One of the FZI's main areas of focus is AI-supported image and sensor processing. Here, algorithms are being developed for the automatic recognition and segmentation of body regions and relevant skin areas. To this end, the FZI selects and integrates suitable camera and sensor systems, trains corresponding machine learning models, and ensures their reliable integration into the robotic platform.

In addition, the FZI is responsible for natural language processing in the project. It is developing a system for processing natural language inputs that recognizes and correctly interprets the information. In this context, large language models (LLMs) are being researched, trained, tested, and optimized for medical use. The goal is to translate voice inputs from specialist personnel into executable robot programs, thereby implementing a no-code programming approach. In addition, FZI is developing an interactive dialogue system that recognizes uncertainties and resolves them through targeted queries.

In the field of system integration and robotics, the FZI develops interfaces between robot systems, backends, and sensor technology, ensuring a modular, scalable, and secure architecture. It also supports data management, program storage, and interoperability with existing medical IT systems.

In addition to its technical tasks, the FZI is also involved in usability evaluation, user testing, ethical reflection, project communication, and knowledge transfer. It analyzes test results, derives optimization potential, and contributes its expertise on data protection, fairness, and acceptance of AI systems. In this way, the FZI makes a significant contribution to ensuring that HELIA can be used in a technically efficient, user-oriented, and responsible manner in the clinical environment.