The Fraunhofer Institute for Integrated Circuits IIS, in cooperation with Nuremberg Airport, conducted a pioneering pilot project for automated passenger assistance. The test centered on a mobile, humanoid robot that served as an interactive information assistant for travelers in the departure hall.
The demonstrator is capable of capturing and immediately answering complex questions about parking, dining options, or check-in counters in various languages. The aim of the research was to evaluate the performance of such assistance systems under the demanding conditions of a busy public space and to investigate the acceptance of the technology among airline passengers.
Technologically, Fraunhofer IIS relies on a so-called edge AI solution. Unlike conventional AI systems or chatbots, which depend on a permanent cloud connection and external data centers, all data processing takes place locally, directly on the end device. For this purpose, three specialized AI models—for speech understanding, dialogue guidance, and speech generation—were specifically compressed and optimized. By using a local Large Language Model (LLM) in conjunction with a special Retrieval Augmented Generation (RAG) system, precise, airport-specific information can be provided without delay. This approach ensures complete data sovereignty, as no audio recordings or personal information leave the device.
A key advantage of this self-contained system architecture is its operational reliability, independent of the existing network infrastructure. Even in the event of a Wi-Fi outage or an unstable internet connection, the robot remains fully functional. Christian Albrecht, Head of Corporate Communications at Nuremberg Airport, emphasized the importance of this practical research for the location, while Dr. Axel Plinge from Fraunhofer IIS highlighted the significance of the models' compact design for data security. The project provides important insights into the user-friendliness and necessary hardware for future, practical integrated systems in the service sector.
The results of this field test will be presented on June 24, 2026, as part of the Nuremberg Digital Festival. In addition to the technical evaluation, the focus will also be on the interaction patterns of travelers in order to further optimize the human-machine interface. The project illustrates the trend toward the decentralization of artificial intelligence, moving away from large server farms toward high-performance, embedded platforms. For Nuremberg Airport, the test represents another step in the integration of modern robotics to make the flow of information for passengers more efficient and to relieve the workload of staff at the stationary information desks.