After starring in countless sci-fi movies, robots trying to mimic the human form have also recently been the stars of a media tsunami. The presentation of Optimus, the bet in this field of Tesla, was responsible for it. In essence, once again, billionaire Elon Musk promised to change our society. On this occasion, through the mass production of his robot prototype and thereby achieving a price of around $20,000. The goal: a humanoid that reaches homes. The why is still not entirely clear.
Projections from different market intelligence firms predict a flourishing future for robotics. Today, industrial robots are the main business; They are followed by the military, the domestic and, light years away, are the humanoids. Robotics is more established in Asia and, to a lesser extent, in the United States, although in Europe it is experiencing strong growth. The list of companies in the sector mentioned in the reports corresponds to the geographical reading of the documents (see graph).
Market data collection firm Statzon estimates that industrial robots had a global market in 2021 valued at $32.2 billion. By 2030, the figure would rise to 86.6 billion. In the military field, defying Asimov’s first robotics law represents a business that in 2020 reached 13,000 million dollars, according to Fortune Business Insights. In 2027 it would be 25,660 million.
In domestic robots, a sector in which those dedicated to cleaning the home unquestionably reign, Markets and Markets forecasts that this 2022 will configure a market of 9,200 million dollars, numbers that would grow to around 19,300 million by 2027, according to your predictions. Humanoids, based on the numbers from the trends research firm Market Growth Reports, represent a much more modest business valued at 681 million in 2022. In 2027 they calculate that it will move about 4,200 million.
Why humanoid?
Fabiola Pérez, CEO of Mioti, believes that humanoid robots do have a future ahead of them. The expert affirms that, although great advances have already been made in robotics, it should be noted that it is one thing to produce three or four prototypes and quite another to produce them en masse. “The Musk robot itself is clumsy and reminds me of the first Boston Dynamics robot. That is to say, Tesla is far behind other companies specialized in humanoid robotics, but the most interesting thing is that promise to do it en masse and that with its financial muscle it may take them less time to close that technological gap, ”she reflects.
Asked why he makes a humanoid robot more useful than today’s vacuum cleaner, Pérez replies that human environments are designed for our physiology, and that a functional robot similar to us would consequently have an easier time interacting with our world. “Roomba is a more massive solution than a humanoid. In the home, introducing a humanoid robot would be like killing flies with cannon fire. They can be very useful in hospital environments, in those places where it is required that there is no fatigue or danger to humans. A Roomba can bring you medicine, but it can’t put down a hospital bed, it can’t help someone up. It does not have that power of force, which is another thing that robotics is contributing. With the aging population, a Roomba cannot take care of an elderly person, but a humanoid could,” she says.
Víctor Mayoral Vilches, founder of Acceleration Robotics, is more skeptical and believes that humanoid robotics is in its infancy. “Nobody knows what they contribute. Today the humanoids do nothing, they are practically not even able to sustain themselves. Anyone has seen movies with advanced humanoids and we can get an idea of their potential capabilities. Humanoids are successful as social companions. It is a reality, but they do not finish penetrating the market. Many research centers and some companies use these humanoid robots as their cover letter. It is a way of showing their status, their technological power”, he describes before pointing out that this problem of a lack of concrete purpose does not mean that research on humanoids should not be continued.
A long way
A world with these widespread devices would undeniably change. “When we talk about robotics, we always have that sword of Damocles on us when it comes to manual work. What is going to happen with the jobs that machines will start doing? It is a situation that we humans have already faced. It is an irreversible path. That is to say, we are going there,” says Pérez. “It will be a change of types of work, but we already experienced it at the beginning of the 19th century with the steam engine and at the end of the 20th century with computers and computing. We have already experienced this, the difference is that now it happens faster” , go on.
According to the expert, in addition to promoting new jobs such as device maintenance technicians, humanoids, if they are effective and functional, would have the ability to completely change the very conception of work. “I’m going to take a very extreme example. Children in the factories of the past. When they were used because their hands reached parts or gears that no one else could reach. That today seems to us to be complete barbarity. I hope that In a few years, we think it’s outrageous for a worker to give up his lower back to do his job,” Pérez wishes.
Although a mass production of efficient humanoid robots would have the potential of great economic and social consequences, everything seems to indicate that there is still a very long way to go before it becomes a reality.
Pérez is referring to Tesla’s self-driving truck, a product that has taken six years to accept orders since it was introduced. It is not yet mass produced. Mayoral, for his part, points out that the reality is that we will not see humanoids in industrial settings or in homes in the coming years, and criticizes Musk for “playing with people’s illusions.”
“Humanoids are not going to be with us next year or the next. After the talk, Musk lowered his claims. We no longer talk about one year, but between five and ten. In robotics, talking about five years is talk about a lot of time. Perhaps there is a future with humanoids in our lives, but it is a distant reality, greatly exaggerated in the media field, “he continues.
“There is an extremely long way to go, and the main problem is not the hardware, which is what Tesla has shown the most with Optimus, but the software. We are not able to endow these humanoid systems with enough intelligence to see them in our houses”, he asserts. The expert defends that in the case of the Tesla robot, by using an operating system other than ROS, the system used according to Mayoral by 95% of the companies in the sector, the Optimus software is obsolete compared to the “robotic brains”. ” that your company manufactures for other companies. “They have wanted to reinvent the wheel and have thrown away decades of learning with it,” he maintains.
Tell secrets to the robot
But aside from the difficulty of making a sufficiently intelligent humanoid, there is another big setback: privacy and cybersecurity.
Mayoral describes the robots as a Trojan horse. “We’ve been researching this for years, and they’re a perfect way to get into the very productive heart of an organization. We have proven to be able to take control of a hospital robot. From making a robotic arm beat itself to breaking point,” he reveals.
The founder of Acceleration Robotics denounces that sometimes the attacks even come from the manufacturers of the devices themselves. “We have discovered through proactive hacking that there are manufacturers that use software updates to slow them down and thus make their robot obsolete and you have to buy a new one,” Mayoral details.
For his part, Pérez warns: “The moment a machine has power, weight and strength to hurt you, it becomes a potential danger. Going back to the Roomba comparison, I can step on the vacuum cleaner, however, it is more difficult to control the humanoid than the Roomba. Automata can be controlled remotely. They can be reprogrammed, they can be hacked. They can hurt you. We have to make sure we have enough regulations to get protection. However, it is a reality to which we will adapt”, foresees the expert.
“You don’t go walking down the street and think at all times that a crane from a construction site may fall on you. It is something that can happen, and that is not why we stop manufacturing cranes. What we try to do is make them safer. As an industry , robotics will have to conform to norms and rules to make it as safe as possible,” says Pérez.
future humanoid
Xabi Uribe-Etxebarria, founder and CEO of Sherpa.ai, believes that we often associate artificial intelligence with futuristic robots or machines that speak to us. But AI goes much further.
“Currently, there are many systems that help us in our daily lives but we don’t know they work thanks to AI. To give an example, the recognition of the fingerprint or facial to unlock the mobile, the voice recognition that we use for different applications, the car park that reads the license plate without having to insert the card, any photo that we take with the mobile or applications most sophisticated in the health sector that help detect cancers that the human eye cannot detect. Systems already exist that are better than humans at specific tasks. This occurs when there is a restricted set of possibilities, for example, in games, detection and diagnosis of diseases in images, etc. Still, in most tasks, we humans are far ahead of any artificial intelligence development,” he explains.
Uribe-Etxebarria exposes a revealing example of why humans are above AI. “A human is capable of learning from very few examples. If you show a child who has never seen an elephant a picture of an elephant, and then show him another of an elephant but in a completely different posture, he will know that he is a elephant. Even more, if you draw an elephant, it will also know how to differentiate it. Currently, for a machine to be able to do that, it would have to be trained with thousands or millions of images and drawings of elephants, “he compares.
Given that it already has so many forms today, it can be inferred that humanoid robots would therefore not be the only future of artificial intelligence; however, they may be one of those futures. One that has a long way to go.
