Department Analytics

The »Analytics« department develops new methods and processes for application-oriented data analytics and AI that go beyond what is currently feasible. The goal is to generate manageable, qualitative data and information for companies from seemingly uncontrollable data volumes and materials.

In terms of methodology, data analytics comprises three areas that build on each other in stages: Descriptive Analytics, Predictive Analytics and Prescriptive Analytics. In the area of predictive and prescriptive analytics, the department is one of the few institutions in the mathematical world to combine forecasting and optimization methods. In this way, it reduces the complexity in the mathematical representation of industrial applications.

A detailed description of why and how we use Descriptive Analytics, Predictive Analytics and Prescriptive Analytics to turn seemingly unmanageable data volumes and materials into manageable, qualitative data and information in concrete applications can also be found here.

Research focus of the department

  • Descriptive Analytics
    In descriptive analytics, historical or status values are recorded and then interpreted. They are used to look backwards, so to speak, and use the results to establish the relationship to the present or to draw conclusions about possible causes of the current situation. In contrast to business intelligence methods, which usually work with existing data records from the past, modern descriptive analytics methods rely on the growing automation and networking possibilities, with which data can be made available in real time if possible and analyzed accordingly. Classic examples are correlation and causality analyses for the comparative analysis of markets and company processes. For association analysis, for example, the Apriori algorithm is often used.
  • Predictive Analytics
    The data collected and interpreted with Descriptive Analytics is the prerequisite for the next maturity level of Data Analytics: Predictive Analytics. Predictive analytics methods are used to automatically evaluate current processes and events on the basis of various data and criteria in order to make predictions - i.e. forecasts - that are as precise as possible. The aim here is to use mathematical methods and models to identify a probable scenario. To do this, data is searched for patterns, models are derived and applied to current data. Classic examples are time series models for the prediction of requirements as well as the classification of processes or for the indication of problems with the help of decision trees and ensemble methods such as random forest.
  • Prescriptive Analytics
    Based on the forecasts of predictive analytics, prescriptive analytics can be used to derive specifications for the planning and control of processes. This requires mathematical models for evaluating future and alternative scenarios, which are fed with current data. The result is individual recommendations for action with which events or trends can be actively influenced. Classical solution methods are mixed-integer programming as well as problem-specific heuristics to determine optimal system configurations.

Four groups: The competences of the department

The competencies required for this, the methods, procedures, models and tools are continuously developed in the following groups:

Data Efficient Automated Learning

Read more about our competencies for more efficiency of AI within the Machine Learning Lifecycle here.

Data Science

Read more about methods, processes, algorithms and systems for extracting insights and patterns from data to generate information and derive recommendations for action here shortly.


Read more about our competences in the area of mathematical optimization and the targeted search for an optimal solution for a well-defined problem here shortly.

Process Intelligence

Read more about our competencies in the area of data analytics for processes here shortly.

Vision and research fields of the department

For a comprehensive description from a visionary perspective of why we want to use analytics to generate manageable, qualitative data and information from seemingly unmanageable data volumes and materials, and how we go about it, see the description of the core competence »Data Analytics Methods«. If you want to learn more about the use cases and domains where we apply our methods and competencies, please follow the research field descriptions linked below.


Core competence

Data Analytics Methods

To extract the raw material »data«, we offer to develop new solutions for Data Analytics and AI as one of three interlocking core competencies. Read more here about why we want to use analytics to generate manageable, qualitative data and information from seemingly unmanageable data volumes and materials, and how we go about it.


Field of research

AI-based demand forecasts for logistics, trade and production

We bring AI-based demand forecasting to bear in logistics, retail, and manufacturing to improve predictions and quantify forecast uncertainties.


Field of research

Sustainability in the digitalized supply chain

Data can be used to design and control processes, organizations and systems in such an efficient, resource-saving and socially responsible way that many of the current challenges can be solved in line with the changed approach to sustainability.


Discovering new cause-effect relationships

Uncovering new cause-and-effect relationships and thus making the right decisions quickly in production and logistics processes in the future.


Digitalized and fully connected transport solutions

Growing traffic volumes and resource shortages require fully networked transportation solutions with intelligent linking of data available both internally and externally.


Supply Chain Analytics

Mathematics can revolutionize the supply chain - with models and algorithms that reduce complexity and interlinked forecasting and optimization methods.