You didn’t understand Tesla’s economic strategy. This is the message that Elon Musk conveyed during his latest quarterly update. Being naive, we thought the company’s main product was increasingly self-driving electric cars. However, the billionaire revealed that he is betting more on Optimus, Tesla’s mysterious robotics project.
First mentioned this past August, the concept seemed to be more than an embryonic idea: no prototype had been submitted (he was doing the show as a dancer disguised as a robot…) and the target audience was ambiguous to say the least. We just learned that this robot will measure about 1.70 metres, have a face screen and will be able to “do whatever humans don’t want to do” (comprehensive program). However, Elon Musk is now teasing the idea of producing the first versions of this robot from 2023. “The public hasn’t taken a measure of the scale of the program, assures the person who is also president of SpaceX and Neuralink. People will understand that the value of Optimus will eventually be more Value from the car market and more than self-driving. I’m sure of that.” So, do we really lack Vista?
Robots have already made giant strides in the past 10 years and think tank Idate estimates they will weigh in at €90 billion in 2030. Advances in artificial intelligence (AI) have given these machines a very precious sense: sight. “Robots can now decide where to put things that should, for example, be put in boxes,” explains Jean-Baptiste Moret, research director specializing in artificial intelligence and robotics at Enria.
limited offer. 2 months for 1 euro without commitment
Also gone are the days when these screwdriver buddies would only work in cages so as not to accidentally injure employees. “We now know how to better measure and dose the power of industrial robots, and make them detect contact. Thus they can often work in close proximity to humans,” the expert specifies. And in the great race for robotic advancement, the human regiment won some pretty impressive medals.
Amazingly realistic face
After announcing its canine robot rotations, Boston Dynamics wowed audiences with Atlas, its bipedal robot capable of running, jumping and doing somersaults. British Engineered Arts created the event with a robot Ameca with breathtakingly realistic facial expressions. In his presentation, the human with his faux-skin-covered face “wakes up”, looking hazy and his eyes trembling, before examining his arms and hands with a questioning look, then showing an expression of deep astonishment as he pretends to discover her body. And everything rings incredibly.
However, Elon Musk seems to be forgetting somewhat quickly the daunting challenges facing the sector in general, and Tesla in particular. “Currently, every humanoid robot on the planet has been ‘handmade’ separately,” explains Justin Carpenter, an INRIA researcher who specializes in the field. Given the technical challenges involved in their design, this may not change anytime soon. If humans now have a firm foundation, it pretty much ends there. “They still have trouble managing the extra points of contact, putting their hands on a wall, a slope, or carrying something fairly heavy,” explains Jean-Baptiste Moret. Those who dream of making them carry their shopping bags to the supermarket have to wait…
Tesla arrives very late in the race. Admittedly, Elon Musk has its expertise in computer “vision” on its side: its in-house software that enables cars is one of the most advanced. But if the eyes are the “window of the soul”, then they are only a small part of a robot. Thus, Boston Dynamics has a huge lead over Tesla in everything else, particularly in the way these mechanical objects move. And even this sector heavyweight can only “achieve his incredible feats in a very tight frame, certainly not on the first try,” said specialist Justin Carpenter.
To attract talent to Optimus, an eccentric billionaire took a checkbook. “They are offering some salaries ranging from $200,000 to $300,000,” a robotics researcher revealed. Not sure this is enough to make up for the accumulated delay quickly. The last shadow on the board that Elon draws for us (and perhaps the most disturbing): the question of how closely the human form fits into a robot. Of course, making creations in his doll is a fad that has been working with humans for ages. But once the “wow” effect of these machines wears off, what is their real added value? The question is important, because these robots that look like us cost a fortune.
‘Robots are too expensive’
The human form actually requires many more complex units than others (for example, several thousand modular components and more than 1,000 specifically designed for the Ameca robot). “The difficulty is that the human body relies on so many axes: wrists, elbows, ankles, knees… so we have to reproduce all of this using mechanical axes and robotic systems,” explains Cyril Capara, founder and CEO of Shark Robotics. , a French company that, for its part, has opted for more classic robotic figures. As a result, the bill is painful and “the cost of a robot is generally a factor of 10 higher than the cost of a conventionally shaped robot,” says an industry expert.
To get the customer to swallow this pill, it is best to be able to prove that their metal foot brings a lot of added value. And that’s where the shoe tweaks for manufacturers. Because in many situations humans are, in fact, more competitive. “No humanoid robot can rival their abilities, Will Jackson fairly admits CEO of Engineered Arts, Will Jackson. Humans self-repair and self-replicate. They maintain a good state of shape for about fifty days. A year and they are very intelligent.” Even for seemingly simple tasks, this insight makes all the difference. “Robots remain mostly machines that have been programmed once and for all, unable to adapt to an unexpected situation,” explains Jean-Baptiste Moret, Director of Research at Enria.
If they are more aesthetically pleasing, then humans also have trouble competing in many areas with basic robots. Example ? Battery life is shorter. A whole part of the bot’s activities (analysis of the environment, etc.) occurs when it is stationary. But for a bipedal machine, staying still presents a very different challenge than a robot on wheels or on tracks: Humans must activate their robotic systems and constantly make subtle corrections in order to maintain their precarious balance.
“Each hour is still a challenge for them, while manufacturers often require devices with a minimum autonomy of four hours,” says Cyril Capbara of Shark Robotics. If a jogger robot throws some of it into the training room, it won’t always be the most effective on Earth. Robots on wheels are able to pull much heavier loads. The tracked machines will play a loose ground where humans are likely to drown.
Another problem is that quadrupeds and bipeds can’t go very fast. While Boston Dynamics’ elegant Spot Dog runs at 6 kilometers per hour, Shark’s Barakuda crawler crawler travels at 20 kilometers per hour. “A humanoid robot wouldn’t be able to lift heavy loads either. That would require tremendous stability and counterweight,” finally highlights Christian Lupsi, expert at handling equipment dealer Aprolis.
For tracking analysis and decoding wherever you are
Download the app
In the end, the situations in which the human figure provides a decisive advantage are very limited. For high-risk interventions (a nuclear site after an accident, a space mission, etc.), this type of robot is of clear interest. “These devices can infiltrate complex and dangerous environments where wheeled robots do not pass and where humans can be infected,” explains Jean-Baptiste Moret from Enria. But in more complex situations, interest in humanoid robots seems negligible. So the enthusiastic promises of Elon Musk who sees Optimus helping us in the medium term in both factory and home seem implausible. But we ask only to be surprised.
The history of Christian Gulier
Transfers from France
History of Jean-Laurent Caselli