Latest
news...

+ products
Ineltec, blog de noticias - Ineltec

Prof. Dion Vlachos (left) and Joshua Lansford. (Photo: Jeffrey C. Chase)

 

They call it artificial intelligence, not because the intelligence is somehow bogus. It is real intelligence, but it is still made by humans. That means AI - a powerful tool that can add speed, efficiency, insight and precision to a researcher's work - has many limitations.

 

It is only as good as the methods and data given to it. By itself, it does not know if information is missing, how much weight it has to give to different types of information or if the data from which it draws is incorrect or corrupted. You cannot accurately deal with uncertainty or random events, unless you learn how to do it. Relying exclusively on data, as machine learning models often do, does not take advantage of the knowledge that experts have accumulated over the years and the physical models that underpin physical and chemical phenomena. It has been difficult to teach a computer to organize and integrate information from very different sources.

 

Now, researchers from the University of Delaware and the University of Massachusetts-Amherst have published details of a new approach to artificial intelligence that incorporates uncertainty, error, physical laws, expert knowledge and data into its calculations. are missing and that ultimately leads to much more reliable models. The new method provides guarantees that are typically lacking in artificial intelligence models, showing how valuable - or not - the model can be in achieving the desired result.

 

Joshua Lansford, a doctoral student in the Department of Chemical and Biomolecular Engineering at UD, and Prof. Dion Vlachos, director of the Center for Catalysis for Energy Innovation at UD, are co-authors of the article published in the journal Science Advances. Also contributing were Jinchao Feng and Markos Katsoulakis from the Department of Mathematics and Statistics at the University of Massachusetts-Amherst.

 

The new mathematical framework could lead to greater efficiency, precision, and innovation for computer models used in many fields of research. Such models provide powerful ways to analyze data, study complex materials and interactions, and fit variables virtually rather than in the laboratory.

 

"Traditionally, in physical models, we build a model first using just our physical intuition and expert knowledge about the system," Lansford said. "After that, we measure the uncertainty in the predictions due to error in the underlying variables, often relying on brute force methods, where we take samples, then run the model and see what happens."

 

Efficient and accurate models save time and resources and guide researchers toward more efficient methods, new materials, higher precision, and innovative approaches that they might not otherwise consider.

 

The article describes how the new mathematical framework works in a chemical reaction known as the oxygen reduction reaction, but it is applicable to many types of modeling, Lansford said.

 

"The chemicals and materials that we need to make things faster or even possible - like fuel cells - are highly complex," he said. "We need precision. And if you want to make a more active catalyst, you need to have limits on your prediction error. By intelligently deciding where to put your efforts, you can define the area to be explored. Uncertainty is taken into account in the design of our model", Lansford said. "Now it is no longer a deterministic model. It is a probabilistic one."

 

With these new mathematical developments in place, the model itself identifies what data is needed to reduce model error, he said. Then a higher level of theory can be used to produce more accurate data or more data can be generated, leading to even smaller limits of error in predictions and reducing the area to be explored.

 

"Those calculations take a long time to generate, so we're often dealing with small data sets. That's where the need to spread the error comes in."

 

This is not a guarantee that the use of a specific substance or method will deliver precisely the desired product. But it is much closer to a guarantee than you could get before.

 

This new method of modeling could vastly improve work in renewable energy, battery technology, climate change mitigation, drug discovery, astronomy, economics, physics, chemistry, and biology, to name just a few examples.

 

Artificial intelligence does not mean that human experience is no longer necessary. Quite the opposite.

 

The expert knowledge that emerges from the laboratory and the rigors of scientific investigation are essential, fundamental material for any computational model. (Source: NCYT Amazings)

 

 

Source: Redacción, N., 2020. Un Gran Salto Adelante Para La Inteligencia Artificial. [online] Noticias de la Ciencia y la Tecnología (Amazings® / NCYT®). Available at: <https://noticiasdelaciencia.com/art/39926/un-gran-salto-adelante-para-la-inteligencia-artificial> [Accessed 27 October 2020].


At Ineltec we offer tailored made solutions to perform any kind of test, doing a strong analysis of all the details so we can provide a tailored answer, that is also the most efficient and affordable solution to our client. If you need more information, don't hesitate to contact us by sending your request to This email address is being protected from spambots. You need JavaScript enabled to view it..

 

If you have suggestions on our blog or need information from our teams, do not hesitate to contact us.

Commercial, Marketing and Communication Department

Email:  This email address is being protected from spambots. You need JavaScript enabled to view it.

Tel: (+34) 938.605.100

 

Development of an organic battery

Tuesday, 20 October 2020 10:21

Mikhail Vagin and PhD student Penghui Ding. (Photo: Thor Balkhed)

 

Researchers at Linköping University's Organic Electronics Laboratory have demonstrated an organic battery for the first time. It is of the type known as "redox flow battery", with a large capacity that can be used to store energy from wind turbines and solar panels, and as a power bank for automobiles.

 

Redox flow batteries are stationary batteries in which the energy is found in the electrolyte, outside the cell itself, as in a fuel cell. They are often marketed with the prefix "eco", as they open up the possibility of storing excess energy from, for example, the sun and the wind. Also, it seems to be possible to recharge them an unlimited number of times. However, redox flow batteries often contain vanadium, a rare and expensive metal. The electrolyte in which the energy is stored in a redox flow battery can be water-based, which makes the battery safe to use, but results in a lower energy density.

 

Mikhail Vagin, Principal Research Engineer, and his colleagues from the Laboratory for Organic Electronics, Norrköping Campus, have succeeded in producing not only a water-based electrolyte but also electrodes made of organic material, which greatly increases the energy density. In this way it is possible to manufacture completely organic redox flow batteries for storing, for example, energy from the sun and wind, and to compensate for the load variation in the electricity supply network.

 

They have used the conductive polymer PEDOT for the electrodes, which they have doped to transport positive ions (cations) or negative ions (anions). The water-based electrolyte they have developed consists of a solution of quinone molecules, which can be extracted from forest materials.

 

 "Quinones can be derived from wood, but here we have used the same molecule, along with different variants of the conductive polymer PEDOT. It turns out that they are highly compatible with each other, which is like a gift from the natural world," says Viktor Gueskine, engineer Principal Investigator of the Organic Electronics Laboratory, and one of the authors of the paper now published in the journal Advanced Functional Materials.

 

High compatibility means that PEDOT electrodes help quinone molecules switch between their oxidized and reduced states, thereby creating a flow of protons and electrons.

 

"Normally it is difficult to control the ion process, but we have succeeded here. We also use a fundamental phenomenon within electrocatalysis in which a special ion in solution, in this case quinone ions, is converted into electricity. The phenomenon is conceptualized by us as ion-selective electrocatalysis, and probably exists in other types of membrane storage devices such as batteries, fuel cells and supercapacitors. This effect has never been discussed before. We showed it for the first time in redox flow batteries ", Mikhail Vagin says.

 

Organic redox flow batteries still have a lower energy density than batteries containing vanadium, but they are extremely cheap, completely recyclable, safe and perfect for storing energy and compensating for variations in load in the electrical supply network. Maybe in the future we will have an organic redox flow battery at home, like a power bank for the electric car. (Source: NCYT Amazings)

 

 

Source: Redacción, N., 2020. Desarrollo De Una Batería Orgánica. [online] Noticias de la Ciencia y la Tecnología (Amazings® / NCYT®). Available at: <https://noticiasdelaciencia.com/art/39861/desarrollo-de-una-bateria-organica> [Accessed 20 October 2020].


At Ineltec we offer tailored made solutions to perform any kind of test, doing a strong analysis of all the details so we can provide a tailored answer, that is also the most efficient and affordable solution to our client. If you need more information, don't hesitate to contact us by sending your request to This email address is being protected from spambots. You need JavaScript enabled to view it..

 

If you have suggestions on our blog or need information from our teams, do not hesitate to contact us.

Commercial, Marketing and Communication Department

Email:  This email address is being protected from spambots. You need JavaScript enabled to view it.

Tel: (+34) 938.605.100

 

Solar panel “checkerboard” design. (Photo: Dr Davide Zecca)

 

Scientists continue to tirelessly research to increase the performance of solar panels, so that it is more feasible for society to adopt and increase the use of fully renewable energy.

 

According to a new study, a new solar panel design, which uses a structure in the form of "checkerboard lines" is capable of increasing its ability to absorb light by 125%.

 

The researchers say this breakthrough could lead to the production of thinner, lighter and more flexible solar panels that could be used to power more homes and used in a wider range of products.

 

The study - led by researchers at the University of York and carried out in partnership with the NOVA University of Lisbon (CENIMAT-i3N) - investigated how different surface designs influenced the absorption of sunlight in solar cells, which are join together to form solar panels.

 

The scientists found that the chessboard design improved diffraction, increasing the likelihood that light would be absorbed and then used to create electricity.

 

The renewable energy industry is constantly looking for new ways to increase the light absorption of solar cells in lightweight materials that can be used in products ranging from common solar panels to roof tiles to boat sails and camping equipment.

 

Silicon for solar use - used to create solar cells - requires a lot of energy for its production, so creating thinner cells and changing the design of their surface would make them cheaper and more environmentally friendly, allowing the creation of solar panels with higher performance.

 

Dr. Christian Schuster from the Department of Physics said: "We found a simple trick to increase the absorption of thin solar cells. Our research shows that our idea actually competes with the improved absorption of more sophisticated designs, also absorbing more light in the deep zone of its plane and less light near the surface structure itself. Our design rule complies with all relevant aspects of light collection for solar cells, clearing the way for simple, practical and outstanding diffractive structures with a potential impact beyond photonic applications and solar panels. This design offers the potential to further integrate solar cells into thinner, more flexible materials and therefore create more opportunities to use solar energy in more products".

 

 The study suggests that the design principle could impact not only the solar panel or LED industry, but also applications such as acoustic noise shields, windbreaks, anti-slip surfaces, biosensor applications, and refrigeration. atomic.

 

Dr Schuster added: "In principle, we would deploy ten times more solar energy with the same amount of absorbent material: ten times thinner solar cells could allow rapid expansion of photovoltaic energy, increase solar electricity production with solar panels and greatly reduce our carbon footprint. In fact, as refining silicon feedstock is such an energy-intensive process, silicon cells ten times thinner would not only reduce this need, but also cost less, which would enhance our transition to a greener economy."

 

Data from the Department of Business, Energy and Industrial Strategy show that renewable energy - including solar energy - already accounted for 47% of the UK's electricity generation in the first three months of 2020. (Source: NCYT Amazings)

 

 

Source: Redacción, N., 2020. Un Nuevo Diseño De Paneles Solares Multiplicará Su Capacidad De Absorber Luz. [online] Noticias de la Ciencia y la Tecnología (Amazings® / NCYT®). Available at: <https://noticiasdelaciencia.com/art/39787/un-nuevo-diseno-de-paneles-solares-multiplicara-su-capacidad-de-absorber-luz> [Accessed 14 October 2020].


At Ineltec we offer tailored made solutions to perform any kind of test, doing a strong analysis of all the details so we can provide a tailored answer, that is also the most efficient and affordable solution to our client. If you need more information, don't hesitate to contact us by sending your request to This email address is being protected from spambots. You need JavaScript enabled to view it..

 

If you have suggestions on our blog or need information from our teams, do not hesitate to contact us.

Commercial, Marketing and Communication Department

Email:  This email address is being protected from spambots. You need JavaScript enabled to view it.

Tel: (+34) 938.605.100

 

In 2016, a supercomputer beat the world champion at Go, a complicated board game. How? Using reinforcement learning, a type of artificial intelligence in which computers train themselves after being programmed with simple instructions. Computers learn from their mistakes and, step by step, they become very powerful.

 

The main drawback of reinforcement learning is that it cannot be used in some real-life applications. That's because in the training process, computers initially try almost everything before stumbling on the right path. This initial trial and error phase can be problematic for certain applications, such as climate control systems, where sudden changes in temperature would not be tolerated.

 

CSEM engineers have developed a method that overcomes this problem. They showed that computers can first be trained in extremely simplified theoretical models before being prepared to learn about real-life systems. This means that when computers begin the machine learning process in real-life systems, they can use what they previously learned in models. Therefore, computers can quickly get on the right track without going through a period of extreme fluctuations. The engineers' research has just been published in the journal IEEE Transactions on Neural Networks and Learning Systems.

 

Pros y contras de la Inteligencia Artificial - Cepymenews

 

"It's like learning the driver's manual before starting a car," says Pierre-Jean Alet, head of smart energy systems research at CSEM and a co-author of the study. "With this pre-training step, computers build a knowledge base that they can draw on, so they don't go blind while looking for the right answer."

 

The engineers tested their approach to a heating, ventilation and air conditioning (HVAC) system for a complex 100-room building using a three-step process. First, they trained a computer in a "virtual model" built from simple equations that roughly described the behaviour of the building. They then entered real building data into the computer (temperature, shutter opening time, weather conditions, etc.) to make the training more accurate. Finally, they let the computer run its reinforcement learning algorithms to figure out the best way to handle the HVAC system.

 

This discovery could open up new horizons for machine learning by expanding its use to applications where large fluctuations in operating parameters would have significant financial or security costs.

 

 

Source: Redacción, N., 2020. Pre-Entrenan A Las Computadoras De Inteligencia Artificial Para Hacerlas Aún Más Potentes. [online] Noticias de la Ciencia y la Tecnología (Amazings® / NCYT®). Available at: <https://noticiasdelaciencia.com/art/39701/pre-entrenan-a-las-computadoras-de-inteligencia-artificial-para-hacerlas-aun-mas-potentes> [Accessed 7 October 2020].


At Ineltec we offer tailored made solutions to perform any kind of test, doing a strong analysis of all the details so we can provide a tailored answer, that is also the most efficient and affordable solution to our client. If you need more information, don't hesitate to contact us by sending your request to This email address is being protected from spambots. You need JavaScript enabled to view it..

 

If you have suggestions on our blog or need information from our teams, do not hesitate to contact us.

Commercial, Marketing and Communication Department

Email:  This email address is being protected from spambots. You need JavaScript enabled to view it.

Tel: (+34) 938.605.100