Henkilöstöresurssien hallinta dynaamisessa projektiympäristössä

Breite, R. (Examiner)
Turunen, J. (Examiner)

Activity: Evaluation, examination and supervision › Supervisor of master student

Period	17 Feb 2025
Examinee	Antti Kinnunen
Degree of Recognition	National

Documents & Links

The main objective of this research was to develop a human resource management model for a maintenance and support business unit of a Finnish project-based industrial company. The central problem addressed was human resource management in a dynamic project environment, where the same human resources serve both projects and continuous service operations. Service level agreements and unpredictable work that require rapid response to customer needs within defined time limits posed a particular challenge. The research was conducted as a case study based on a theoretical framework derived from a literature review and analysis of the target company's current state and needs. The analysis aimed to identify best practices in organizational structure, management practices, and resource planning suitable for a dynamic project environment. The development work was complemented with empirical data analysis, particularly in developing demand and capacity needs forecasting models. The developed model integrates Sales & Operations Planning (S&OP) process with hybrid organization principles, where the monthly S&OP process serves as a framework for combining quantitative forecasting models and team-based human resource management. The study developed forecasting models that applied the LightGBM machine learning model and the Prophet time series model for demand forecasting, and a logistic regression-based model for capacity needs assessment. As key results, the research developed a comprehensive human resource management model that was integrated from four core elements. The model elements consist of the S&OP process, team-based organizational structure, decentralized decision-making, and applicable dynamic project environment management methods. In tests conducted with historical test data, the short-term LightGBM model achieved MAPE values of 24-47% for different work types in demand forecasting, and the longer-term Prophet model achieved MAPE values of 20-34% with a 6-12 month forecast horizon. In the capacity needs forecasting model, modeling project dynamics with a logistic function produced a high coefficient of determination (R² = 0.91) based on test data. The research provides a new perspective on human resource management in dynamic project environments by combining S&OP process and hybrid organization principles. The developed model is tailored to the needs of the target company, but its basic principles are applicable to similar operating environments. A key limitation is the lack of practical testing of the model, which constitutes an important topic for future research.