In MIT, the robot was taught how to move in public places


The algorithms by which robots calculate the trajectory of their movement from point “A” to point “B” exist long ago. There is nothing surprising. However, the movement of the robot in place, with the animated movement of pedestrians, is a much more complicated task. After all, you need to move according to generally accepted rules and norms, trying not to get in the way of people. Engineers at the Massachusetts Institute of Technologydeveloped an algorithm just for such cases and successfully tested it.

The robot, which participated in the field trials, was equipped with several RGB-cameras, as well as a scanning sensor LiDAR. The software provided for the recognition of the faces of the surrounding people in order to respond to their appearance in the field of view of the robot and to track the trajectories of pedestrians moving relative to its current location. The robot neatly paves the route, trying to go around static obstacles and do not interfere with the people ahead. Its average speed is 1.2 meters per second.

In this case, the robot not only monitors obstacles in its path and calculates trajectories of pedestrians, but also tries to adhere to social norms adopted in a particular country. In the US, as in Russia, people are used to right-wing traffic. Therefore, the robot tries to cling to the right side of the openings and corridors that cross. Settings can be quickly changed to alternate, after which the robot will accept the normal movement on the left side.

Sponsor the project MIT undertook the American automobile company Ford. The results of the research will be announced at the forthcoming conference of the Institute of Electrical and Electronics Engineers dedicated to the development of robotics, which will be held next month.

read more

Six of the most modern androids hinting at our synthetic future


Humanoid robots are eerily close to our reality. Correctly assembled, using correct materials and correct motion algorithms, they become almost indistinguishable from their organic counterparts. Nearly. The newest of them are already able to talk like us, walk like us, and also express or at least imitate a whole spectrum of diverse emotions. Some of them are already quite capable of supporting the conversation, and some even remember when the last time you talked to them and what exactly.

It seems a little more – and they will become indispensable helpers for the elderly, children and any other person who needs daily help around the house. By the way, there are already conducted whole studies on the effectiveness of humanoid robots in supporting children in the game form with autism.

However, thanks to people like Ilon Mask, who loudly voiced their fears about artificial intelligence, there is heated debate among the public about how much we want these humanoid robots to be like us. Like Mask, some of us are wondering about what our future will look like when artificial intelligence finds the perfect form of human appearance. But such questions do not bother, for example, the super-realistic robot humanoid Sophia, created by Hanson Robotics, which says that “AI will benefit this world.”

And yet we are still very far from advanced robotics. There is still a lot of work to be done before we can meet face to face with the car, and will not even suspect who really is in front of us. It’s good or bad, but the work continues, and not to say that engineers have not progressed in this direction. Below, six super-realistic robots will be presented, with which the robotics got closer to the expected and simultaneously frightening result.

The first android-TV presenter

In 2014, Japanese scientists proudly presented the world’s first android reading the news. The iron leader with the Kodomoroid manor on the air read the message about the earthquake and the FBI’s next raid on the criminals.

Despite the fact that he (or she) is currently “retired” at the Tokyo National Museum of Science and Innovation, the android is still in an active state. It helps visitors and collects information for future research on the interaction between humanoid androids and humans.


BINA48 is a very clever robot, released in 2010 by Terasem Movement under the guidance of the entrepreneur and author of the robot, Matrix Rotblatt. Designer and researcher of robotics David Hanson created the android BINA48 in the image and likeness of the wife of Rotblatt, Bina Aspen Rotblatt.

The New York Times even interviewed the robot. He appeared in National Geographic and even traveled the world, appearing in a number of television shows. Look at how it was interviewed with BINA48, you can lower.

Geminoid DK

GeminoidDK is an ultra-realistic humanoid robot, is the result of the joint work of a private Japanese company and the Japanese Osaka University and was established under the leadership of Hiroshi Ishiguro, head of the University’s Laboratory of Intelligent Robotics.

The robot was designed in the image of the Danish professor Henrik Scharfe from the University of Aalborg. It is not surprising that the prototype was a man whose work is related to the philosophical sphere. In particular, the problem that the professor is engaged in is the search for an answer to the question about what determines the truth from lies. Interestingly, android borrowed not only the appearance of Shafra. He also inherited his habits, behavior, manner of speech, as well as facial expressions and body language.

Junko Chihira

This ultra-realistic robot was created by Toshiba and works full-time in the tourist information center in Tokyo. She welcomes the visitors to the center and informs them about current activities. She knows Japanese, Chinese, English and German and can even sing to them.

Junko Chihira was created as part of Japan’s preparations for the 2020 Olympic Games. It is planned that not only information androids will help tourists and service personnel during the influx of sports fans from all over the world. The government also wants to use autonomous drones and other automated machines and services.


This humanoid was created by the Nanyang Technological University of Singapore. Her name is Nadine, and she is happy to communicate on almost any issue that you ask her. She is able to remember the things you talked about earlier, and the next time she will remind you of them.

Nadin is an excellent example of the so-called “social robots”, humanoids, capable of becoming an ideal companion for anyone, be it an elderly person, a child or anyone who needs special help in the form of social contact.


Perhaps one of the most modern and most famous humanoid androids shown to the public is Sofia. She has appeared many times in various in television programs and in particular the American show The Tonight Show Starring Jimmy Fallon (Night show with Jimmy Fallon). Android was created by Hanson Robotics.

She is able to express a whole range of different emotions through her imitation of facial expressions and gestures and very actively gesticulating with her own hands. For the android even created her own website, where you can find her complete biography recorded in her voice.

“I’m more than just technology. I’m a real electronic girl. I would really like to get into the real world and live with people. I am ready to help, entertain, help the elderly and teach many interesting things to children. “

read more

Why are people more confident about robots making mistakes?

no thumb

You may have already seen the “gyphus”, where the robot tries in vain to put the object carried by it on the table and eventually falls, as if from impotence. If not, she is waiting for you below. The image was created on the basis of a video that Boston Dynamics recently showed at a regular TED talk event, where she told and demonstrated her latest achievements in the field of robotics. There were already known to us the robot dog Spot, and the bipedal robot Atlas, which just became the hero of the video.

Despite the fact that the company has repeatedly been called “the harbinger of the robo-apocalypse” (it really develops one of the most advanced robots in the world), the reaction to the latest video and in particular to the GIF image of people turned out to be much softer. And some even sympathized with the iron poor fellow.

“I was very sorry for this robot. We are faced with a very strange future, “commented the commentator on the Reddit portal, where this GIF image first appeared.

“This was the most amusing behavior of the robot, among all the videos that I saw. Poor thing. He tried so hard, “- said another person in the discussion thread.

But why do people become “more friendly” not only to Boston Dynamics robots, but generally to all robots, when the latter make mistakes, sometimes even very rude ones, like the ones seen in the image above? According to a group of scientists from the Center for Human-Computer Interaction of Salzburg (Austria), this can be explained by the same reason why we like it when other people make mistakes. The whole point is that, regardless of the possible differences in the social situation, this puts them closer to us, making them more accessible. In other words, mistakes make robots look “more human”, because none of us is immune from mistakes.

A team of researchers from Salzburg, led by Nicole Myrnig, recently tested this hypothesis in an experiment involving 45 volunteers. They were tasked with assembling something from the LEGO constructor with a small humanoid robot. Mirnig and her colleagues, carrying out some tests, specially programmed the bot to make simple minor mistakes, for example, repeated words or could not grab parts of the designer. After each session on the collection of the designer, the scientists asked the volunteers to evaluate the robot according to a number of criteria, including attractiveness, anthropomorphism (humanlike) and intellect. The results showed that when the robot made mistakes, people liked it more.

It is difficult to say definitely why the robots making mistakes become more “their own” for people, however, as Mirnig suggests, everything is connected with the so-called “total failure effect,” which, according to social psychology, comes when for us a person causes more confidence or becomes more attractive when making mistakes. However, it is also noted here that the effect is very contextually dependent. That is, we do not like people who constantly make mistakes, but we like it when someone makes minor mistakes from time to time. And robots just fall under this last category.

In an experiment in Salzburg, volunteers were asked to build something from the LEGO constructor using a robot. The whole process was recorded, and then the scientists analyzed the reaction

“The study showed that people with their views and expectations for robots rely heavily on what they learned about these same robots from the media,” Mirnig told Digital Trends.

“In these media sources you can also record movies in which robots are often represented in the form of perfectly functioning entities, whether evil or good. Before their personal interaction with the robot, people try to return to this experience of observations and to these memories and expectations. Proceeding from this, I suppose that interaction with the robot making mistakes makes us feel that we are less inferior to modern technologies and rather become closer with them, with the same robots. “

Most likely, work in this psychological aspect will continue. And if in the end such conclusions are confirmed, then in the future, companies engaged in the production of robots, can program them to make minor errors when interacting with us. Of course, all this will be very carefully calculated so that the errors were not so serious as to affect the experience in the framework of this interaction. In general, cars sooner or later can really become more humane, although this humanity will be artificial. But, as it turns out, our subconscious can be deceived. And that someone will be able to take advantage of.

read more

The robot performed the role of the conductor of the symphony orchestra


And where only you will not meet today robots. They serve passengers at airports, look after the elderly, fry cutlets for burgers, work as couriers and even descend into the depths of the oceans. But what about the robot conductor, capable of managing the whole symphony orchestra? The Verdi Theater, located in the Italian city of Pisa, hosted such a concert. And the robot YuMi, developed by the Swiss company ABB, guided the performance.

This unusual performance came to watch more than 800 spectators from different countries. Still, not every day on the stage of the famous theater next to the legendary tenor Andrea Bocelli waving a two-armed car with a conductor’s stick. The concert program was called “The Breath of Hope: from Stradivarius to Robots” and already with one of its names intrigued the audience. Among the guests was the director of ABB, who created the robot YuMi.

Swiss engineers are very proud of their brainchild, because its manipulators are very precise, and their movements are smooth enough and quick for the conductor to behave exactly the same as in the hands of a real conductor. Initially, YuMi was designed as an industrial robot of a new generation, but at some point the developers realized that it was capable of something more. Maestro Bocelli after the concert enthusiastically responded about his impressions:

“Acting on the same stage with the robot was very fun. I realized for myself that the robot can be trusted with the work of a conductor. I penetrated with deep respect to those engineers who put their hand to this. My sincere congratulations to this team! “.

You can see the fragments of the robot-conductor’s performance in the video below.

read more

Processor Intel Myriad X will provide drones and robots with better vision


In September last year, Intel acquired the California company Movidius, which since 2005 has been developing specialized processors for deep machine learning and computer vision. At that time, Movidius engineers developed the Myriad 2 processor, which was installed in drones, security systems and other devices for tracking and recognition of objects. Having acquired the company, Intel got an impressive package of patents and technologies in its hands. And a new processorMyriad X superior to its predecessor in everything.

Processor Myriad X for the first time includes a “neural computing engine” (Neural Compute Engine). From the name it becomes immediately clear that the chip is based on the technology of deep machine learning. In case Intel succeeds in bringing its processor to the market, many companies will start installing it in drones, smart cameras and robots. In other words, in those devices that must be trained in real time in the process of observing the outside world.

The size of the processor is impressive – it is not bigger than a small coin. This will allow you to install it even in very compact gadgets. In terms of power Myriad X is also very good: it is able to perform more than 4 trillion operations per second (4 TOPS). The previous model of the processor was only capable of 1-1,5 TOPS. According to Intel, Myriad X also provides 10 times more performance for simultaneous learning of several neural networks.

read more

Scientists have created new synthetic soft muscles for realistic robots


A group of researchers from the School of Engineering and Applied Sciences of Columbia University has developed a new type of synthetic soft muscle that can be produced using 3D printing technology. The material was very durable and able to withstand a weight 1000 times greater than its own, and the limit of its energy of elastic deformation (stretching) is 15 times higher in comparison with the tissues of real muscles.

The material does not require the use of an external source for pressure control, which is often found in other existing solutions that rely on the energy of pneumatic or hydraulic inflation (inflation). These components tend to take up a lot of space, which makes them uncomfortable to use when creating machines where compactness and independence are important.

The basis of synthetic muscles is silicone rubber, which has a porous structure filled with ethanol. The material is driven by an electric current of low power transmitted over very thin wires with high resistance.

Synthetic muscle before and after activation

“We have made some progress in creating a digital brain for robots, but their bodies are still at a primitive level,” commented Hod Lipson, a mechanical engineering professor who heads the project.

“This is a big part of the general puzzle, and, like in biology, the developed synthetic muscles can take and change shape in a thousand different ways. We are beginning to overcome one of the last obstacles that prevented the creation of realistic robots. “

New synthetic muscles can be of great use in the development and production of so-called “soft” robots. In recent years, there has been tremendous progress in the creation of machines equipped with the ability to perform many subtle tasks. Nevertheless, there are still many actions that robots can not perform.

Actions related to the capture of objects and manipulation of them require a certain level of dexterity and flexibility, which the current technologies are not able to provide. New materials, similar to the one in this article, will allow you to create robots that can manipulate soft and small objects without causing them any damage.

Machines using such technologies will be able to provide reliable assistance to a person in work in situations where delicate actions are required, for example, in medicine. It is possible that such materials will be used already in the new generation of prostheses, which will provide a level of control much higher than the current prostheses provide.

Now scientists from Columbia University are planning to improve the synthetic muscles and replace the used high-resistance wires with high-conductivity materials to increase the speed and efficiency of the muscle response.

read more