Adaptive, self-optimizing nonwoven production based on reinforcement learning

As part of our software solutions for automation, visualization and optimization of processes, we are also active in research.

We shape the digital future

An example of this is our research project AdaNowo. This improves visibility, minimizes energy consumption and reduces waste.
+49 2364 5049012


Our goals

The research goal is to economically optimize product changes and long-term process management in nonwovens production using an online-capable reinforcement learning algorithm.
MAT Solution is the network coordinator
and responsible for database interfaces, dashboards and marketing.

Supported by:

Federal Ministry for Economics Affairs and Climate Action on the basis of a decision by the German Bundestag

Other partners are:

RWTH Aachen University, Institut für Textiltechnik (ITA)
BNP Brinkmann GmbH & Co. KG, Hörstel
Dilo Spinnbau GmbH


AdaNowo as part of a simulation

Is the simulation of the process control and setting of the carded nonwoven production in integrated into a playful environment, then further optimization potential can be used.

Areas of application

  • For training future specialists as a simulation environment that promotes learning
  • To demonstrate the potential of systematic production data collection
  • Validate intelligent production planning algorithms quickly and cheaply

Features of a simulation

  • Interaction with simulated process
  • Realism through the use of industrial data and ML
  • Simulation cost dynamics
  • Simulation quality (tensile strength, basis weight, process variation)
  • Virtual production orders

    Your contact person:

    Ruben Kins
    Institute of Textile Technology

On-Premises or in the cloud?

AdaNowo can be operated locally or directly on the server (on-premises).

Alternatively, we also offer AdaNowo as a cloud solution from the well-known service providers Amazon Web Services (AWS), Microsoft Azure or Google Cloud, thus enabling global access to the data.