As companies increasingly invest in Artificial Intelligence (AI) research and implementation, there is a growing demand for skilled professionals capable of solving real-world AI challenges. Universities of applied sciences play a crucial role in bridging the gap between academic knowledge and practical application. Developing a didactic learning model for teaching AI skills is essential to equip students with the necessary competencies to meet industry demands. This research aims to demonstrate an effective teaching model that enables students to tackle complex AI problems in corporate settings, fostering innovation and driving economic growth. We seek to answer the following research question (RQ): What could a teaching-learning model at a university of applied science look like to enable students to solve real-world AI problems for companies and develop students AI skills?

To answer the RQ, we follow the structure of a Design Science Research approach. First, we use an AI skills model for higher education with the dimensions AI innovation, AI technology, AI management and AI application areas to identify the relevant skill areas. We ask our cooperation partner Siemens Digital Industry (DI) which AI skills will be particularly important for employees in the area of specification, development and operation of IT systems (problem relevance). Second, we adapt and apply a teaching-learning model (artifact) that improves relevant AI skills of master’s students in computer science programs and train them to solve real-world problems of companies. Third, as part of the evaluation, we ask those involved in the course (students, lecturers, cooperation partners) whether the students' AI skills have improved as a result of participating in the course (evaluation).

The results of the Siemens DI survey show that AI skills in the dimensions AI innovation and AI application areas are particularly important for our cooperation partner. AI innovation skills focus on creatively applying AI to address real-world problems, while skills in the AI application areas involve recognizing opportunities across diverse industry sectors.

The didactic teaching model we apply is an innovative educational artifact designed as a project-based learning course. Teams of master's students of computer science and related fields at the Nuremberg Institute of Technology compete to solve real-world problems presented by Siemens DI. The course follows a six-phase structure and focuses on both technical and soft skills development. By bridging academic learning and industry needs, this artifact not only enhances graduate employability but also shows the industry partner new perspectives on their own real-world problems through the unbiased student approach. We call our artifact Information Management Challenge. The adaptability of this course allows it to address diverse learning content. As a result, we can expand our focus to AI skills in summer semester 2024 with a project from our industry partner on using large language models in logistics.

The evaluation of the artifact shows that students' AI skills improved in all four skill dimensions as a result of participating in the course summer semester 2024. The improvement was disproportionately strong in the skill dimensions of AI innovation and AI application areas, which Siemens DI rated as particularly important. Therefore, we take this result as an indication that the artifact is working in the desired direction.

Keywords: Project-based learning, university-industry cooperation, master level course, artificial intelligence skills, teaching model.