The Chair's research focuses on the development and application of methods in predictive and prescriptive business analytics to optimize decisions in the design and management of operations systems. This includes performance analysis and optimization of stochastic systems, applications of methods from artificial intelligence, or combinatorial or nonlinear optimization. Application areas are broad with projects in Industry 4.0, airport operations, automotive production, workforce planning, or management of distribution centers.
Current research projects are in three area i) operations scheduling, ii) management of dynamic systems, and iii) design of lean operations systems.
In this field, our focus is on scheduling operations in a multi-product setting, i.e., allocating heterogeneous jobs to (heterogeneous) resources. We consider both tactical and operational planning situations, for example with sequence-dependent processing or set-up times. Additionally, different aspects of fairness are considered, especially in workforce planning and task assignment.
Another stream of our research is related to tactical and strategic planning of capacity under unceratain and time-dependent demand and service level restrictions. In this setting, production processes are often highly time-dependent, e.g., due to the influence of capacity ramp-ups, seasonal demand patterns, or time-dependent reliability. We support diverse managerial decisions in such stochastic and time-dependent environments.
This research area is related to the analysis and optimization of the performance of manufacturing systems under stochastic conditions, e.g., due to uncertain supplier's capacities, machine breakdowns, or customer demand variability. The aim is to support decisions in the strategic configuration of lean production systems that are robust with respect to planned and unforeseen changes.
The corresponding application areas are broad with projects in Industry 4.0, airport operations, automotive production, workforce planning, or management of distribution centers. Some projects are funded by industry partners, the European Union, or the German science foundation (DFG):
As shown in the figure below, several techniques from predictive and prescriptive business analytics, especially from Operations Research, are used to analyze and to optimize production and operations systems under static and dynamic conditions.
Methods from discrete optimization are applied to deterministic planning tasks. Quantitative models of decision problems are derived and solved using optimization algorithms. These include, e.g., Benders decomposition, Branch & Bound, and dynamic programming. For large-scale optimization problems, efficient heuristic solution approaches are developed.
For the performance analysis of queueing systems, smethods of queueing theory are applied to analyze stationary models to support long term decisions. Growing emphasis has been placed on the analysis of stochastic and time-dependent systems. Besides simulation studies, fast and reliable approximation approaches are developed to support managerial decisions in such stochastic and dynamic environments.
The third stream combines both former operations research directions into the robust optimization of stochastic systems. Here, the capacity of servers, the size of the system, or the acceptance or release of orders are considered as decision variables in an uncertain environment. To solve such Stochastic Programming problems, advanced decomposition and sampling approaches are developed and analyzed to support robust managerial decisions.