AI Software Academy

Operating Time: 07/2021 - 12/2023
Source of Funding: Land BW
Project webpage:


  • Prof. Dr. Steffen Staab. Head of the Department for Analytic Computing (AC), Institute for Parallel and Distributed System, University of Stuttgart.
  • Prof. Dr.-Ing. Wolfgang Nowak. Head Institute for Modelling Hydraulic and Environmental Systems (LS3/SimTech)
    Co-spokesperson of the Cluster of Excellence EXC 2075 "Data-Integrated Simulation Sciences"
  • Tim Schneider. Doctoral researcher, Department for Analytic Computing (AC) and SimTech.

Our Mission

Artificial Intelligence (AI) intrinsically requires dedicated software and the related software engineering (SE) skills. Additionally, profound domain knowledge (X) in engineering, sciences and didactics are usually needed in order to efficiently and effectively use AI in academia and industry. AISA aims at educating specialists that are equipped with interdisciplinary skills in all contributing disciplines: AI+SE+X.


  • Artificial Intelligence (AI) is a key technology in today's industrial and academic life. It comprises a huge variety of different methods, tools, and use cases. Understanding the relation between domain-specific problems and the proper AI tools to be used is a key competence that is often lacking. Also, the development of your AI methodologies requires profound knowledge.
  • Software Engineering (SE) is a competence of neglected in academia. When developing research codes, sophisticated individual AI solutions, or simulation tools, things like documentation, testing, code readability, robustness, and so forth are often neglected. However, SE skills can boost efficiency in this regard. They can offer many benefits in larger teams, and for community wide roll out of academic and industrial solutions, and they improve on the sustainability of software and the related methods.
  • Application domains (X) come along with specific requirements. For instance, basic physical principles must be obeyed. Knowledge about the data may be available. Side conditions not captured in the data such as experimental conditions may be relevant. Result representation can be critical in order to trigger acceptance. Interaction with domain-specific research data infrastructure may be required.

By fusing competences from AI, SE and X, AISA will equip specialists with the needed tools, methods, workflows, and didactic skills in order to trigger acceptance of novel technology.




  1. Schwindt, S., Meisinger, L., Negreiros, B., Schneider, T., & Nowak, W. (2024). Transfer learning achieves high recall for object classification in fluvial environments with limited data. Geomorphology, 455, 109185.

Project Members

To the top of the page