Add idea


«    May 2019    »










DateDate: Today, 05:55

New Zealand engineers worked with a door mat. Now he can distinguish people by footprints, given the shape of their soles and the weight distribution of the shoe holder.
Smart door mat will remember the trail of the owner
The device creates a scan of a person's shoe when he comes to the door. The proposed development can be one of the stages of fast multifactor homeowner authentication in the future, which will eliminate the use of additional USB tokens, putting fingers on appropriate scanners, etc.
Suranga Nanayakkar and his colleagues from the University of Auckland presented a novelty at the conference AH2019. Inside the doormat there is a 15.3-inch capacitive sensor. He is able to detect an object above the carpet at a height of 4 centimeters. This property allows you to scan the shoes of the person who came to the door.
Unfortunately, the technique proposed by the craftsmen is not perfect and is not 100% accurate. As it turned out, on the second day of development testing, the recognition accuracy dropped from one hundred to ninety-six percent when analyzing the soles of 15 volunteers. But engineers are not losing optimism, they are confident that in the future they will be able to bring the work of the smart carpet to perfection.

DateDate: Yesterday, 05:58
Geksapod began its existence without having the slightest understanding of his environment. Thanks to the recursive function developed for self-improvement, the robot can gather information about its environment in order to fine-tune the behavior. The more experience a robot gains, the better it works.
At first glance, nothing complicated, but it is worth considering the fact that the robot must be able to determine not only its location and orientation in space, but also to “feel” the impulses of the sensors located on the machine’s lap.
By optimizing the behavior of the robot and concentrating on the task, scientists from Facebook managed to teach the robot to “walk” in just a couple of hours, not days.
It would seem that further on the hexapod should go and learn everything new. But it’s not so easy to evoke the passion of traveling with robots. However, this is exactly what the Facebook team intends to do in the next research phase.

DateDate: 22-05-2019, 05:57

Forecasts come true with an accuracy of 75% and 92%
The machine learning system has studied thousands of recent solar flares and determined that their occurrence could have been predicted several days before emissions from a sharp increase in the strength of electric currents in the future zone of their origin. This was told by Mark Chung from Stanford University, PNAS reports.
American scientists have suggested that all flares are generated by similar processes in the depths of the sun.
For artificial intelligence training, specialists manually analyzed more than 1000 photos of active regions on the surface of the star that caused flashes in 2010–2016 and were studied using the GOES meteorological probe tools.
The machine learning system rather quickly learned to recognize the “quiet” regions of the Sun and the future centers of flare activity with an accuracy of 75% and 92%.
At the same time, differences between them occurred and persisted for 3 days before and after the birth of the release, which indicates an extremely “long-playing” character of such activity zones.
In their predictions, AI relied on how the electrical activity of the solar photosphere changed in those zones where flashes appeared.
The subsequent analysis of data from solar satellites will help scientists understand what exactly forces solar matter to “eject” from the surface of the star and how such cataclysms can be predicted with an accuracy of 100%.

DateDate: 21-05-2019, 06:46
Traditional cement production accounts for about 8% of global carbon emissions. Since it is still impossible to do without cement, a team of geologists from the Martin Luther University decided to create an ecological alternative to it from industrial waste. This writes the Eurek Alert.
The main raw material for cement is limestone, which is heated in large kilns. At the same time, about one ton of carbon dioxide is emitted from each ton of limestone. Scientists propose to replace this material with waste from the production of kaolin and aluminum.
"This is industrial waste, which can still be used very efficiently, just for the production of alternative forms of cement," says geologist Herbert Pelmann. The study showed that "green" cement will have all the useful properties of the usual, but it will not emit greenhouse gases.

DateDate: 20-05-2019, 06:40

Researchers at the University of Cambridge have developed a new type of so-called electronic currency, the use of which will allow people to conduct financial transactions at any time, from any location and without any other restrictions. But the most important advantage of this electronic money is that the system ensuring their work is ultra-safe, it will be able to withstand and successfully resist the cyber attacks that will be carried out in the future with the help of full-fledged quantum computers, reports DailyTech. The basic principles of the organization of new electronic money, which are called “S-money”, guarantee a completely secure procedure for authentication and identification, the absence of any loss of information, a faster exchange of information than any of the existing financial communication systems. And all this is achieved by combining the power of quantum theory and the theory of relativity. Due to this, the new system will make it possible to carry out financial operations almost in real time not only from anywhere in the solar system, but also from our galaxy, although trading at the galactic level is still only science fiction. Scientists from Cambridge plan to conduct a practical test of the theoretical model developed by them a bit later this year. And, of course, this check will be conducted at the highest level available today - at the level of the entire planet. S-money technologies require very fast calculations in large quantities; nevertheless, all this is already quite realizable at the current level of information technology development. The electronic money S-money is based on virtual tokens, which are generated by the financial network, like the generation of traditional blockchain-based cryptocurrencies. At the same time, some rules for generating tokens ensure that these tokens appear in the right place and at the right time, which allows you to conduct financial transactions almost instantly without the need for lengthy verification and confirmation procedures. At the same time, some of the used laws of quantum mechanics completely eliminate the risk of double selling or using the same token. Note that this implementation of electronic money is not the first attempt to use the technology of "quantum money". Modern technologies already allow operations with quantum money, but their full use as an e-currency is not yet possible. After all, this requires a minimum of the ability to store the state of quantum particles for an infinitely long time, which is still either unattainable or will be incredibly expensive in some cases.