The AI4Mobile project, funded by the German Federal Ministry of Education and Research (BMBF), aimed to investigate how artificial‑intelligence techniques can optimise connection parameters in cellular networks for data transmission. The study covered both public 4G/5G networks and private campus networks, targeting applications such as car‑to‑car, car‑to‑infrastructure (C2X), and industrial processes involving autonomous mobile robots (AMR) and automated guided vehicles (AGV). Götting KG, a small‑to‑medium enterprise specialising in wireless communication for industrial environments, was responsible for defining use cases, specifying system requirements, building a 4G campus network, and developing measurement and analysis tools.
The project’s technical work focused on two industrial use cases: Cooperative Transport of Goods (CTOG) and Manual Teleoperated Driving (MTOD). CTOG requires low‑latency communication as part of a closed‑loop control loop, while MTOD demands high uplink data rates to transmit multiple video streams from a teleoperated vehicle to a control centre. Götting KG built a 4G (LTE) campus network operating in the newly licensed 3.7–3.8 GHz band, which was deployed at partner sites ENWAY in Berlin and Bosch in Hildesheim. Measurement campaigns were carried out along AGV tracks at Bosch, and scripts were written to capture position‑dependent radio‑quality parameters. A custom sensing node was fabricated to receive RF signals, and accompanying software processed IQ samples for jammer detection using convolutional neural networks.
The AI component of the project involved creating models that use the collected data to make location‑ and time‑dependent decisions about radio‑parameter configuration, thereby improving quality of service (QoS) for the defined use cases. Although specific numerical performance figures are not reported in the summary, the work demonstrates the feasibility of deploying a licensed campus network for industrial mobility and the potential of AI‑driven optimisation in such environments.
Collaboration was organised under the AI4Mobile umbrella, with VDI/VDE/IT in Berlin acting as the project sponsor and Fraunhofer Heinrich‑Hertz‑Institut coordinating the effort. Partners included ENWAY GmbH, Bosch GmbH, Ericsson GmbH, Deutsche Telekom AG (associate partner), and several universities such as the Technical University of Dresden and the Technical University of Kaiserslautern. Götting KG’s role encompassed use‑case definition, requirement specification, campus‑network deployment, data‑collection scripting, sensing‑node hardware design, and IQ‑sample analysis. Bosch led the demonstrator deployment, while ENWAY provided the measurement environment in Berlin.
The project ran from 15 March 2020 to 14 March 2023. The onset of the COVID‑19 pandemic in March 2020 limited early activities to virtual meetings, but the team continued to progress on the technical milestones. The final report outlines the results, discusses progress relative to other initiatives, and presents a plan for exploitation and publication of the findings.