If you've seen a video online of robots doing surprising and sometimes a little creepy things online in the last few years, it was almost certainly a video from Boston Dynamics. In fact, in recent years, videos of their robots running, jumping, holding and manipulating objects, interacting with courageous dogs, opening doors, opening doors despite someone trying to stop them, pulling trucks, jumping have been seen on the internet. deadly and dance. Wheeled robots, bipedal robots, four-legged robots: all that move with remarkable agility, even on rough terrain and in complex environments, even when some brave Boston Dynamics employee tries to make them fall.
In short, in the very complicated field of robotic engineering, Boston Dynamics is doing very well and its videos – which are more than thirty on YouTube – have been viewed and shared millions of times. But a company that excels in the very complicated world of robotic engineering costs a lot of money and cannot be satisfied with this: at some point from viral videos you have to move to the real world. Boston Dynamics, which has existed for more than 25 years and has practically never done so until now, will have to try from here on out, and given how famous it is, what will happen will be very important for the future, at least in the short to medium term, of the robotics.
In a recent article on the future of Boston Dynamics, The Verge, wrote that the robotics sector is “notoriously ruthless” and that over the years several companies have gone bankrupt because to make robots you need a lot of money and a very high competence, but above all a lot of patience to work in a sector with very great prospects, albeit hypothetical for now. Boston Dynamics is therefore to be considered a rarity and has survived because, as Quartz summed up, “it went from making robots that looked like drunk children to making robots that looked like snappy ninjas”.
Boston Dynamics was founded in 1992 by a group of researchers from MIT, the Massachusetts Institute of Technology (United States). In 2013 it was acquired by Alphabet, the Google holding, which however in June 2017 sold it to the large Japanese group SoftBank for an undisclosed amount. Over the years, the company has not only received money from Google and SoftBank but also from the US Department of Defense (which turned down BigDog, one of Boston Dynamics' robots, because it was deemed too loud), among others.
However, these were funding for research and development of robots, not actual purchases. “We've been in the R&D phase for a long time, working hard to make robots that look like what people think a robot should be,” Raibert said. Erik Nieves, founder of automation company Plus One Robotics, said more or less the same thing, but in less beating about the bush: “They were taking money from the government and Google, they didn't have a real mission, they just had to be cool.” At the moment Boston Dynamics robots are of seven types and the main prerogative of the most interesting is mobility: something that, writes The Verge, has escaped almost all robots of the past, which were perhaps resistant, but clumsy in their movements. point to run aground at the slightest obstacle. As Hod Lipson, an engineering professor at Columbia University, said: “We think knowing how to play chess is great, but even more so is being able to coordinate hundreds of muscles.” Not surprisingly, most of the robots used today are predominantly static and mostly used in warehouses or assembly lines. Something passes in front of them and, often with a mechanical arm, they carry out a task that is usually constant and rather repetitive.
Boston Dynamics has other plans. The robot on which the company intends to focus in the near future is Spot, a four-legged robot 84 centimeters high and weighing about thirty kilos. Together with Atlas, a bipedal and humanoid robot, Spot is the most agile and surprising, the one whose videos have so far been the most successful. Spot is also the robot that Raibert has been carrying around the world for some years, often to events and technology conferences, making him jump, run and dance. Raibert told The Verge that he wanted to do with Spot, and later with Atlas and the company's other robots, what Android did with smartphones. That is, he wants Spot to become a “customizable platform” on which various customers can make the necessary software or hardware changes, adding or removing features according to their needs.
According to current plans, Spot will therefore be sold only to companies or public bodies which, perhaps by ordering more than one robot, will be able to request or make the necessary customizations, which may concern mechanical arms capable of gripping, lifting and moving different types of objects, sensors. , cameras of different types or tools to communicate with Spot.
After talking to Raibert, The Verge wrote that the Boston Dynamics commercials could be used on construction sites, warehouses and even as surveillance tools. The Verge also asked Raibert if there is anyone he would not sell the Spots to, “for example the police or the army” and he replied that it is a complicated issue but that he is “enthusiastic about any responsible use of robots” .
The sale to private individuals does not seem to interest Boston Dynamics at the moment and for now it is not known how much Spot will cost, nor is there a precise date for the start of its marketing. When he talks about it, Raibert says it will happen “shortly”. In April Raibert said that the goal is to produce at least 100 Spots by 2019; in May it said it had 50 Beta Spots ready and was aiming to produce 1,000 a year. But in the technology sector and particularly in the robotics context, these announcements must be taken with a grain of salt. As the Futurism website points out, Raibert had previously made unfulfilled announcements.
Meanwhile, some simpler, wheeled robots manufactured by other companies are already being used to survey large spaces. The pros are that robots are less expensive than a human employee and can be used for particularly tedious or dangerous tasks, or in isolated or hostile places, such as an oil rig in the middle of the sea or a nuclear power plant. The downside is that they are not always very efficient, like the time when a robot used for surveillance of a building complex in Washington DC ended up in a fountain while it was making one of its patrols.
However, Boston Dynamics is not the only company in the robotic engineering sector: among the most promising are the Chinese Unitree Robotics, the American Ghost Robotics and the Swiss ANYbotics, whose most famous product is called ANYmal. Those who deal with these issues agree that the presence of multiple companies is a good sign, because it helps to legitimize a sector about which little is known and in which companies must have the courage to invest. For now, robots exist and work when it comes to having to automate some industrial process, the challenge in the coming years will be to make them exist in other contexts that require the ability to move, interact and, in some ways, improvisation.
Among its products, Boston Dynamics has Pick, a robot that does not move and consists only of a mechanical arm, but also relies heavily on Handle, a 2-meter, 100-kilo robot mounted on wheels that can then be moved wherever it is. it seems. Handle has much more potential, but it will be more expensive and presents a series of problems: the fact that it has a battery and that it must therefore be charged and the fact that its mechanical arm, supported by wheels and therefore not well fixed to the ground, can lift and move only certain weights and certain objects.
Erik Nieves, the founder of Plus One Robotics, told The Verge about Handle: “It's like playing Tetris in 3D, in real time and half of the things that have to be moved aren't even rigid. I am convinced that Handle, as it is at the moment, will never enter a warehouse ». Raibert commented on this saying that in fact the technology behind Handle is still to be developed and that for now there are no plans for its imminent commercialization, but that its potential is still “enormous”.
More generally, there are also the criticisms of those who believe that in many cases robots are just a fad that will pass. The economists Daron Acemoglu and Pascual Restrepo, for example, spoke of the so-called “so-so automation”, with reference to the contexts in which technology and robotics do not bring an effective and concrete improvement compared to human work, but only savings for those who have to buy robots. rather than paying employees. Raibert said: “I think robots will change our lives for the better. I think they will increase productivity and I think they will free people from what is boring, dirty or dangerous. It would be terrible to see that a great technological opportunity like this is wasted due to the fear of possible negative effects “.
