Simulated Annealing (SA) is a probabilistic method calculating the global minimum of an objective function that may possess several local minima. Basically it works by emulating the physical process whereby a solid is slowly cooled so that when eventually its structure is “frozen”, a minimum energy configuration is reached.

Quantum Annealing (QA) has the same goal of simulated annealing, i.e. finding the global minimum of a given objective function over a given set of candidate solutions (in a certain search space), but this time the adopted process is based on quantum fluctuations. A “state” (i.e. a candidate solution) is randomly replaced by a randomly selected neighbor state if the latter has a lower “energy” (i.e. the value of the objective function). The process is controlled by a “tunneling field strength”, a parameter that determines the extent of the neighborhood of states explored, i.e. the mean distance between the next candidate state and the current candidate state. In other words QA has the possibility of quantum tunneling across the width of a barriers between states, instead of just scaling their heights (as in simulated annealing) with a thermal jump.

There is theoretical and experimental evidence of the advantage of solving classical optimization problems using QA over SA.

In 2011, D-Wave Systems announced the first commercial quantum annealer: D-Wave One. The company claims this system uses a 128 quantum bits (qubit) processor chipset, specifically designed for quantum annealing. Essentially, it involves preparing some sub-group of the qubits into their lowest-possible energy state, or ‘ground state’, and then performing a series of operations to put it into a more complex ground state that can’t be easily solved using classical methods.

It has been recently announced by Nature News Blog that D-Wave quantum computer can solve the protein folding problem, which is very complex (for traditional computation). The model consisted of mathematical representations of amino acids in a lattice, connected by different interaction strengths. The D-Wave computer found the lowest configurations of amino acids and interactions, which corresponds to the most economical folding of the proteins (the global minimum of the free-energy function is conjectured to be the native functional conformation of the protein).

Quantum Device and Qbit configurations

Actually, according to the paper, 10,000 measurements using an 81 qubit version of the experiment gave the correct answer just 13 times (checked with a conventional computer). Even though this problem still can be solved on a classical computer by exact enumeration, it is remarkable that the device anneals to the ground state of a search space of 2⁸¹ possible computational outcomes.

This study provides a proof-of-principle that optimization of biophysical problems such as protein folding can be studied using quantum mechanical devices. This paves the way towards studying optimization problems using quantum devices in several domains, also in future Internet (see my next post on that).

The Moore’s law has its days numbered if we were stuck with silicon. The progress will be reaching a thresholds by the end of this decade, probably sooner. But there are other avenues to explore, and indeed researchers have already something up their sleeves.

Rendering of 2D molybdenum disulfide structure

One option is to use graphene, a particular form of carbon in a 2D structure. Carbon can provide better performances than silicon but it is trickier to control. That is why, in spite of the success in creating transistors made with graphene the production is still at the starting blocks.

Another avenue, just opened by research at the MIT, as reported in an article published on NanoLetters, is to use molybdenum disulfide. This material has similar property of graphene but works around some of its problems making production of chips easier.

According to the MIT researchers, by using this substrate it would be possible to create chips as large as a whole wall, thus using your room wall as a screen, or using the surface of your cloth fabric as a computer.

In perspective, any surface will become a screen, will have interaction capability and will be able to process and store information. This is going to take probably a few decades, but the first examples will be available sooner, by the end of this decade. So, in a way, the iPad of the future will be any surface, and it will be your personal “iPad” since that surface will be able to recognize you and will customise interactions, information and services to you!

As we understand more and more about the code of life, we are now becoming able to identify larger building blocks, each one enabling a specific functionality, be it the luminescence of a cell once a substance is detected in the environment, or the transformation of hydrocarbon into CO2 and water.

This has enormous implication on our ability to create new substances or to improve existing process by “teaching” bacteria to do what we want.

Biocounter schematics, Pamela Silver Openwetware

This has given rise to a new science, synthetic biology. And this is what Pamela Silver at Harvard Medical School is doing.

The goal is to be able to first model through a computer the functionality of basic blocks, then to assemble these blocks in the computer modelling to create more complex functionality and eventually to create these functionality by coding the genes and inserting them into bacteria.

This is already done today by “splicing”, taking a gene sequence from one cell and inserting into another, thus transferring the coding instructions.

With synthetic biology it becomes possible to create a gene from scratch, after having modelled its desired functionality. The goal is quite ambitious, and fraught with big question marks, even bigger than those we already have when splicing is done. In this latter we know what that “code” is doing in the original cell but do we know what it will be doing, as potential side effect, in the spliced cell?

In the former we know, after computer simulation, what that coding is going to do (create a specific protein) and the resulting “main effect” but what about unexpected side effects? These are no minor challenges but it is something researchers are studying because the potential benefits are enormous.

Now, looking at these LEGO bricks, if I may call them so, we are also looking at “information” . And there is going to be plenty of information. In the same way that by surfing the web we can get information and use it, in the future we will be able to surf the multitude of code of life and use them. It may take another decade, or even more, to have something comparable to the World Wide Web in biology, but my guts tell me it is going to happen.

Our nose provides smelling sensations that although often underestimated are important to our everyday life. Of course the “sensitivity” of our nose is really low in comparison with a dog’s nose. Still, replicating the capability of sniffing and smelling in a computer (equipped with sensors) has been a big challenge to researchers.

Nanotechnology based sensors on a chip to power an electronic nose

Over the years researchers have been able to create artificial, electronic, noses that can detect specific odours but not a broad range of them.

Now it looks like we are getting closer to “iSniffing”, an artificial nose that can exceed our nose sensitivity and that can be applied in a variety of fields.

Nosang Myung, a professor at the University of California Riverside, has created a sensors based on nanotechnology that has a broad and good sensitivity to odours.

Olfactor Laboratories, a US company, has developed the chip. The whole system is about 4 by 7 inches but the goal is to reduce it to the size of a credit card. The chip as such has been designed to be fit for embedding in a cel phone or a tablet.

The applications can range from agricultural (detecting pesticide levels), production in industry (detecting leaks, emissions) and also as warning system for bio-terrorism.

It may also have medical applications, such as studying children’s asthma for correlation with the insurgence of symptoms as a function of air pollution.

In the future we may expect to see cell phones equipped with these sensors, like today we have got used to have gyroscope embedded in out cell phones. Robots are also like to use these sensors to get a better perception of their environment.

Localizing a source of information in a complex network, and the way it diffuses, is a desirable but rather challenging task. This could have several interesting applications: imagine, for example, understanding how a disease spreads through society, or how a gossip diffuses through a social network or how a virus spreads over the Internet. 

Actually, most of today methods have been focused on the so-called forward problem of epidemics. Consider for example a disease: there are many models representing the diffusion process and its dependence on the rates of the infection and the cure, as well as on the network structure. Common to these models is the need of knowing the states of all nodes of the networks under consideration…and the networks might be very large. Interestingly, in this letter, they have solved problem changing the point of view: inferring the original source of diffusion, given the infection data gathered only at some of the nodes in the network. Advantages of this method are good localization accuracy for much lower computational costs (even for large networks).

Given size of real networks it is usually almost impossible to observe the state of all nodes, so researchers of the École Polytechnique Fédérale de Lausanne have developed an algorithm allowing to estimate the location of the source from measurements collected by sparsely-placed observers.

In the picture, for example, there are three observers, which measure from which neighbors and at what time they received the information. The goal is to estimate, from these observations, which node the information source.

“Locating the Source of Diffusion in Large-Scale Network”, Pedro C. Pinto, Patrick Thiran, Martin Vetterli, École Polytechnique Fédérale de Lausanne (EPFL).

In particular, they have demonstrated the effectiveness of their method using data about a cholera outbreak in the KwaZulu-Natal province in South Africa in 2000. They say: “By monitoring only 20% of the communities, we achieve an average error of less than 4 hops between the estimated source and the first infected community“. 

In summary, this method may provide an effective alternative to the monitoring of all nodes states, which is not practicle or even feasible for large networks. So it is very promising even if there are still two critical points to be improved: it may be difficult to determine exactly the underlying network graph; and the most effective choice of observers in the network which strongly affects the performance of the proposed algorithm.

Just imagine, what could be more Nature friendly than a car moving on air … as a fuel? Seemed impossible to me but I have learnt to be prepared for the unexpected!

Tata Airpod car, running on air!

Tata has announced a prototype of a car, the Airpod, that use a tank full of pressurised air to move piston in the engine. The tank is 175 litres and can be filled by special refuelling stations or by using an electrical motor to pressurise the air.

A full tank will let you drive for 200 km and a refuelling cost about 1 euro! That’s less than 1/20th of the cost of a gasoline minicar.

The cost of the Airpod is estimated in 10,000$ and of course such a car would be  emission 0!

It is interesting to discover every day that some new avenue can be explored to take us into new solution spaces.

Now, I do not expect to see this car anytime soon in my Country but just the possibility of the existence of sic a car makes a big difference to me.

Of course, it is not a completely 0 emission system, you need to take into account where is the power for pressuring the air coming from. If it comes from electricity being produced by coal there you have an emission… And even if it comes from a photovoltaic systems you should take into account the emissions generated by its production…

Still, we have found a new path worth exploring, a path I didn’t know could ever exist.

Take a look at the CNN video showing the car … in action:

Printing a home from scratch in a day

3D printing has been mentioned in this blog several times and it is actually recurring more frequently in this last year. It is an indication of the progress in this area, a progress that we foresaw at the Future Centre several years ago when we decided to study how the world will change, in its distribution and customisation aspects, once 3D printing would become commonplace.

However, we did not envisage, at that time, that one day we might print “our home” and do that in less than a day!

This is the story that is being told in a recent TED presentation by the USC professor Behrokh Khoshnevis who is part of the Center for Rapid Automated Fabrication.

At the Centre they are working on using special 3D printing technology that can be applied to the construction of buildings, resulting in fast and cheap construction. It would help in solving problems in many world developing countries and as a prompt intervention in case of extensive building damage such as the ones occurring in earthquakes.

The 3D printing machine ejects concrete at a pressure of 10,000 PSI, that is 3 times as much as the one you find in normal concrete structures. Since the printing is made layer by layer under the direction of a computer there are really no limitation to the kind of shapes that can be made.

You should take the time to look at the TED presentation:

Image advertising the Tawkon App

Few days ago I run into an article on Wired talking about a new App released by Tawkon that is providing information on the level of radiation you have been exposed during the day (or week, month …  whatever) by using your cell phone.

The article provides an explanation of some basic terms like SAR and also some background on the potential effect of electromagnetic radiation.

I am not an expert in the field, although I have been working for several years with researchers studying SAR for cell phones and taking care of the scientific communications, so I do not want to comment on the potential damage, if any, resulting from exposure to cell phone radiation.

What I am interested in is the thin line existing between data and information. Unfortunately, most of the time we tend to mix the two, we perceive data as information and this is misleading.

In this specific case Tawcon App is measuring the electromagnetic field of your cell phone taking data from the cell phone chips (your cell phone has to know what kind of power to use to send signals to the remote antenna) and applies an algorithm to calculate the cumulative effects. However, there is no scientific proof indicating that electromagnetic radiations have a cumulative effect on bio-material (that is your head..) and the cell phone does not know how you are using it (e.g. you may have it in contact always on the left ear, sometimes you switch ear, sometimes it is not in contact at all …).

Hence, the data accumulated are basically meaningless. Tawcon is careful to state that the apps is just to warn you of the use of your cell phone and it is up to you to decide if that usage level is too much (too dangerous). But if the data are meaningless, what actually is the information provided? Zilch!

What is wrong, in my opinion, is that by providing data we create a perception that does not correspond to a scientific fact. And this goes both ways. You may scare people, or you may provide a false sense of assurance.

I guess this is one of the big problems we have always faced, but in the Internet age with the abundance of data the risk of being misled is even greater.

Well, you still need to buy a ticket if you plan to fly oversea, or even to some nearer place. But now you can use 3D printing technology for printing a model aircraft!

Maker Plane model aircraft to be printed at home with a 3d printer

According to a news on Engadget Maker Plane is now offering complete instructions for a 3D printer to print a model plane.

What I find interesting (I am not into aircraft modelling at all) is the potential behind this progress.

First, this news shows that 3D printers are now within the (economic) reach of the mass market (as it happened for the paper printers years ago – you may not remember but there was a time, in the seventies, that printing was our of reach for consumers, since printers were far too expensive…).

Second, there is the capability to download printing instruction thus cutting the cost of mailing and opening up the door to customisation.

Third, it is very easy to imagine that crowdsourcing will quickly step in, letting people collaborate to develop products. This may change the rules of the game in many sectors. As well as opening up new biz opportunities.
You no longer need to set up a shop to produce goods for sale, rather you can lean back on your chair, design your product at the computer and then … sell it. It will be “manufactured” at your client’s home.

We are already seeing this happening in the books area, iBook is a clear example of proving do it yourself tools for creating a book. We have been discussing this 5 years ago at the Future Centre and it is nice to see it happening now.

The idea of a humanoid robot, spread during the last century by technology advance and science fiction writers, is still pursued by several research labs and we are seeing some experiments in interaction with toddlers, kids and elders.

We are so captured by the imagination of a robotic alter-ego that we are missing to perceive that robots are already among us. In Turin we have a robotic subway, in some surgery operation robots are performing ever more complex procedures.

Honda's Asimov presenting the gardening robot Miimo

But what is really interesting is to see robots becoming parts of our life. That is the case with Roomba, the robot-vacuum cleaner, and it is the case with the new Honda Miimo, our future gardener.

Miimo is already in production and will go on the shelves early next year. You just leave it in your garden and it (he?) will take care of trimming the grass, 3 mm at a time for an always perfect lawn. It will do that several times per week and the clippings are so tiny that can act as fertiliser.

It will find the perfect pattern for mowing, although you can instruct it to follow certain direction. If it steps onto higher grass it will keep mowing, 3 mm at a time, till it has got the perfect lawn. If it encounters an obstacle it finds a way to work around it.

It also learns where the powering station is and when its batteries run low it will go back to recharge.

Get ready to buy one, starting next year. The price? 2500 euros for the bigger model, fit to manage 3000 square meters of lawn. If you have a smaller patch you may want to consider its little brother, selling at 2100 euros.

I think this is really showing that robotics are becoming part of our environment, to the point that in a few years we will take them for granted. Technology disappears from our perception radar when it is successful.