– Optimizing Your Processes

Better Results with Machine Learning.

Our AI platform enriches the world of engineers with the vast and yet untapped potential of Machine Learning. To achieve this, we employ a technology called Reinforcement Learning. enables process optimization in every dimension: Be it saving costs, maximizing output or achieving overall higher quality. In addition, we offer basic AI functionality like anomaly detection and predictive maintenance.

How does it work?

First, instead of replacing the control logic of single machines, we focus on process interdependencies. To do this, we create a Digital Twin, digital representation of your process, together with your engineers. This comprises all relevant sensors, actuators, relevant exogeneous factors and the optimization goals.

You can also use our platform to expand your existing IoT eco system with AI functionality. If you already have a specific application in mind, you can benefit from our pre-developed solutions HVAC Control and processControl.

After a training phase, HVAC Control will optimize your heating, ventilation and air conditioning processes and reduce your energy consumption by an average of 20%.

processControl for industrial environments will help you to translate your domain knowledge into a Machine Learning framework and thus helps to achieve more with less effort. learns like a human to achieve more

Benefits automatically improves over time

The result: An ever-growing degree of automatization.

Process optimization in every dimension

Flexible optimization goals- be it quality, energy, throughput or performance. Or all of them combined, and at the same time.

Risk-free for your ongoing operations

The system is based on a functional Digital Twin, meaning a training phase without any interference to your live processes.

Predictive maintenance already included

Based on our system architecture, more classical applications like predictive maintenance and anomaly detection are readily available.


  • Easy to implement:
    The dynamic control policy is based on the graphicx AIoT platform and therefore seamlessly connects to industrial assets.
  • Based on predictions:
    We are relying on proven existing data that is publicly available (such as weather forecasts) to improve the predictions of our system.
  • Full control:
    The platform provides you with a visualized feedback of everything going on. This is how you stay in charge of your operation and gain full transparancy over every process interaction.
  • Step-by-step approach:
    If you feel uncomfortable handing over control to the platform from the start, you can always come back to it at a later point. You will still benefit from predictive maintenance, anomaly detection and optimization suggestions.
  • Trains with a Digital Twin:
    The dynamic control policy explores process interactions on a Digital Twin. That is how you can demo actions before they get executed in the real process.
  • Easy definition of your optimization goals:
    You define how the algorithm works by stating objectives, constraints and parametrization. It is an easy way to shape it towards your needs and goals.

Looking for a customized AIoT solution?

Get in touch.

  • +49 228 299 745 99