Hi, this is Subham Datta from the Knott and today I am going to talk about something called neuromorphic engineering.
Recently, in a study, a team led by scientists from the University of Bath, in the UK described a procedure to make silicon chips that, as they say, can mimic brain neurons. According to the researchers, these chips one day can repair parts of our nervous system that are damaged by diseases and will also be able to help with paralysis. More importantly, the chips only need one-billionth of the energy required by a microprocessor which means that they can be used in medical implants.
But this concept of neuromorphic engineering is not exactly new. Back in the late 1980s, Carver Mead came up with the concept of using VLSI systems to mimic neuro-biological architectures in our nervous system. Back then, the systems were too large and they used tabletops to make simple neural networks with sensitive mirrors and lenses. However, they have passed a long way to finally reduce it to a chip that spans a few millimetres.
The new small silicon chip presently simulates a network of 16 neurons in four layers of four. Data enters the chip in the form of the light beam as optical computation is efficient and can travel and interact on their own. The light beam then split into four smaller beams. The brightness of each beam signifies a different piece of information. The processings are done while the beams cross and interact.
Initially, the researchers came out with a simulation to model how neurons in respiratory systems and hippocampus in rats response to stimuli. The models were then translated to silicon chips inside which they created replicas of biological ion-channels to transmit signals.
Theoretically, these artificial neurons should be able to replicate healthy functions of failing nerve cells to pass electrical signals to different parts of the body. For example, if an organ fails and neurons in the brain do not respond properly, these artificial neurons can fire the right signals to get the organ back on track.
Though they are accurately working on rats, it is a long way to go to actually come up with something for medical implant in humans. So let’s wait and watch.
That’s all for today and if you have any thoughts on Neuromorphic Engineering, don’t forget to write it in the comments section.
Thank you and Stay knotted.