- il y a 1 jour
Beyond Human Limits: Robotics to Walk and Empower
Catégorie
🤖
TechnologieTranscription
00:00Bonjour tout le monde, je suis super excitée de vous aujourd'hui avec vous et avec Jean-Louis
00:05pour partager un peu plus sur la histoire de Wondercraft.
00:10Juste un petit mot sur moi, je suis le président de l'ealthech investment à Eurasio,
00:16une firme de l'investissement basée en France.
00:19Nous investissons en tech start-ups de C à Série B, de toute l'Europe.
00:24Et bien sûr, Wondercraft est une de nos compagnies de portfolio.
00:29Donc je suis super excitée de prendre un peu de temps avec vous aujourd'hui
00:31pour partager un peu plus sur la histoire de cette entreprise.
00:36Jean-Louis, peut-être que vous pouvez vous présenter?
00:38Merci Maxime, je suis Jean-Louis, je suis un des trois fondateurs de cette entreprise
00:42que nous allons vous présenter à Wondercraft.
00:48Et c'est ça.
00:48Je suis très heureux d'investir à Eurasio depuis longtemps, sept ans.
00:55En 2018, sept ans, oui.
01:01Jean-Louis, vous avez différents positions dans différents telco et des entreprises
01:07dans le passé.
01:09Pouvez-vous nous expliquer ce qui a fait vous rejoindre Mathieu et Nicolas
01:14dans cette nouvelle aventure de Wondercraft et quelle est l'ambition de l'entreprise
01:19au début ?
01:20Oui, c'est parce que mon fils, Oskar.
01:24Il est le gars sur le...
01:28...on le droite.
01:30On le droite.
01:31Et quand il était 5 ans, il m'a dit,
01:34« Dad, vous êtes un ingénieur robotique.
01:37Donc, au lieu d'avoir un ingénieur robotique, pourquoi ne pas vous faire un robot pour m'aider ? »
01:42Et puis j'ai dit, « Wow,
01:45marcher dans le robotique est le plus difficile.
01:49Et c'était, à l'époque, le grail de la robotique.
01:53Et, à l'époque, je n'avais pas fait beaucoup de robotique dans les dernières 30 ans.
01:57À l'époque, j'ai rencontré un autre gars, Nicolas,
02:03qui était frais de l'ingénieur de l'école en France,
02:07avec son meilleur ami Mathieu, de la même school.
02:10Et Nicolas'un autre brother,
02:12l'autre personne,
02:13avait la même pathologie que mon fils.
02:16Mais il savait comment transformer la physique et la maths en un robot.
02:22Et, à l'époque, en 2012,
02:25nous n'avons même pas les vidéos de Boston Dynamics.
02:29Donc, un robot walking n'était rien que un rêve.
02:34Et il avait toutes les intuitions et les visiones technologiques
02:38qui nous rendent où nous sommes aujourd'hui,
02:41que je vais vous montrer.
02:44Et que, en fait, ces deux personnes walkent aujourd'hui.
02:49Donc, c'était, en fait,
02:51qui m'ont répondu à leur question,
02:52qui m'a permis de lancer Warnercraft.
02:57Super, super intéressant.
02:59Et, Jean-Louis,
03:00peut-être expliquer à nous pourquoi,
03:01aujourd'hui,
03:02en 2025,
03:04nous pouvons toujours voir des gens
03:05dans des wheelchairs,
03:06dans la rue,
03:07et nous pouvons voir des gens
03:08guidés par exosquelettons.
03:12Qu'est-ce qui fait que c'est difficile
03:15de mettre exosquelettons dans la rue
03:17et de remplacer les wheelchairs ?
03:19Tout d'abord,
03:20les gens dans les wheelchairs,
03:21ils ne sont pas malheureux,
03:24en général.
03:25Et les wheelchairs,
03:26c'est la chose qui nous a aidé,
03:28qui nous a aidé,
03:29à récupérer la liberté
03:30et à avoir une vie.
03:31Et ils ont une vie,
03:33comme j'ai une vie,
03:34vous avez une vie.
03:34Donc,
03:35nous offrions quelque chose de différent
03:37en termes d'options dans cette vie.
03:40Probablement plus de mobilité.
03:42Et une très importante chose,
03:43c'est que,
03:44quand vous êtes en train d'un wheelchair
03:46tous les jours,
03:47votre health degrés,
03:49décès,
03:50rapidement.
03:51Et votre lifespan
03:53peut être augmenté
03:55par 10%, 20%, 30%,
03:5740%,
03:58depending sur votre pathologie.
04:00C'est juste sitting tous les jours.
04:02Donc,
04:03standing up et walking
04:04est une grande avantage.
04:06Et pourquoi il n'y a pas d'exoskeletons ?
04:08Actually,
04:09Wondercraft est l'unique
04:10qui fait un exoskeleton
04:13que nous pouvons montrer dans la vidéo,
04:16qui peut marcher naturellement.
04:19Donc, nous sommes toujours les uns.
04:20C'est extrêmement complexe
04:22pour développer.
04:23Il nous a fait des années.
04:24Et puis,
04:26quand il est développé,
04:28vous aurez de,
04:30un,
04:30parce que ce n'est pas comme
04:33que vous voulez
04:33pour une vidéo.
04:36Oui,
04:36on a usually une vidéo.
04:38Oui,
04:38OK,
04:38perfect.
04:39Vous vous trouverez
04:41quelqu'un
04:41inside
04:42qui est fragile.
04:43Donc,
04:44il n'y a pas de façon
04:44que vous pouvez faller.
04:46Vous voulez
04:46éviter de faller
04:47à tout prix.
04:48Donc,
04:49c'est Thibaut,
04:49le même
04:51jeune homme
04:52qui était dans la photo.
04:54Et Thibaut
04:55qui s'est walking
04:55par la scène,
04:56on a fait
04:56la vidéo
04:56pour le NVIDIA
05:01keynote
05:01à la CES
05:02en Las Vegas
05:04ce soir.
05:07Donc,
05:08vous voyez,
05:10arrivé à ce point
05:11of technology
05:12took us
05:12a lot
05:13of time.
05:13Il y a
05:14only
05:14us
05:14today
05:15who
05:15can
05:15do
05:15that.
05:16And then,
05:17you have to
05:18make it approved
05:19by the FDA.
05:20Because that's going
05:21to be reimbursed
05:22in the USA,
05:23which is great.
05:24But before,
05:25you have to go.
05:26And we are currently
05:27in the last clinical trials
05:28for filing to the FDA
05:29and hopefully having
05:31an approval
05:33at the year end
05:33or early 26
05:35and being on the market
05:36at that time.
05:36but it took us
05:37like 12 years.
05:39Definitely,
05:40it takes always
05:41a lot of time
05:41in healthcare
05:42but even more
05:43when it's about
05:44having a robot
05:45in a real condition.
05:47So,
05:48when we invested
05:49in 2018
05:50in Wondercraft,
05:51it was just a team
05:52of brilliant engineers
05:53with an amazing
05:55ambition and projects.
05:57Now, seven years after,
05:59can you maybe explain
06:00us what makes
06:01Wondercraft so special?
06:03What are the secrets
06:05behind
06:06behind the technology,
06:07the solution
06:07that makes it possible
06:10to have a robot
06:11in the street
06:12to guide patients?
06:14Yeah, well,
06:15we have a tech
06:16and the whole team
06:17that's quite,
06:19every startup says
06:20we have an unbelievable team.
06:22We do have a team
06:23that's very strong
06:24in technology
06:25and in science
06:26and that's not so frequent.
06:28Some investors
06:29don't like science.
06:31That's true.
06:32You did accept it.
06:33So, thank you.
06:35We have a small number
06:37of people
06:38whose vision
06:39we're rolling out,
06:41including Nicola,
06:43brother of that person
06:45who's walking,
06:46who are outstanding,
06:48meaning that you don't
06:49find 10 of them
06:51on the scene
06:51of robotics and AI.
06:54we were able to recruit
06:56the best
06:57because we're doing
06:58something that's cool.
07:00Actually,
07:00we're doing something
07:01that doesn't need to explain
07:03why it's impactful
07:05or why it's good
07:06because it is good
07:08and that's important.
07:11we made good technological choices.
07:14So, you can say
07:16we were lucky
07:17or the people
07:17who made the choices
07:18were smart.
07:20It's probably
07:22and we're in a good country.
07:24France is good
07:24because we have
07:25very good AI engineers,
07:26very good roboticists,
07:28a good science background
07:29and we had good investors
07:31because you have to be
07:32extremely patient.
07:35When we pitch to investors,
07:36we always say,
07:37well, it's going to take some time.
07:39It's going to take some time
07:40because we're from science
07:41to technology
07:42to FDA to market
07:44and they say yes
07:45and they may not think yes.
07:47Actually,
07:48our investors
07:49were okay to
07:51walk with us
07:52in this long journey
07:53to the market
07:54and now with new developments
07:56that are quite cool.
07:58And maybe,
07:58so you mentioned
08:00you mentioned rapidly AI.
08:02I think a few years ago,
08:04we didn't have the same level
08:05of technology
08:06than today.
08:09What is the main
08:13change today for you
08:14with new AI technologies?
08:16How do you include that,
08:18embed that in the robots
08:19and will it make a difference
08:21or not for your technology?
08:24AI in robots plays
08:25at least at two levels
08:27and then on others.
08:29There is an AI
08:30that activates the robot,
08:33makes the actions,
08:34pilots the motors,
08:35makes it walk
08:37or move the arm.
08:38And there is another level of AI
08:40which is more mission oriented.
08:42what does the user want to do?
08:45If it's in a factory,
08:47it has to understand
08:48this is a tire,
08:49I have to take it,
08:50carry it somewhere,
08:51put it on the car,
08:53check the quality
08:53and that's the mission.
08:54and that is more
08:56a generative AI like Mistral
09:00and the robotics AI is more like
09:03custom and no one today,
09:06to my lineage,
09:06makes a unified VLA
09:08which is completely unified AI.
09:11we have some things
09:12that's more context related
09:13and some things
09:14that's more action related.
09:16Then you have also neural networks
09:18to analyze what the robot sees
09:21or perceives via sensors, etc.
09:24So AI has changed a lot.
09:27The work,
09:27the quality of the work
09:28you have on the exoskeleton
09:30that's on the scene here,
09:32that's AI,
09:33that's robotics AI.
09:35by comparison with what we call model-based,
09:39model-based algorithms,
09:41they are basically solving
09:45non-linear differential equations
09:47in real time
09:48that are very complex
09:49and based on extremely complex
09:50mathematical models.
09:52And that was good for work
09:54until I would say now,
09:58but the border,
09:59there were boundaries.
10:00It was very difficult to work faster
10:02and still be balanced.
10:04It was very difficult to stabilize
10:06up and down the stairs.
10:08And every time we felt
10:10that the border was closer.
10:12But luckily,
10:13we started to work on AI
10:14thanks to Nicolas five years ago.
10:17No one was talking about AI in robotics.
10:19And now we are AI-ing
10:21or AI-fying most of our algorithms,
10:26not everything today.
10:29And that helps us make the boundaries
10:31of robots and their motion
10:33very far away.
10:35I don't know if there are any,
10:36but we're going to run.
10:38We're going to walk in the sand.
10:40We're going to do stuff
10:41that the robots don't do yet,
10:42but they will.
10:44No, definitely.
10:44And the only limit today
10:46is to just to wait
10:47for the regulatory approval
10:49of the different versions
10:50of the exoskeleton.
10:52So we talked a bit about
10:53the medical applications
10:54for Wondercraft.
10:56But in a recent press release,
10:59I think two days ago or three days ago,
11:01you announced a new partnership
11:03with Renault
11:04to work on something different,
11:06new applications for Wondercraft.
11:09Can you maybe explain to us
11:10what could be the future
11:12for Wondercraft
11:13with these new applications?
11:16Of course.
11:18Also, five years ago,
11:20we put to the market
11:23the first exoskeleton,
11:24not the one that you saw,
11:26one that is a little bigger
11:27and it sells to hospitals
11:28to help people who had a stroke,
11:31for example,
11:33relearn how to walk
11:34with a physiotherapist
11:35and doctor, etc.
11:37And that sells very well.
11:38We're very happy.
11:39It's called Atalant.
11:41And then we realized five years ago,
11:43okay, this is the first
11:45humanoid robot on the market.
11:48There were Boston Dynamics.
11:49There weren't anything
11:51about Tesla figure, etc.
11:53There were prototypes in labs
11:54in universities.
11:56And we said,
11:57well, nothing industrial product
12:00are on the market.
12:01So we said,
12:02okay, we have a humanoid robot.
12:05It doesn't have arms.
12:08But that's going to be okay.
12:09So we started to work backstage
12:11on a humanoid.
12:13The reason why we started to work on a humanoid
12:16is that Oscar and Thibaut,
12:19they have weak arms, weak hands.
12:22So they can walk in the exoskeleton,
12:24but for most of their daily life activities,
12:27they need an aid.
12:30Usually it's a human.
12:31It could be a robot.
12:33Human aids are getting more and more rare.
12:36Demography makes that
12:37they won't have aids in the future.
12:38Same for old people.
12:40There will be more old people
12:41and less aiding humans.
12:43We need robots to help people
12:46with limited autonomy,
12:47not only exoskeletons.
12:49So that was our sort of
12:51load the star idea that we pursued.
12:53And we didn't make a full prototype.
12:56And then we met Renault at the right time,
13:00eight months ago.
13:02And we discussed with Renault
13:03who wanted to work with a specialist in robotics.
13:07They had been visiting all robot makers in China,
13:10the USA, etc.
13:11And they chose us.
13:12Which is quite cool because they didn't choose to work
13:15with all the names I quoted.
13:18But they thought that our technology was more reliable
13:21because we've been making exoskeletons
13:24that are safe with people for five years.
13:28And because they are stronger.
13:29A person can be 120 kilos.
13:32So our robots, exoskeletons,
13:35our robots are already very strong.
13:36Reliable, strong, with a strong team.
13:40And they chose to work with us.
13:43So yes, we have this partnership with Renault
13:45to develop humanoids, firstly in factories.
13:50And then we will also pursue my idea
13:53that there are 25 million people in Europe only
13:58with limited autonomy.
13:59There is going to be no one to help them.
14:02So there are 25 million robots out there.
14:06It's not tomorrow.
14:08The technology is not there yet.
14:09But in five, six, seven years, we can have them.
14:14So this new application wasn't the focus of the company,
14:18but it was a strong interest over the past few weeks.
14:22Maybe we can show a video of what you've achieved
14:25in just a few days, weeks, about the new applications
14:31for Wondercraft robots.
14:33Jean-Louis, if you want to comment on that.
14:34It's called Calvin40.
14:36So that's the name of the family of robots
14:38we're going to make.
14:40Calvin40 is a proof of concept, if you want.
14:43It's named Calvin40 because it was developed in 40 days.
14:47Literally from zero to this.
14:51And it does more things, actually.
14:53We're not showing everything.
14:54It has hands, for example.
14:57And actually, it's really based on the exoskeleton platform.
15:01We have a very strong platform.
15:03And we could, in 40 days, make that with arms.
15:07And the context mission management AI.
15:14And mostly everything that's needed to work in a car factory.
15:19Because the beauty of the agreement we have with Renault
15:22is not only they are investing in Wondercraft,
15:25but they'll be the first customer of our robots.
15:27We are going to work the data and the use cases with them.
15:31So we will be ready to sell to the whole supply chain of car making,
15:37of automotive, very soon.
15:40And very soon, actually, is very soon.
15:45And they are going to manufacture our exoskeletons and robots.
15:50And this is essential because there is only one...
15:54Anyone knows how much your car costs per kilogram?
16:00A car costs 10 euros per kilogram out of the factory.
16:071.5 tons, 15,000 euros.
16:11No one in the world can make such complex objects as cars at such a low cost than the car
16:19industry.
16:20So working with car makers will make our exoskeletons and robots probably 10 times cheaper
16:27in a relatively short future with the volumes.
16:31So it's really strategic to work with a car maker.
16:34We like four or five robot companies to have a strong tie with a car making company.
16:40Renault is actually the...
16:42You may not know that, but the highest gross margin car company in the world.
16:47So they are super high performance and it's French.
16:51So we're really happy to have been chosen and to have this stronger...
16:55And we are very ambitious.
16:57We're really going to be the European leader in humanoids.
17:02And so it doesn't mean for the company that you give up the healthcare applications.
17:06It means having in parallel two different streams,
17:10following your objectives in terms of healthcare delivery and exoskeletons,
17:16but also trying to build this new version of Calvin for manufacturing purposes.
17:22Well, Oscar still lives with me.
17:24So if we would stop Eve, he would kill me.
17:29No, our DNA is to help people who can't make something, make it.
17:35I think that's where the value lies.
17:38There's something that I can't do and we'll help you do it.
17:43And that's what our mission is.
17:45So you can't walk, we let you walk.
17:48You can't put on your shirt, you will have a robot to help put on your shirt.
17:51You can't make this job in a factory because it's like near fusion metal, 1500 Celsius.
17:59Let the robot make it.
18:01That's our job.
18:02So it's the same thing, the same technological platform.
18:06But Eve, which is the name of our exoskeleton that you saw walking by the seine,
18:11is going to the market next year.
18:13And that's going to be our key product for years.
18:16Of course, we keep it.
18:18And that's what made us live.
18:20So we're going to be an exoskeleton robot company.
18:24That's a good news for everyone.
18:27And for our investors.
18:29Of course, we invested on the healthcare thesis,
18:31but super happy also about the perspectives of these new applications.
18:37Jean-Louis, we have a few seconds left.
18:39Today, Wondercraft is already known in hundreds of hospitals, trying to support patients.
18:45Tomorrow, it will be present in the houses of hundreds of patients, but also in factories.
18:52How do you perceive Wondercraft in a few years?
18:57How would you define Wondercraft in five years?
19:00A bigger team of people super proud of helping others and selling thousands of exoskeletons
19:08and ten thousands of robots per year.
19:11Maybe hundreds of thousands of robots in a little more time.
19:16Great, amazing.
19:18Thank you very much, Jean-Louis, for this talk.
19:20And thank you, everyone, for attending to this talk.
19:23We were super happy to have you and to share a bit about the story of Wondercraft.
19:28Thank you, Maxime. Thank you, everyone.
Commentaires