Artificial Intelligence

As the volume of data for all areas of life and applications increases, demand is growing for intelligent systems that can autonomously evaluate and interpret such data, and translate it into decisions and actions.

This is where machine learning methods and concepts come into play. They often involve highly complex evaluations based on artificial neural networks, which perform at their most efficient with hardware architectures inspired by the brain.

In addition, special latency, efficiency and security requirements often make it necessary to evaluate the data directly where it is generated. That’s why we focus on local data processing and direct our efforts toward giving devices and systems artificial intelligence.

Machine learning in distributed systems

Distributed machine learning takes place directly in the sensor system (embedded processing) or close to the network (edge processing). We develop energy-efficient and robust solutions to integrate machine learning into edge or embedded devices, for instance, by utilizing neuromorphic hardware.


Embedded machine learning

Implementation and integration of machine learning algorithms on embedded devices


Neuromorphic hardware

Use of neuromorphic hardware for machine learning: consulting, design and implementation