Researchers have developed a brain-like computing device that can learn by association.
The secret of the device lies in its new organic synaptic transistors, which simultaneously process and store information, just like the human brain. It is noted that the transistor can mimic the short-term and long-term plasticity of synapses in the human brain, relying on memories: this is how it learns over time.
Due to its similarity to how the brain works, the new transistor could potentially overcome the limitations of traditional computing, including its power-hungry hardware and limited ability to perform multiple tasks at the same time. The device is also more resilient: it continues to run smoothly even when some components fail.
The modern computer is an outstanding invention, but the human brain can easily surpass it in some complex and unstructured tasks such as pattern recognition, motor control, and multisensory integration. This is due to the plasticity of the synapse, which is the main building block of the brain’s computing power. These synapses allow the brain to work in a parallel, resilient, and energy efficient manner.
Jonathan Rivney, Senior Study Author
In conventional digital computing systems, there are separate processing and storage units, as a result of which tasks that require large amounts of data consume large amounts of energy.
The memory resistor, or memristor, is now the most advanced technology that can perform combined processing and memory function.
However, memristors suffer from energy intensive switching and less biocompatibility. These shortcomings have led researchers to create a synaptic transistor – a special organic electrochemical synaptic transistor that operates at low voltage, has continuous memory rebuilding, and has high compatibility.
Next, the developers optimized the conductive plastic material inside the organic electrochemical transistor, which can trap ions. It is a structure through which a neuron can transmit signals to another neuron using small molecules called neurotransmitters. In a synaptic transistor, ions behave similarly to neurotransmitters, sending signals to form an artificial synapse.
The authors emphasize that the synaptic circuit is made from soft polymers such as plastic, so it can be easily fabricated on flexible sheets and easily integrated into soft wearable electronics, intelligent robotics, and implants.