Researchers are developing flexible neuromorphic electronics that imitate the brain’s information-processing abilities, offering a potential alternative to rigid silicon-based AI hardware and enabling deeper human-machine integration.
Modern artificial intelligence systems can already surpass human performance in areas such as image recognition and medical data analysis, yet they still face a major limitation when it comes to operating inside the human body.
The core challenge is straightforward: human tissue is soft, elastic, and in constant motion, while traditional electronic hardware is not. Even the most advanced silicon-based chips are rigid, which makes seamless integration with organs, muscles, and skin extremely difficult. As a result, devices placed on moving parts of the body—such as a beating heart, expanding lungs, or flexing joints—can cause irritation, lose stable contact, and ultimately fail over time.
Researchers are now exploring a fundamentally different strategy. Rather than making the human body accommodate electronic devices, they are designing electronics that can adapt to the body.
A review in the International Journal of Extreme Manufacturing describes the development of soft neuromorphic electronics—a new category of systems that integrate sensing, memory, and computation within materials capable of stretching, bending, and conforming to biological tissue. These devices draw inspiration from the human brain, not only in how they process information but also in how they physically interact with their environment.
Brain-Inspired Electronics
Unlike conventional circuits, which rely only on electrons moving through rigid metal conductors, these systems use soft materials such as flexible polymers and gel-like ionogels that transport both electrons and ions.
This process, called organic mixed ionic-electronic conduction, more closely mirrors the electrochemical communication found in the nervous system. The active materials are capable of taking in and releasing ions from their environment, which allows their internal electrical properties to continuously change and adapt.
As a result, a single soft transistor can emulate synaptic plasticity—the biological process by which connections between neurons are strengthened or weakened over time. In this way, the hardware itself can display behaviors that resemble the brain’s own learning mechanisms.
Flexible and Energy-Efficient
Recent progress in materials science has enabled these devices to achieve remarkable flexibility. Certain components can stretch up to 140% beyond their original length, exceeding the natural elasticity of human skin and allowing reliable performance even in highly dynamic areas of the body.
These systems also require very little power to function. Instead of relying on high electrical currents, they use efficient electrochemical mechanisms to carry out complex functions, such as heart rhythm classification, at operating voltages under 0.5 volts.
Operating at such low voltages reduces both heat production and electrical stress—two essential factors for electronic devices intended for continuous use in direct contact with biological tissue.
This technology could also transform the way engineers design wearable devices. Instead of attaching rigid sensors to flexible materials, engineers could directly print soft, integrated computing networks that merge sensing, memory, and processing into one stretchable medium. Such a design could make it possible to create electronic skin and soft robotic limbs that process touch and motion data locally, without needing to continuously transmit information to an external computer.
Advancing Beyond the Laboratory Setting
Despite significant progress, major technical challenges still need to be addressed before soft neuromorphic electronics can be used in clinical applications.
A key limitation is memory stability. Many existing soft memory devices are unable to retain information for long periods, as stored signals tend to fade quickly once the input stops, restricting their effectiveness for long-term data storage.
To overcome this problem, researchers are developing island–bridge architectures. In these designs, engineers place stable memory elements on small rigid “islands” that protect them from mechanical deformation, while flexible coiled connectors link the islands together and enable stretching.
Scientists suggest that integrating this approach with chemically stable and biocompatible materials could offer a viable route toward long-lasting neuromorphic devices that can safely interface with the human body over extended periods.
Read the original article on: SciTechDaily
Read more: Researchers unveil a floating buoy engineered to repel jellyfish