A specialist in exoskeletons, the French company has developed an autonomous humanoid as part of a partnership with its minority shareholder Renault. With big ambitions.
Wandercraft opens a new chapter in its development. Previously specializing in exoskeletons, the French company developed Calvin-40, an autonomous humanoid, as part of a partnership with Renault, which took a minority stake in its capital. 350 units of the robot are to be deployed over the next 18 months on the automaker’s production lines, with the aim of then moving to a larger-scale deployment.
The CEO of Wandercraft, Matthieu Masselin, aims to create a European, or even global, champion of the sector.
JDN. How did Wandercraft begin its transition to humanoid robotics?
Matthieu Masselin. For ten years, our only obsession was to make people walk again. This involved the deployment of stabilization and walking algorithms for our exoskeletons. We looked at what Boston Dynamics was doing and partnered with the University of Michigan and the California Institute of Technology (Caltech). We then obtained the necessary regulatory approvals in Europe and then in the United States: exoskeletons are considered medical devices and regulated in the United States by the FDA and by the European Commission in Europe.
When did you consider developing an autonomous humanoid robot?
In 2023, Wandercraft was starting to work well, and hundreds of patients were already walking thanks to the exoskeletons. That’s when, seeing that many companies like Figure were moving into humanoid robotics, I realized that these robots needed legs, while we were producing the best on the market. I went to Jerry Pratt, the CTO of Figure at the time, to offer to provide them to them. He was enthusiastic, but considered that it was a core business and wanted to favor internal development. But that’s when I realized that we already had, in a sense, a humanoid robot.
How did your alliance with Renault come about?
Renault came to see us explaining that it wanted an autonomous humanoid and asking us if we were capable of producing one. We answered in the affirmative, with ten years of experience in robotics, specifying that we would probably need a year. In the end, it only took us 40 days to present a first demonstration of Calvin, capable of moving tires.
What specific needs did you have to meet?
Initially, they showed us around their factories, already largely automated, with robotic arms, AGVs (Automated Guided Vehicle, editor’s note) and AMRs (Autonomous Mobile Robot). They told us that 60,000 people currently work in their factories, often on arduous manual tasks involving heavy lifting. Most of the time, they hire temporary workers to do it because it’s tiring, and it’s difficult to keep up over time. However, it turns out that our exoskeletons were designed to carry a person weighing up to 100 kilos. This capacity is greater than those of Unitree, Boston Dynamics or Figure robots. We then only had to add certain elements, such as manipulation or autonomous decision.
What role did Renault play in Calvin’s development?
The head is a good example. Many companies put a head on their robot. Renault didn’t interest them. They only wanted the robot to be able to see, because that’s what powers AI and its understanding of the world. We therefore placed vision sensors in different strategic locations. But the head itself has no functional benefit, while representing an additional cost. We also found that it was not necessary to have a hand with five fingers for the use cases that Renault showed us.
How is the integration of robots into factories taking place?
We provide engineers to support the deployment and train employees on security. For the rest, Calvin is autonomous and can complete a task in its entirety without the presence of a teleoperator, which constitutes an important difference compared to our competitors. And a major advantage of humanoid robots is that they can be deployed and integrated very quickly.
When do you plan to move to industrial scale production?
Concretely, Renault asked us to respond to certain use cases, with a defined level of performance. Once this objective is achieved, we will move on to a larger-scale deployment phase. We are also working with Renault to reduce Calvin’s material cost. Our agreement provides that from a certain volume, they will directly take charge of manufacturing the robots. Renault already has the necessary capabilities to industrialize a product on a large scale.
What is the threshold at which Renault will take charge of production?
This will be done on a case-by-case basis: producing 1,000 robots, each integrating 20 identical motors, implies, for example, manufacturing 20,000. On these components, Renault will be able to intervene more or less early in the industrialization process. We will therefore not need to build a “megafactory”, which will allow us to focus our investments on artificial intelligence and business development.
To what extent will Renault be able to rely on its automotive know-how to produce humanoid robots?
There are many overlaps between humanoid robotics and automobiles. For example, Renault is one of the largest manufacturers of batteries and electric motors. However, motors are a critical part of a robot because they are located at the heart of the actuators, which are crucial. Renault also produces a lot of metal parts, plastic elements, and of course the cars are full of electronics. So many components that can be transposed for the production of Calvin.
Concerning the software part, do you use partners or do you develop it internally?
We have always been full stack: embedded software, electronics, hardware. And every time we had to add a layer, we did it. Afterwards, it is true that partnerships can make it possible to move faster. There are many interesting players in the European ecosystem.
How many robots do you plan to produce in the long term?
We do not want to give figures at the moment, but according to various studies there could be a million humanoid robots in the world by 2030. We want to become the world leader in the sector, so there will be a lot of robots to produce.
How do you plan to stand out against Chinese and American competitors?
We are pushing for the reindustrialization of Europe, and robotics is one of the levers to achieve this. For the moment, the Americans have not announced large-scale deployments, despite the level of investment from which they have benefited. And regarding China, there is a kind of myth about the prices of Chinese robots. A little robot that does nothing, they will actually be able to release it at a low price. But for a real humanoid, it goes up very quickly. I don’t think we should say that we lost the game against the Chinese. We have a super favorable ecosystem in Europe. We must not let what could be the most important industry of the 21st century pass us by, giving up when we could be creating a world champion.
