OPM 661: Business Analytics: Robust Planning in Stochastic Systems

Business Analytics helps to optimize decisions for the design and management of operations systems and production processes. A major driver of the performance of operations systems is stochastic variability. For example, production systems often operate in an uncertain environment due to uncertain demand, unreliable machines, or random processing capacities.
In order to support robust decisions, we apply analytical solution approaches based on techniques from predictive and prescriptive analytics. The basic concepts of the analysis of Markovian queueing systems are explained in detail and performance evaluation approaches are implemented in Python. To create digital twins of operating systems, simulation techniques are introduced and implemented. This allows to analyze the sensitivity of system parameters on the main performance measures. Advanced topics such as queueing systems with general distributions, heterogeneities, and time-dependent input parameters are covered. Additionally, general managerial insights, for example economies of scale and the value of flexible capacities are discussed. Methods and performance measures of robust planning and optimization are introduced. Students become familiar with concepts and tools for predictive and prescriptive business analytics.
Moreover, we will implement those concepts using the programming language Python to perform sensitivity analyses and to develop managerial insights for stochastic operations systems. During the course the students will work on several case studies and assignments (individual and in groups).

Learning outcomes
Students learn to understand the impact of stochastic variability in operations systems. After this course students are familiar with the theory and practice of the analysis of stochastic systems. They learn to implement, adapt and to apply methods and tools from Business Analytics e.g. analytical approximations, simulation, and robust planning methods to support managerial decisions.

Necessary prerequisites
OPM 561 and OPM 560 OR: OPM 561 and „Schlüssel­qualifikation 1: Programmierkurs Python“ (Angebot der WIM)

Recommended prerequisites
Participants should be familiar with the fundamentals of production and operations management. The course further assumes a basic knowledge in mathematics (including linear programming) and in statistics (probability distributions).

Forms of teaching and learningContact hoursIndependent study time
Lecture with integrated exercise4 SWS9 SWS
Exercise class2 SWS8 SWS
ECTS credits8
Graded yes
Form of assessmentAssignments and presentations (70%), written exam (45 min) or oral exam
Restricted admissionyes
Further information
Performing lecturer
Prof. Dr. Raik Stolletz
Prof. Dr. Raik Stolletz
Frequency of offeringSpring semester
Duration of module 1 semester
Range of applicationM.Sc. MMM, M.Sc. WiPäd, M.Sc. VWL, M.Sc. Wirt. Inf., M.Sc. Wirt. Math.
Preliminary course work
Program-specific Competency GoalsCG 1, CG 4
LiteratureWill be announced during the lecture
Course outlinePlease note that this outline will be subject to changes:
Introduction to performance evaluation and simulation
  • Queueing systems, decisions, and applications
  • Performance measures and simulation of queueing systems
Performance analysis of Markovian queueing systems
  • Analysis of stochastic processes and Markov chains
  • Performance analysis and economies of scale
Impact of variability in queueing
  • Queueing systems with general distributions
  • Time-dependent analysis of queueing systems
Optimization and queueing
  • Optimization concepts and approaches
  • Robust planning with scenarios
Practical insights
  • Predictive and prescriptive analytics with Python
  • Guest lecture