Science Literature

More and more people are realising that the living world is like a treasure trove packed full with engineering marvels. The agenda of biomimetics is to actively research the potential of applications inspired by animals and plants. The human body supplies some of these design ideas, and the one considered in this blog concerns the inner ear, or cochlea. Undoubtedly, researching organs like this leads to a new appreciation of the sophistication of biological systems.

Sound is collected in the outer ear, travels through the ossicles (or bones) of the middle ear and delivers pressure waves through the oval window of the cochlea (the inner ear) (Source here)

The people involved in the newly published work describe the cochlea as "an amazing sensory instrument that transforms sound frequencies into spatially and temporally-varying excitation patterns of the auditory nerve". It is particularly interesting to electrical engineers because: "It performs this task over a wide range of input frequencies and amplitudes using very little power. In humans, the approximate values of these performance metrics are three decades, 120 dB, and 14 W, respectively". The mechanism is described thus:

"The cochlea is a hydro-mechanical system; incoming sounds set up travelling waves on the basilar membrane (BM) and in the fluids that surround it. The properties of the BM scale approximately exponentially with position: The membrane gradually becomes wider and less stiff, and resonates at lower frequencies. Thus, high frequency sounds excite responses towards the beginning, or basal part, of the cochlea, while low frequency sounds excite responses towards the end, or apical part. In other words, the cochlea uses a frequency-to-space transformation to perform audio spectral analysis."

There is broader medical interest in any research with potential relevance to problems of deafness. Cochlear implants have been available for some years. They give some benefit to people whose hearing is impaired, but do not restore normal hearing. Electrodes are implanted to stimulate the the cochlear nerves using electrical impulses generated from sound reaching the subject. For an brief overview of the history of cochlea implants, go here. For more technical information, a paper by Kissiah (2007) is helpful.

However, medical applications are beyond the scope of the reported research. The researchers have their sights on constructing an electronic cochlea - not implanting electrodes but making a complete system. They are interested in radio-frequency devices, not audio-frequency applications. Their goal is to achieve a high speed of operation, an ability to handle a wide range of input frequencies and a reduced power requirement in use.

"The human ear is a very good spectrum analyzer," said Rahul Sarpeshkar, a professor at MIT who co-authored the paper [. . .]. "We copied some of the tricks the ear does, and mapped those onto electronics."
[. . .] To detect electromagnetic waves instead of pressure waves the MIT scientists used circuits, in place of cilia. Starting on the outside edge of the 1.5-mm by 3-mm-chip are tiny squares, each one corresponding to a different size radio wave.
As they spiral into the center, the squares become larger and larger. The outer spiral detects the highest energy, shortest frequency waves, while the center circuits detect less energetic, longer frequency waves.

The team have produced an "electromagnetic ear". This detects a very wide range of frequencies with no more energy than is used by a typical cell phone. What they have done is to combine an analogue spectrum analyser with digital signal processing. This reduces the power requirement to about 1% of a purely digital system.

"A simple cell phone takes 300 millivolts to detect one carrier wave," Sarpeshkar said. "We can do all 50 carrier frequencies with 300 millivolts." Other devices do exist that can examine a range of radio frequencies. They just require much more power to do so. The low power usage of the electromagnetic ear means it would be ideal for portable electronic devices.

One assessment of this work is as follows: "People have tried to construct electronic cochlea before, but this is the first demonstration that imitates the amplification we think happens in the ear to produce a device that works." The team is now working on RF transmission as well as signal analysis, because this has the best potential for commercial exploitation.

Since the discovery of DNA, it has been increasingly popular to refer to life as "digital". Darwinists, particularly, have latched onto this concept, because their mechanism (incremental changes filtered by natural selection) can be understood in terms of digital mutations. Artificial life software like Avida is 100% digital, and enthusiasts consider that their digital world gives them the power to experiment in an unprecedented way. One researcher, Richard Lenski, is quoted thus:

"It's also the power to manipulate almost any variable one can imagine, to measure variables with absolute precision, to store information that then allows one to trace back a complex chain of events, and to take evolved organisms and subject them to new sorts of analyses that one might not even have anticipated when first collecting the data."

But what if life is both digital and analogue? What if analogue information is independent of digital information? These are questions that Darwinians do not ask because they stretch beyond the horizon of the gradualist paradigm. What if analogue systems point to complex specified information that cannot be modified gradually without ruining functionality? The case of the RF silicon cochlea is significant. This device was only developed by the focused effort of intelligent design engineers. When we consider the human cochlea, we need to ask the question whether it could have been developed by digital tinkering or whether the sophisticated design principles embedded in the physical structure of the organ point to an explanation involving intelligent agency.

A Bio-Inspired Active Radio-Frequency Silicon Cochlea
Mandal, S.; Zhak, S. M.; Sarpeshkar, R.
IEEE Journal of Solid-State Circuits, June 2009, 44(6), 1814-1828 | doi: 10.1109/JSSC.2009.2020465

Abstract: Fast wideband spectrum analysis is expensive in power and hardware resources. We show that the spectrum-analysis architecture used by the biological cochlea is extremely efficient: analysis time, power and hardware usage all scale linearly with N, the number of output frequency bins, versus Nlog(N) for the Fast Fourier Transform. We also demonstrate two on-chip radio frequency (RF) spectrum analyzers inspired by the cochlea. [. . .] Our work, which delivers insight into the efficiency of analog computation in the ear, may be useful in the front ends of ultra-wideband radio systems for fast, power-efficient spectral decomposition and analysis. Our novel rational cochlear transfer functions with zeros also enable improved audio silicon cochlea designs with sharper rolloff slopes and lower group delay than prior all-pole versions

See also:

Bland, E. Human ear inspires universal radio antenna, Discovery Channel (June 8, 2009)

Is Life Analog or Digital? A Question for Edge discussion group from Freeman Dyson (2001)