Nature has kept evolving for at least 4 billion years on Earth, transforming random interactions into progressively more complex “random but probabilistically directed” interactions that all together create emerging behaviours that in turn increase probability and in a way decrease randomness.

The basic steering force in this evolution can be seen as Darwin said “the survival of the fittest” or as other biologists put it “the success of the most adaptable”. A physicist would probably say “the drive towards lower and more efficient power consumption against the second law of thermodynamics”.

Indeed, if we look at biological systems we see an amazing success in minimising the power requirements, from the flight control system of an insect that selectively activates just what’s needed in a specific moment to the automatic temperature control in termite nests.

A network of hundreds or thousands of dissociated mammalian cortical cells (neurons and glia) are cultured on a transparent multi-electrode array. Activity is recorded extracellularly to control the behavior of an artificial animal (the Animat) within a simulated environment. Sensory input to the Animat is translated into patterns of electrical stimuli sent back into the network. (Credit: Thomas B. Demarse et al./Autonomous Robots)

Engineers are trying to reduce as much as possible, nowadays, the power budget in their systems and the overall power budget in system aggregation. This latter is much more challenging since the aggregation results from a multitude of systems, each one optimised but those optimisations when aggregated do not necessarily result in an overall optimisation. We need to move from local optimisation to an overall optimisation and that requires that each individual system can evolve and adapt over time. A very big challenge indeed. So why not learn from Nature that had billion of years to perfect strategies and went through billions of missteps eventually coming to good solutions?

This is what many scientists are actually doing. More specifically, this post is originated by having read a news from the National Science Foundation reporting on the work of a team at the   Real Power and Intelligence Systems Laboratory at Clemson University.

This is a team of neuroscientists that have decided to approach the problem of controlling the complexity of electrical grids using live neurones grown in a culture dish. By leveraging the ability of neurones to process complex data (and understanding patterns) the neuroscientists hope to create a “smart grid”.

According to Venayagamoorthy, the team lead researcher:

“What we need is a system that can monitor, forecast, plan, learn, make decisions. Ultimately, what we need is a control system that is very brain-like. The brain is one of the most robust computational platforms that exists. As power-systems control becomes more and more complex, it makes sense to look to the brain as a model for how to deal with all of the complexity and the uncertainty that exists”.

The Sun baths our Planet with energy, every single day since the birth of the Planet and will do so for billions of years to come. All life on Earth depends on this energy and is in equilibrium with it and the life capacity to harvest part of this energy.

Science and technology have shifted this equilibrium by adding artificial capacity to harvest several sources of energy (although most eventually can be tied to the Sun, like hydroelectric, wind, oil…). The human kind is now using more and more energy to multiply its own activities and in doing that is depleting what has been accumulated over million of years.

A rendering of the solar mirror proposed by IBM researchers

Still, the energy we receive every single day from the Sun greatly exceeds our present demand (and the one in the foreseeable future).

Hence, one possibility would be to enhance our capability of harvesting the solar energy reaching us. Just 2% of the solar energy received by the Sahara desert would satisfy our present demand.

One of the problems is how to increase present efficiency in energy capturing and conversion. And this is what is being addressed by researchers at IBM and other companies in Zurich.

They have been awarded 2.4 million $ over the next three years to create a system capable of converting up to 80% of the solar energy received by a mirror into usable forms of energy. This grant has been obtained by demonstrating a prototype where several advanced technologies are being leveraged to reach this kind of efficiency.

Today conversion efficiency is below 20%, so this would represent a tremendous increase able to lower the cost of solar based energy to 10c per KWH, which compares to the production cost of electricity from coal and oil (the cheapest source of energy).

This feat is achieved by a mirror composed by many individual mirrors that can focus the incoming light to a point that the receiving photovoltaic panel gets 2,000 times the energy irradiated by the Sun on that specific surface. This is why they label this project as “the energy of 2,000 Suns” (which of course is not true but makes for good headlines…). The whole reflector and the individual mirrors are controlled by a computer that orients them in such a way to maximise the harvesting of light beams.

The problem is that such an intensity would heat the receiving panel surface to such a degree that it will stop functioning soon and start to melt.

This is where another technology is needed. The photovoltaic panel is composed by individual cells (chips) each one able to produce 250W and cooled by a fluid (water) using an approach similar to what is being used in our body by the ramification of vessels (capillaries) to bring oxygen to all cells and exhaust heat.  This is also been used by IBM to cool some chips using micro-channels.

The combination of these two technologies (and a few more – read the article) increases the conversion efficiency from solar energy to electricity to 30%. An additional 50% is converted in usable heat by leveraging the cooling fluid increase of temperature, yielding a total of 80% conversion, a staggering increase from present conversion efficiency.

I attended the European gathering of the STS Forum, Science and Technology in Society, in Paris, and the themes are as usual focussing on the big World Challenges that the Kyoto Agreements (not fully translated into concrete actions as of today, after almost 10 years…) aimed at tackling.

Water, food, population, energy. They are all different aspects of a single issues: our relation with the environment as a global race.

Technology has developed over the century to make it possible to do more, and in order to multiply the body capabilities it has to tap into more and more energy.

Its success has increased the chances of living (the average human life span has doubled thanks to technology in its many forms) and supported a multiplication of humans on our (only) planet ten folds in 300 years.

Technology success have brought forward even greater challenges to the point that technology alone cannot provide the solution. But technology is an essential part in any solution that assume as a global right the aspiration and fulfillment of a better life.

We are now in the need of changing some centuries old paradigms. Rather than focussing of producing “more” we should aim our endeavor to use less.

We should cut drastically the usage of water (this is not about cutting the water we drink but cutting the amount of water being used for crops), we should cut drastically the use of energy because the use of energy (second law of thermodynamics) creates heat and this destroys the Earth present equilibrium. It does not pay to create smarter technology to convert solar light into usable energy: it is true that the Sun is bathing the Earth with thousands fold more energy than what we are using. But most of this energy is reflected into spaces keeping Earth temperature and climate within the range suitable for current life forms.

If we harvest more, we dissipate more and this creates havoc. Cooling giant data centers by placing them in the sea, is not a free lunch: it heats the sea. You might think that would be a risible increase in temperature but it is not so: estimates indicated that if un-checked our use of energy will lead to a two degree increase of ocean temperature by the end of this century and that spells for disaster: as water temperature increases the CO2 capacity of absorption decreases (oceans are the greatest absorbers of CO2) and more CO2 will remain in the atmosphere leading to further increase in temperature in a never ending cycle that can bring our Earth surface to 100°C within two centuries.

Hence the real focus should be on usage, not on production.

As Richard Feynman used to say “there is plenty of room at the bottom”. Indeed, our bodies, as all living things, are thousands folds more efficient in using energy than our machines (chips included). To store a bit we use million folds more energy than what is required by physical laws. There is really plenty of space to decrease energy usage.

A drosophila fly image of neurones firing at “take off”

ICT is considered to be a key component and an important lever in this quest for efficiency.

But current ICT operates on inefficient platforms (chips) and we need to reinvent these platforms.

The graphene project of the European Commission may provide some basic building blocks but architectures are likely to be inherited by Nature, by observing and replicating Nature most advanced control systems, like the system to control the flight by a fly. A few thousands neurons using as little as   375nW  do the job (one might even say a better one) that the A380 flight computers that use  in the order of 1KW, that is 3 billion times more! And don’t be mistaken by the size of a fly vs the size of an A380 (an A380 weights 138.4 billion times more than a fly…)! We are talking about control system, not about the actuators energy requirement…. There is basically little reason why the two need different amount of power.

In the photo an image of neurones of the fly (a Drosophila) that fire immediately at the take off… The insect’s brain activates the eyes to focus on the take off path to look for obstacles. Insects eyes are composite and to save energy only part of them are active at any given moment. So the brain provides the instructions needed to ensure a safe take off. Quite sophisticated system, isn’t it?

As I reported in the very beginning, technology alone cannot provide the answers. We need the contribution of humanities, economics and social sciences to reach a sustainable answer. What has been stated is the need for a new Renaissance, where these different aspects are not evolved in parallel but together. We need ICT schools that teach much more than ICT and we need humanities and economy schools that teach ICT.

A thought to be considered by the ICT LABS. It is now starting to be considered by the IEEE at the Future Direction Committee that I chair.

Leading in technology is not going to lead “anywhere”. But leading without technology is simply impossible.

Electronics is slowly but steadily make ways in our bodies. Microchips to release drugs, sensors to detect a variety of conditions are already part of the landscape.

The crucial problem to solve remains the powering of these devices. As electronics evolution (Koomey’s law) keeps decreasing the demand for power new approaches are becoming feasible, such as the one proposed by Jia Hao Cheong, a professor at the A*START Institute of Microelectronics in Singapore.

He is addressing the problem of monitoring grafts in the body that surgeons are implanting to create an artificial blood vessel (or substitute a faulty one). The problem with these grafts is that they may become clogged and it is important to monitor what is going on. The solution proposed is to insert in the graft material nanowires that change their resistance as they are stretched, somthing that happens when the blood pressure increases. A sensor can measure this pressure increase by detecting the change in resistance of the nanowires but of course the sensor needs to be powered.

Cheong has found the solution by sending the power wirelessly from the reading device that the doctor holds in her hand, like it is done when reading RFID tags.

As shown in the figure the graft (in this case used in people with kidney failure that need to have dialyses every few days and have a graft implanted to avoid pricking the veins with possible damage and infection) is being monitored by a tablet like device that is sending wireless power sufficient to activate the sensor for both reading the strain in the graft and sending the information, wirelessly, to the tablet.

The power transmitted by the tablet is absorbed also by the body tissue but it is sufficient to power a 12.5 microW sensor up to a depth of 50mm, which is plenty for a graft inserted in arm.

Solar panels are “old stuff”. They have been around for quite a while and scientists have tried to increase their efficiency in converting light energy into electrical energy. Progress have been made but we are still “wasting” a lot of light energy.

The trade-off is between cost and efficiency. Engineers have been able to create much more efficient solar panels but their cost makes them unpractical. Once in a while we get the news of another ingenious way to increase efficiency, like this one developed by the EPFL (Losanne) and the Niels Bohr Institute in Denmark.

They have discovered that nanowires are very good in focussing light, reaching an intensity 15 times greater than the one in normal solar panels.

Nanowires are smaller than the wavelength of light and this creates a resonance that in turns results in greater focussing capability. You need 10,000 nanowires to get the size of a hair.

The problem, as in several other discoveries, is to translate this into an economical industrial process and the jury is still out to see when this will be possible.

On the other hand, this adds to the list of wonders that can be achieved working at the nano level…

“Information transfer in nanoscale structures, by means of surface plasmons, is referred to as plasmonics“

This is what you get when you look on wikipedia to learn about plasmonics. Actually you get more but that “more” is even MORE complex than that sentence. SO let me try to explain it in more accessible, may be not completely correct, terms.

Plasmonics is a new branch of physics that studies what happen to electrons as they move at the edge of two materials. If these two materials are not the same, from a physical conductivity point of view, some strange things happen. There is no more the usual conductivity that you see in a copper wire where electron moves pushed by an electromagnetic field from one point to the next. Their movement is affected by the diversity of the two materials. They start to oscillate and out of this oscillation, that depends on the characteristics of the two materials you get different transfer of energy.

And here comes the interesting part. When a photon hits the surface its energy is transferred to an electron that being at the  edges of two materials start to oscillate and to transfer that energy in very specific ways. By finding out the right materials we can force the electron to transfer the energy in ways that are more efficient to our goal, in this case converting sun photon energy into electrical energy.

This is what a research team at the University of Toronto managed to do.

The researchers developed special quantum dots (the same I have been mentioning in some of these posts as a new technology for high resolution bright display) made by nano gold particle on a thin film.

This film overlaid on a photo voltaic solar panel can increase its efficiency by 11%, which may not seem to be such a progress but advances in energy conversion efficiency are painfully slow. The second law of thermodynamics is there to make everything more complex…

So, it is just another step, although a little one, on the way to harvest the energy the keeps bathing our planet.

A team of researchers at the Northwestern University and University of Illinois have demonstrated a stretchable lithium battery that can expand up to 300% of its linear size.

The trick is made possible by connecting the various components through tightly S shaped wires so that when the battery is stretched the wires unfurl, like yarn unspooling.

A battery is composed by many components (hundred in the prototype) that have to be stacked one on the other and this can also be done by connecting them sequentially by wires. Each component is tiny and flexibility is achieved by making the interconnection flexible. The connection is made through an S shape that in turns contains many S wiggles. Once stretched first the big S gets streamlined then the wiggles inside. It is like having springs within a spring. As the stretching subside the springs bring the wires back into their original shape.

The team is working on flexible electronics to fit wearable as well as for implants in a human body. A flexible battery was missing.

They tried it out on an elbow, a typical area where you get continuous alteration in shape and where flexibility is a must. The battery can be recharged wirelessly, using an antenna in the shape of a coil so that the antenna too can stretch.

Writing at the computer as I am doing now on a table having a recharging pad on its surface would recharge the battery in my elbow…

Take a look at the full paper.

The future will see a lot of embedded electronics and part of this will have to live inside stretchable materials, including our bodies. Hence the interest for this kind of research.

Graphene has lot of promises. A new European flagship project has just been launched with an investment exceeding 1 billion € over a seven years period.
Scientists now have a good understanding of the properties of graphene and the potential in several areas. A big challenge is to create industrial production processes that would support massive applications in several fields at a low cost.

This is where the work of UCLA researchers comes in. They have developed a process to create supercapacitors based on graphene that piggy-back on the industrial processes used to produce DVDs.

Supercapacitors are devices that can charge and discharge hundreds, and even more, times faster than normal batteries. Using graphene is a way of creating such a supercapacitor, by exploiting the amazingly good conductivity properties of graphene.

Most supercapacitors developed so far are very expensive and bulky. What the UCLA researchers wanted to do was to create micro-supercapacitors that could be used to power tiny devices, like a pacemaker, and of course they wanted a production process dramatically decreasing their cost.

One of the big challenges faced in electrical energy storage is that size matters and batteries do not scale down. By using graphene on the contrary it is possible to scale down graciously.

So graphene it is. But the traditional method for producing and assembling graphene into meaningful, and useful, structures are costly (based on the same lithographic processes used for chips). UCLA researchers have been able to use modified DVD burners to create graphene supercapacitors thus lowering the production cost whilst keeping the mass production capabilities.

We can expect graphene to permeate more and more our everyday world, sensors are likely users of these supercapacitors and I would expect to have many surfaces covered with a one atom thick graphene layer providing electrical properties to many commonly used objects. Notice that one atom layer of graphene is completely transparent so you won’t be able to notice it. A wooden table top will look and feel exactly a wooden table top, even though it might double up as a display screen you can touch and interact with…

Just yesterday I reported of scientists studying ways to harvest light energy and fighting the low yield of the technology used.
Today I can report on another team of scientists that have discovered a way to improve energy conversion efficiency by pairing carbon and silicon.

Silicon based photovoltaic panels are a reality although their efficiency is far from stellar and their cost is still hampering a massive adoption.

On the other hand carbon based photovoltaic panels (as the layers described in the previous post) have even lower efficiency.

Yale researchers have discovered that by layering a carbon nanotube film on a normal silicon photovoltaic panel one can increase the efficiency to 11% (over the magic 10% figure).

The nanotube layer sort of focus light in a way that it can be better harvested by the underlining silicon panel.

In the drawing a rendering of this idea, with the carbon nanotube layer (above, in white) capturing stray beams of light, photons, and directing them to the silicon panel for transferring their energy to electrons.

We already have the industrial manufacturing technology for coating surfaces with nano films, so we are really close to manufacture this kind of solar panels.

We love light, and we make sure to have big windows to get it in hour homes. Although we seldom realise it, we are actually interested in just a small portion of light, the ones colouring the rainbow.

Our eyes cannot see ultraviolet light nor infrared light. Hence, we have no us for it.

Here is where Ubiquitous Energy, a US start up, comes in. They are trying to create transparent layers (that is transparent to the wavelength of the rainbow…) that can absorb ultraviolet and infrared lightwaves converting them into electrical energy.

Such layers over  a window will in no way obscure the light we are interested in since the light waves blocked are the ones we won’t be able to see anyhow.

The layers are made up by organic molecules that can convert specific wavelengths into electrical energy and in principle they could be placed on any type of surface.

There are still significant challenges ahead: their transparency to the rainbow colour is of the order of 70% (a window glass has a transparency between 55 -very poor- and 90% -very good) so it actually blocks a bit of light and their efficiency (the ability to convert light energy into electrical energy) is around 2% whereas the target should be 10%.

Still, in these years we have seen scientists learning more and more how to develop materials with specific characteristics and I trust it is just a matter of time before they can manage to increase both efficiency and transparency. Once they do that the process for coating glasses is already in place, since today most glasses are chemically coated for a variety of reason, like decreasing glare, becoming dust repellant, …

There is so much energy permeating our ambient, if we can manage to harvest part of it we will be able to decrease the cost of creating smart, self aware ambient. The decreasing power budget of electronics will also make any tiny energy scavenging useful. Here we really see a convergence that delivers: less power consumption and increased ambient energy scavenging.