From sound to movement: BrainScaleS-2 processes analogue signals in real time

In the demonstrator, the owl figure turns, driven by a servo motor, toward the direction of detected sound sources. IMAGE: Stradmann, Y. et al. IEEE Xplore, 2026.

Two microphones hear a sound. A small owl turns towards it. Behind this simple demonstration is a neuromorphic computing system that processes the microphone signals directly in analogue form and uses the result to control a motor in real time.

Researchers from the EBRAINS community have demonstrated this approach using the BrainScaleS-2 platform. Instead of first converting the microphone signals into digital data, the signals are fed directly into the analogue computing circuits of the BrainScaleS-2 chip. This allows the system to process information from the sensors without analogue-to-digital and digital-to-analogue conversions.

In their demonstration, two microphones pick up the same sound at slightly different times. A spiking neural network running on BrainScaleS-2 uses these differences to determine the direction the sound came from. The result is then used to control a servo motor, which turns the owl towards the detected sound source.

Because BrainScaleS-2 can run neural computations around 1,000 times faster than biological real time, the system can respond quickly to transient sounds such as short noise peaks. The demonstration shows how sensing, computation and movement can be connected in a single neuromorphic system.

Processing sensor signals close to where they are generated could be useful for applications where efficiency and fast responses are important. By working with analogue signals directly, neuromorphic systems such as BrainScaleS-2 can avoid some of the data conversions required in conventional computing systems.

The work was presented at the Neuro Inspired Computational Elements (NICE) Conference 2026 in Atlanta, USA.

BrainScaleS-2 is available to researchers through EBRAINS for experiments in computational neuroscience and neuromorphic machine learning.

The next BrainScaleS hands-on tutorials will be offered on 7 October 2026 at the tutorials day of the EBRAINS Users Days in Heidelberg, Germany, and at the virtual training days online on 13 October 2026.

EBRAINS researchers present latest neuromorphic computing advances at NICE 2026

Read the full paper:

Y. Stradmann, J. Schemmel, M. A. Petrovici and L. Kriener, "Real-Time Processing of Analog Signals on Accelerated Neuromorphic Hardware," 2026 Neuro Inspired Computational Elements (NICE), Atlanta, GA, USA, 2026, pp. 1-6, doi: 10.1109/NICE69539.2026.11567519.

Author: Helen Mendes
Contact: press@ebrains.eu