Checkmate of Routine
The Use of AI in Project Management
When, in 1997, a computer developed by IBM named „Deep Blue“ defeated the reigning world champion Garry Kasparov in chess, the astonishment was immense. While computers had already been able to solve complex mathematical problems for some time, the skills necessary for chess – such as spatial reasoning, pattern recognition, foresight, judgment, and decision-making—had, until that day, been attributed more to humans than to machines.
Prof. Dr. Reinhard Wagner
Since then, not only has the performance of computers significantly advanced, but also their range of application possibilities. With the help of neural networks, powerful computers today can make long-term predictions, use „machine learning“ to recognize complex patterns from vast amounts of data, and ultimately prepare or even make decisions under uncertainty. The development of chess computers was merely the starting point for a trend that has had profound effects on our society. The following focuses on the consequences of using artificial intelligence (AI) for routine tasks in projects.
Despite the Uniqueness of Projects, Routine Dominates
In many definitions of projects, their uniqueness is emphasized, and a clear distinction is made from routine tasks (DIN 2009a). For example, the development of a new vehicle is organized as a project, but its mass production is not. Development involves creative collaboration within a limited timeframe with the goal of creating the most innovative product possible. Production, on the other hand, is the repetition of finely coordinated processes, optimized for efficiency, to produce as many vehicles as possible. While numerous robots are used in production, working together through digital control and automation solutions, project execution is still approached more traditionally. Although software solutions are used for planning and controlling projects, they are not yet highly automated, digitized, or optimized using AI.
unique, which leads to a preference for human involvement over computers. However, when analyzing the steps in projects more closely, many routine tasks become apparent that occur in similar or even identical forms in other projects. This starts with the definition of a project, for example, using a project brief, and continues through the planning steps, progress analysis, reporting, and concluding with evaluation and lessons learned. Even though the specific data may differ from project to project, there is a significant similarity that presents potential for digitalization, automation, and optimization using AI. Lean Project Management (Hüsselmann 2021, Wagner & Erasmus 2023: p. 126) highlights the potential within projects and helps minimize resource waste by focusing on the value stream. It is important to emphasize that humans are not made redundant in projects but are instead relieved of routine tasks, allowing them to focus on other activities, such as communication, leadership, and other social aspects of project management.
In Which Areas AI Can Be Applied in Projects
Projects have a clearly defined beginning and end. In between, specific tasks must be completed. Focusing on project management, process models such as DIN 69901- 2 (2009b) describe the tasks to be performed, from the initiation, definition, and planning stages to the control and completion of a project. These process descriptions provide an ideal basis for assessing the potential of AI in project work within a company. Although the use of AI in projects is still in its early stages, an analysis by Tiba Managementberatung GmbH (Tiba 2023, Wagner & Wagner 2024) shows that Machine Learning, Deep Learning, and Generative AI applications are currently being primarily applied during the planning phase, while their use in the pre- and post-project management phases remains relatively rare.
For instance, Mariani et al. (2023) report on the use of „Unsupervised Machine Learning“ to analyze and assess stakeholders in the complex environment of construction projects. This AI-driven approach is significantly more powerful than previous human methods, as AI can process much larger datasets in a shorter time, generate more detailed analyses, and thus provide better decisionmaking support for stakeholder management. Mancini et al. (2023) present an example from the decommissioning of nuclear facilities, where a smart approach using „Fuzzy Bayesian Belief Networks (FBBN)“ is employed to manage the risks of a complex project, estimate delays, and forecast cost overruns within acceptable tolerances.
„Generative AI“ is also being used in projects. This ranges from drafting project briefs and creating project plans to developing „smart contracts“ for a project‘s suppliers (Ritsche et al. 2019: 91). These contracts can then be automated using blockchain technology and payments via cryptocurrencies. Particularly in distributed project networks, where transparency of interactions and trust in legally and financially binding processes between partners are crucial, such a combination of AI, blockchain, and financial transactions via cryptocurrencies makes sense (Wagner et al. 2024). Many other applications are conceivable.
For example, Microsoft applications such as PowerBI or PowerAutomate can be used to collect relevant project data and prepare it in a predefined dashboard for project leadership and the steering committee. Project meetings can be recorded, transcribed, and analyzed using Copilot applications. This can help identify team sentiments, track discussed actions and measures, and capture experiential knowledge. The latter is particularly important for ensuring the availability of lessons learned within the company.
At the program and project portfolio level, AI can also be used to great advantage. By analyzing large datasets across all projects within the program or portfolio, cluster risks can be identified, resource allocation can be optimized, and experiential knowledge can be extracted in line with pre-set goals. Some project management software providers are now integrating AI applications into their programs to make them even more powerful, although we are still at the beginning of this development. A study by IPMA® (2020) identifies significant savings potential in routine tasks in project management.
Benefits, Requirements, and Unwanted Side Effects
In the current discussion about the use of AI in projects, more is being written about the dangers - such as job losses and significant changes in project management tasks—than about the benefits. Yet, the use of AI to handle routine tasks in projects promises a significant acceleration of processes, saving of resources and costs in project execution, and an improvement in the quality of results, both in terms of the project‘s product and the process, including process documentation. This is achieved by the ability of computers to analyze large amounts of data („big data“), allowing project management to spend minimal time correcting or supplementing information. As a result, the quality of information increases, and the error rate decreases with the use of AI in projects.AI is also objective, which can be an advantage in the analysis of stakeholders, risks, and decision options, especially when there are „cognitive biases“ within the project team. AI can recognize patterns across various projects or project data that might be overlooked by humans, which is particularly important for decision-making under uncertainty. In these situations, AI can help prepare information for critical decisions, making the work of the steering committee easier. Finally, the use of AI is also appealing to Generation Z, which seeks modern technology in their daily work and often chooses employers based on this criterion (Tiba 2021).
To realize the benefits of using AI for routine tasks in projects, a variety of prerequisites must be established. These include formulating a clear strategy and specific goals for the use of AI within the company, as well as establishing AI governance to ensure that AI is used in an ethically responsible and transparent manner.
Adherence to both internal and external compliance regulations must also be considered when zmplementing AI. This includes issues such as data protection and information security. Additionally, precautions must be taken regarding the quality of AI-generated results, as AI can sometimes „hallucinate,“ meaning it can produce incorrect or misleading outputs. It is advisable to subject AI results to a quality control check (Galgenmüller & Wagner, 2024). One often overlooked prerequisite for the use of AI is the availability of data. This may sound trivial, but it presents a real challenge for many companies, as information is often stored in various systems and formats. The lack of system integration can significantly hinder the use of AI and requires investment to resolve.
The qualifications of project participants are also a critical factor in successfully utilizing AI. This includes competencies in using digital tools and technologies, methods for data analysis („data analytics“), and the ability to assess the quality of AI outputs. Finally, the implementation of AI should be supported by change management, which focuses on addressing the concerns and fears of those affected, ensuring their involvement, and guiding them through the introduction process.
Nevertheless, the use of AI in a company can lead to unwanted side effects. For example, the introduction of AI may cause fear among some employees, leading to rejection or „passive resistance,“ where they no longer engage as fully as they did before or only perform tasks strictly by the book. Another side effect is that employees might fully rely on AI and stop critically questioning its outputs.
This is particularly problematic for project managers, who are ultimately responsible for the project’s results and must make their own assessments. Technical dependence on systems and data can also pose issues, especially in the case of a power outage or system crash. Before using AI, it is essential to ensure system integration, otherwise, crashes and errors may occur. Some applications (such as blockchain) require significant computing power, leading to high energy consumption, which raises questions about the sustainability of AI use. Who has access to which data? Are the data protected from unauthorized access, and how can the violation of third-party rights be prevented? Data protection and security are central risks associated with AI applications and should therefore be addressed as part of an AI governance framework.
Checkmate for Project Management or the Project Manager?
The literature almost unanimously emphasizes that the use of AI in projects brings more advantages than disadvantages. The IPMA study (2020) highlights that project managers would not be replaced by AI; instead, AI serves as an „assistant,“ providing valuable support by relieving project managers of routine tasks, allowing them to focus on „soft skills“ such as leadership, communication, and collaboration. Peter Taylor (2022) similarly argues that AI can be seen as a „team mate,“ enabling project managers to work „smarter, not harder.“Project management has continuously evolved over the past decades, and AI will likewise drive an evolutionary advancement in project work (Bernert et al. 2023). However, this also necessitates a realignment of the skills, approaches, and technologies required for projects. In line with Lean Project Management and the concept of the „Lazy Project Manager“ (Taylor 2015), the goal is to reach the next level of productivity in project execution by largely digitalizing, automating, or optimizing routine tasks through AI. This will lead to an acceleration of processes and a significant reduction in resource use in projects. As a result, project participants can focus more on the aspects necessary for effective collaboration, ensuring the success of their projects.
In this article, we have also addressed the benefits, the creation of appropriate prerequisites, and the potential unwanted side effects of using AI. Companies should therefore approach the topic strategically and accompany the introduction of relevant AI technologies through a systematic change process (Lang & Wagner 2022).
Literature
Bernert, C., Scheurer, S. & Wehnes, H. (Eds.) (2024). KI in der Projektwirtschaft. Was verändert sich durch KI im Projektmanagement? UVK-Verlag.
DIN (2009a). Projektmanagement - Projektmanagementsysteme - Teil 5: Begriffe. Beuth.
DIN (2009b). Projektmanagement - Projektmanagementsysteme - Teil 2: Prozesse, Prozessmodell. Beuth.
Galgenmüller, A. & Wagner, R. (2024). Künstliche Intelligenz in der Praxis – Aller Anfang ist schwer.
[retrieved on June 23rd, 2024].
Hüsselmann, C. (2021). Lean Project Management. Hybride Methoden wertschöpfend anwenden. Schäffer-Poeschel Verlag.
IPMA
®
(2020). Artificial Intelligence impact in Project Management.
[retrieved on June 23rd, 20244].
Lang, M. & Wagner, R. (2022). Das Change Management Workbook: Veränderungen im Unternehmen erfolgreich gestalten. 2nd edition. Hanser-Verlag.
Mancini, M., Mariani, C. & Manfredi, C.M. (2023). Nuclear decommissioning risk management adopting a comprehensive artificial intelligence framework: An applied case in an Italian site. Progress in Nuclear Energy, 158, pp. 1-12.
Mariani, C., Navrotska, Y. & Mancini, M. (2023). Unsupervised machine learning for project stakeholder classification: Benefits and limitations. Project Leadership and Society, 4, pp. 1-12.
Ritsche, F.-P., Wagner, R., Schlemmer, P., Steinkamp, M. & Valnion, B. (2019). Innovation Project EPC 4.0. Unleashing the hidden potential.
[retrieved on June 23rd, 2024].
Taylor, P. (2015). The lazy project manager: How to Be Twice as Productive and Still Leave the Office Early. 2nd edition. Infinite Ideas.
Taylor, P. (2022). AI and the Project Manager. How the Rise of Artificial Intelligence Will Change Your World. Routledge.
Tiba (2021): Transformationsbedarf für Unternehmen aus Sicht der Generation Z. Tiba.
Tiba (2023): Künstliche Intelligenz im Projektmanagement – Chancen, Gefahren, Anwendungsmöglichkeiten.
[retrieved on June 23rd, 2024].
Wagner, P. & Wagner, R. (2024): The evolution of technology in artificial intelligence and its impact on project management. In: Hemanth, D.J., Kose, U., Patrut, B. & Ersoy, M. (Hrsg.) Innovative Methods in Computer Science and Computational Applications in the Era of Industry 5.0. ICAIAME 2023. Engineering Cyber-Physical Systems and Critical Infrastructures, 10, pp. 268-293. Springer.
Wagner, P., Wagner, T. & Wagner, R. (2024). The impact of blockchain technologies on project management and the complementary role of AI. Proceedings of the 12th IPMA Research Conference “Project Management in the Age of Artificial Intelligence”, April 19-21, 2024, College Park.
Wagner, R. & Erasmus, J. (2023). Projektmanagement in der Automobilindustrie. Effizientes Management von Fahrzeugprojekten entlang der Wertschöpfungskette. 6th edition, SpringerGabler.