAP5 The focus of your work lies in the industrial research: AI-based design of engine architecture. You investigate efficient and physics-informed ML/AI strategies for the preliminary design of engine concepts, in particular the semantic integration of the "Preliminary Engine Architecture" tool and the GTlab framework for Large Language Models (LLMs), as well as multi-agent architectures for the automated synthesis and evaluation of design workflows. As customer inquiries for new propulsion concepts require the rapid, integrated evaluation of very large and complex data sets from preliminary design tools, AI methods shall make new technically feasible and requirement-compliant engine concepts quickly accessible.

Your main responsibility

• AI-supported component selection: investigation and preparation of existing data sets for automated preliminary design, generation of missing data, and provision of a database suitable for ML/AI evaluation.
• AI-supported evaluation and automated analysis of engine concepts: investigation of ontology-based metadata models for the machine-readable description of model structures, workflows and parameter dependencies, as well as a multi-agent architecture (LLMs, Retrieval-Augmented Generation, function calling to GTlab).