How is neuromorphic touch changing robotics?

By: researcher Alejandro Pequeño Zurro

Why touch was not a priority for a long time

Mimicking human touch is complex. A hand contains approximately 17,000 mechanoreceptors: specialised sensory cells that perceive pressure, movement and texture. This wealth of signals enables sophisticated interaction, but also poses a challenge in robotics.

In traditional AI systems, touch sensors send signals to a processor at fixed intervals, regardless of whether anything has changed. Even when a surface remains still, the data stream continues. Multiplied by thousands of sensors, this leads to a continuous stream of repetitive information. Processing this information consumes energy, slows down response times and puts strain on the system without always providing new insights.

Thinking like the nervous system

Neuromorphic touch takes a different approach. Not everything is equally relevant; change is what counts most.

Pequeño Zurro compares it to the human body. When you touch the palm of your hand, you expect information from the hand, not the elbow. Neuromorphic sensors follow the same logic. Instead of measuring at fixed intervals, they only respond when something actually changes. This approach has two important consequences. Firstly, the amount of data to be processed is greatly reduced, which leads to lower energy consumption. Secondly, the response speed improves. The system does not have to wait for the next measurement round, but responds at the moment that matters.

The problem of the slippery pen

A simple example illustrates this difference. Suppose a robot picks up a pen. Initially, it succeeds, but the pen slowly starts to slip. In a traditional system, this change is registered, but it disappears in the large amount of other data. The response comes too late and the pen falls.

In a neuromorphic setup, that small change in pressure would immediately activate a signal. The robot can adjust its grip before things go wrong. The difference is not in strength, but in timing and relevance.

Touch and trust

Neuromorphic touch also affects the way people experience robots. Research into social robotics is looking at how tactile interaction influences feelings of trust and closeness. In a collaboration between CogniGron and the University of Southern Denmark, a pilot study tested a robot that could give a personalised handshake. Participants rated this robot as friendlier and more reliable. This can make a difference, especially in contexts such as elderly care or therapy, where safety and human contact are central.

Feeling textures, acting carefully

Neuromorphic touch also offers new possibilities in healthcare, for example in surgical robotics. During an operation, it is important to recognise different structures and levels of resistance. Traditional sensor systems often provide too much information at once, making quick and accurate interpretation difficult.

Neuromorphic systems make it possible to process tactile information directly and in a targeted manner. This opens up a form of robotic action in which the robot not only grasps or cuts, but also actually perceives what is happening.

Fundamental research with meaning

This research is part of CogniGron, an interdisciplinary research programme at the University of Groningen, funded by the UEF. It shows how fundamental, brain-inspired research can contribute to more efficient technology and more natural interaction between humans and machines.

These are developments that require time, collaboration and trust. Step by step, technology is emerging that is not only smarter, but also better suited to the world in which people live.

This article is based on a CogniGron story and has been editorially adapted for publication by the Ubbo Emmius Fund.