PROSTEP publishes three articles on artificial intelligence

"Projektmanagement neu denken" (Rethinking Project Management) is the name of the GPM's publication series in which the volume entitled "KI in der Projektwirtschaft 2" (KI in the Project Economy 2), was recently published. On over 300 pages, a number of different authors turn their attention to topics such as the cross-industry use of AI, industry-specific AI applications and other practical use cases, as well as the changes that AI has brought to the roles in project management. PROSTEP has contributed three articles, which reflect the practical experience with the use of AI we have gained in customer projects and research projects.

Requirements management is one of the application areas in engineering where the use of AI promises considerable potential for savings. In their article on the use of AI in requirements management systems, Elaheh Nabati, Luciana Kröseler, Frank Pfirmann and Noah Sentürk present three promising use cases, namely improving the description of requirements using AI-based sentence structure checks, automatically filling fields in defect and test management based on the evaluation of error descriptions, and the analysis of semantic similarities between requirements with the aim of avoiding duplications and creating automatic links.

The authors not only explain how the three use cases were implemented but also address the challenges that had to be overcome. These include not only the computing capacity required for training the large language models and procuring and processing the training data but also the question of how personal and company-specific information in this training data can be anonymized in such a way that the training results do not allow any conclusions to be drawn about individual employees.

AI is turning an increasing number of business processes into black boxes. Although the results of these processes can be assessed and documented, how the processes themselves came about cannot. AI also reduces the extent to which employees are involved in the decisions made by a company. But first and foremost, AI can only develop company-specific learning content if it is brought into close contact with the knowledge holders of this content at every level. This has the positive side-effect that employees see AI as an opportunity rather than a threat.

In their article on employee participation and AI, Elaheh Nabati and Knut Stang describe how employee participation in AI projects can be ensured and increased. According to the authors, a prerequisite for successful AI projects is employee involvement. They use two customer projects to demonstrate what this can look like in practice.