Adrian Cheok, from University of Singapore and Keio University of Japan, delivered an inspiring keynote at the Mobile HCI Conference early this month in Stockholm. His vision is that communication will move from informational to experiential. Experiential communications use visual, auditory and tactile imagery to capture attention, stimulate the senses in many ways: physically, rationally and emotionally. Adrian showed a number of prototypes on experiential communication that he is working on with his team:  e.g. giving hugs to your pet friends, hugging distant people (as explored in the Huggy Pajama project), enriching paper communication. He also showed examples of communication involving food printing, as part of his project called Food Media. The project is exploring new digital food and cooking media for virtual and remote communication in order to enhance remote family communication networks and allows for novel intergenerational interactions between family members through food and food activities. One platform is based on a

3D food tele-generating printer that transforms eating into an interactive communication experience by creating edible layered messages. It produces food by printing layers of flavored agar pixels that are liquefied by a silicon heater and then solidify upon contact with the environment. The system prints the layered information according to drawings, photos, or texts submitted by the sender.

Personally I would be even happier if this way I could plan and cook my dinner remotely from my computer in a way that just after being back home from gym or work, all I have to do is setting my table!

3D food printer

When we switch on a lamp or a computer an electrical current is involved: electrons flow and activate components, make transistors work. Living beings also use a sort of electricity for communications and activation but this electricity is not based on electrons (that are difficult to control) but on ions and in particular protons (in practice they use ionized hydrogen that when stripped of its electron becomes positively charged since all what remains is a proton).

Proton pumps are used by our stomach to activate cells producing acid (and every day I am taking a pill to stop these proton pumps since acid is not good for my stomach…), protons are used to have synapses clicking and therefore to make our brain work and let us feel…ourselves, thinking machine.

Interacting with our body with electrons is feasible (think at the pacemaker) but if we were able to interact with proton then we would be able to talk to cells and individual cells directly.

The scheme of a proton based chip

This is now becoming feasible with the result obtained by scientists at the University of Washington.

They have created a FET (Field Effect Transistor) that can send pulses of protons using a special substance, chitosan, extracted from the shell of squids or crap shells and modified to fit the bill.

This substance, has been discovered by University of Washington scientists is very good at extracting protons from water (absorbing the electron and leaving the proton alone) that can be used to generate the electrical flow.

The development of a complete circuit based on protons flow is still in the future but still a first step has been taken.

The University of Delaware and energy supplier NRG have formed a new national company, eV2g, to tap and distribute electricity stored in fleets of electric vehicles; the”vehicle-to-grid” hookups will be managed by a system that continuously evaluates grid needs and assembles or “aggregates” power from enough electric vehicle sources to supply those needs.

Charging the car or... recharging the grid!

Architects of the eV2g venture say the approach will give owners of electric vehicle fleets and even individual car owners a secure way to sell surplus, stored energy back into the power grid without jeopardizing the availability of cars and trucks for regular use.

Thirty percent of the cars in the United States,  would have the same capacity as all electric generation in the United States. Local pilot tests have provided as much as 15 kilowatts of electricity to the grid, on demand, from a handful of Autoport and university test vehicles. That amount is more than enough to supply 10 homes for a short period. Efficient use of stored but unused electricity in the batteries of parked vehicles can help regional electricity systems manage sudden spikes in electricity demand or abrupt falloffs in supplies from other sources.

Our Society is producing data at a staggering pace. But we are just learning how to mine these data a nd the wealth they hide.

At the Future Centre we have been looking into data for several years, convinced that this is possibly the most important opportunity for Telecom Operators and for the Society as a whole. This view is being reinforced every day more but news coming from many areas, such as the one posted by Antonello few days ago. And again, by this news published on Technology Review on Sept. 19th.

Discovering possible side effects by mining health care records

Researchers at Stanford University are taking advantage from the availability of Digital Health Care Records.

They looked at Paxil, an antidepressant being used since 1992, and Pravachol, a cholesterol lowering drug on the market since 1996. Of course both had extensive trials and passed with flying colors. However, no-one thought about possible interference and side effects when taken together.

It would be basically impossible, on the other hand, to check for potential dangerous of any two (or even more) drugs. But now, with the enforcement of digital health care records in the United States we are starting to have several years history of million of patients and it becomes possible to mine these data for answering exactly those kind of questions.

Indeed, they found that the use of the two drugs in conjunction is leading to an increased glucose level in the blood, something that is not healthy, in general, and that can be dangerous for diabetics.

It is just an example but it clearly points out at the potential of data and the opportunities that can derive from opening them up to researchers.
Clearly, along with the opportunities there are always potential dangers, in terms of privacy violation, hence the need to apply protocols and procedures to avoid them. This is an area where Telecom Operators may contribute and may generate revenues.

The original goal of AI was developing “thinking machines”, i.e. machines with human-like intelligence capabilities. Due to the challenges of this goal, the last few decades have been more focused on the so-called “narrow AI” i.e. developing methods with limited intelligence ambitions and more focused on highly constrained tasks.

Recently, however, there is return of flame back to the original ambitious goal, under the name of Artificial General Intelligence (AGI). Have look at this nice talk:

http://www.youtube.com/watch?v=x18yaOXBSQA

OpenCog is an open-source software initiative which aims at providing research scientists and software developers with a common platform to build and share artificial intelligence programs. The long-term goal of OpenCog is acceleration of the development of AGI. Framework includes components for knowledge representation (AtomSpace), task scheduling (CogServer), AI algorithm containers (MindAgents), connectors to instant messaging and virtual world systems, and other components.

http://opencog.org/

Nice bottom-up initiative, but one of the main concerns which makes me cautious in believing in “this” way toward mimicking real human intelligence is still how it will be possible capturing it and above how mimicking all humans forms of irrationality (with these, which to me, are advanced but still in-the-box approaches).

With traditional computing systems and programming methods “narrow AI” makes still more sense.  Neuroscientists say that brain is the best parallel fabric for integrating together processing, communication and storage; we know also that it does not use Boolean operations on defined stable states but it is likely that it involves complex dynamics transitions to code and to elaborate information in an highly parallel and distributed way.  So, the way to mimic brain is still unclear. And even before this, what real “intelligence” it is exactly.

I have posted a number of news on advances in 3D printing. This is just one more and it is notable for the flexibility it brings into 3D printing. Take a look at the clip:

The Media Lab is a pioneer in this area and I always enjoy seeing the amazing forms they print as demo. Clearly moving from the lab stage to the workshop or better to the home for on demand object printing is still far down the lane….

Intel has announced a microprocessor that can power a PC and be powered by a solar cell embedded in the PC.

The new Claremont chip can be powered by a tiny solar cell

The chip, code named Claremont, was presented this week by Intel at a conference in San Francisco. It has an extremely low power consumption, 10 mW in idle state. Over the next 10 years Intel expect to release in the mass market chips consuming 1/300 power of current chips having the same processing power. Note that this promise comes after the one of reducing a hundred fold the power consumption of chips Intel did, and kept, 6 years ago.

The energy needed is so low that a solar cell is sufficient to power it. We can imagine to embed the solar cell in the case of the PC so that no external power supply is needed.

During the demonstration a lamp illuminated the solar cell and the computer crunched its numbers. As they switched off the lame the computer froze. That’s why you might find yourself asking for opening the blinders in the near future.

Of course this is just the processing part of the PC. You still need to power the screen and that requires much more power. Nevertheless the effort to keep energy consumption in check goes on.

Actually in Italy the TV share is statiscally measured by a company called Auditel that has installed in more than 5 thousands families about 14 thousands auditel TV meters.

Auditel house configuration and architecture

The news that I want to highlight is that a group of researchers of Münster University in Germany has just published the result of their work on the analysis of detailed data of power meters connected to traditional TV set already present in all houses.

example of detailed consumption in an house

example of contents and channel

They have discovered that it’s possible to understand exactly which content is displayed by any commercial TV set. Any content has infact its scenes and any scene has distinctive dark areas that change dynamically the power consumption of the TV set. The analysis of detailed power consumptions permits to understand what exactly a person is watching (channel and content), when he changes channel and so on. In this case it could be possible to understand what all italians are watching just using a cheap power meter instead of the Auditel meter on a limited sample of the population.

Household appliances

The most interesting result of this research is that with the same technology applied to other household appliances like for example microwaves, washing machines, dish washers, …, it could be possible to analyze and to understand the consumers behaviour and interests for other kind of “programms” like for example the kind of cooked food, the kind of washed dresses and so on. These information could become essential for food and detergents company (some multinational companies have business in both areas) and combined with viewed TV contents can help to target the best advertising on the right channels.

I think that there will be money for all companies that will understand the business opportunity and invest on the idea to provide the technology and the service.

Projected social interaction from MIT PoCoMo

Several geeks, keen on developing interfaces for mobile devices, gathered last August at the Mobile HCI conference in Stockholm.

The MIT fluid media group brought one of the most inspiring presentations. They presented their vision in future projected social interfaces. Personal projection devices are becoming more common. Last year 500.000 picoprojectors were sold worldwide.  Their price is about 30 USD per unit. A couple of mobile phones with embedded picoprojector are out in the market: the  Samsung Beam i8520 Galaxy Beam “Halo” and the Indian manufactured Spice M9000 Popkorn.

Leveraging on these devices,  it is thus possible to imagine a scenario where mobile phones can project images on walls and track other ones’  images by the use of the phone cameras.  The phones would be aware of each others’ projections and use them for social interaction, as proposed by the MIT PoCoMo team (that’s the name of the project).

In my opinion this scenario represents an evolution of AR as conceived today, where we use mobile phones to see-through and add information about the surrounding environment. Using projectors we will be able to add more pieces of information on the real things around us, making possibly easier the process of augmenting reality.

The MIT chip for generating electricity out of vibrations

Researchers at MIT have created a chip that converts vibrations into electricity. As more and more electronics will permeate our ambient the challenge is on powering it.

There are a number of solutions to harvest vibration energy from the ambient, most common the piezoelectric effect is leveraged. For this micro quartz crystals, or material with similar properties, are glued together in a chip and a cantilever beam moves in synch with vibration squeezing the crystals which displaces electrons that are picked up by tiny electrodes.

The problem is that there is only a tiny range of vibrations that effectively produces an electrical current, hence the amount of energy transformed is limited. By introducing in the chip cantilever beams of different length, thus resonating at different frequencies it is possible to leverage a broader range of vibration frequencies but this increases the cost of the chip (in the range of 10$, far too much for powering a sensor that costs a few cents).

The researchers at the MIT have managed to obtain the same result fabricating a single layer MEMS with a cost of less than 1 $.  The MEMS looks a bit like a bridge whose pavement is made by piezoelectric material (PZT). This architecture responds to a larger range of vibration frequencies and can generate up to 45 micro watts with just a single layer of PZT which represents a hundred fold increase of efficiency.

The team is now at work to increase the effective frequency range and to increase the power output to 100 micro watts.