Neura says cognitive robots such as these MAiRA arms can learn to improve their performance. Source: Neura Robotics
In the race toward general-purpose robots, developers are focusing on different parts of the technology stack. Neura Robotics GmbH has been developing “cognitive robotics” including collaborative robot arms, mobile service robots, and a humanoid. These systems integrate sensors, proprietary artificial intelligence, and other components, offering a unified platform for Neura to develop applications with its partners.
The Metzingen, Germany-based company recently named Dr. Oliver Wolst, who has experience with automotive electronics from Bosch, as chief technology officer. He succeeded Jens Fabrowsky, who has become chief operating officer.
“Cognitive robotics is the most exciting technology field of our time,” Wolst stated. “Its potential to revolutionize our lives is immense.”
The Robot Report spoke with David Reger, founder and CEO of Neura Robotics, about the company’s approach to general-purpose robots and global competitiveness.
Cognitive robots gets capital to advance
In January, Neura Robotics raised $123.3 million in Series B funding. What was it like getting investment during a period of high interest rates and restrained venture capital?
Reger: The investors came to us, so we did not have any outreach. We made this investment round happen in less than a month, so that is a record. We basically had only inbound inquiries. We were talking to some investors for some time, but we were not in need of money.
Then, by the end of November, we just got a term sheet on the table. We also got other term sheets on the table before that we did not take, and this one was attractive, but with a time limit. They simply said, “We want to close it before Christmas.”
Most of the investors which had already talked to us decided to invest in a very short time, and we were also oversubscribed. It’s not about the need for cash; it’s more about how you need to spend it efficiently and how much dilution you want in what time.
Lately, the term “AI robots” has become popular. How is your education of the market going for “cognitive robots?”
Reger: Cognitive robots are AI robots. Nobody wants to use the term “cognitive robots” because if they Google it, Neura comes up.
We started doing this in 2019, what the world is now bullish about. Our unique selling point is that everything that most companies envision, we have already placed in the field. Cognitive robot arms are enabled to see and feel and are already streaming the right data to build AI models. You can start making your AI models on real data with a timestamp.
We’re German … we’re just proving that our technology is good, and we have a huge order book. Now it’s on us to enable the scale combined with very good quality for customer satisfaction. This is the next big hurdle, because scaling hardware is actually hard.
Register now so you don't miss out!
Neura follows plan for global expansion
As Neura Robotics continues growing its staff beyond 300 people, are you looking at U.S. locations?
Reger: We are investing in building our robot manufacturing right now. We’re the source of most components, which is good. We’re also releasing the Neuraverse platform for everyone, and we’re hiring a lot of people.
We have very ambitious targets, and we don’t just want to participate in the humanoid sector; we want to win. We’re hiring people in Detroit, Boston, and Silicon Valley. A lot of our partners are there, so we need to be close to them and support them.
Speaking of partners, are you looking for more relationships like those with Kawasaki and Omron?
Reger: We have some partnerships in the humanoid sector, and we’re working with Delta Electronics. There are some partnerships with automotive groups, but for now, they want to keep it to themselves. Our go-to-market strategy is that we’re an enablement company. We’re building our teams around them, not to compete.
Neura has developed a wide range of systems for industries including manufacturing. Source: Neura Robotics
Mobile manipulation a key step to a robotic future
You have the 4NE-1 platform — when do you think humanoid robots will be out in the world?
Reger: We see a lot of humanoid companies popping up everywhere, and it’s a good thing having all these form factors, because then you can judge. Right now, we’re seeing two arms on an unstable platform.
Humanoids will probably converge on one form factor. I believe that every robot we have in the filed with cognitive abilities contributes to the development of humanoids.
We’ve showed MAiRA building with blocks and then the same functionality with a humanoid with two arms, but the accuracy is not there. In my opinion, the issue to be solved is actually closing the loop with having enough sensor integration to close the simulation-to-reality gap.
We’re working hard to get there. The new Gen 3 will have more sensor capabilities and more power in its legs to be a bit more efficient. We’ll show it at Automatica in Munich.
The MIPA service robot. Source: Neura Robotics
What other models are you working on?
Reger: We also now have the third generation of the MIPA service or household robot coming up. This platform is basically a mobile manipulator with two arms, but this is the faster way of getting closer to human [capabilities].
You can already gain a lot of data and make this physical AI brain more efficient. And you’re working without fighting with instability. That’s why I’m always telling my team, it’s not about making a backflip.
It’s important to focus on getting the actual work done, like assembling, gluing things, filling up dishwashers. These are things you can do on a mobile platform with wheels.
Out of Neura’s different robots, which is seeing the most demand?
Reger: We’re selling LARA, MAiRA, and MAV to the market right now. If you’re looking at the quality piece, then it’s more in the LARA space. Quantity-wise, MAiRA is catching up, and I think it will be bigger because it’s cognitive, and you don’t have to add on like with LARA.
Automation experts know there’s still some way to go for humanoids, so they’re now deciding to go with mobile manipulators. But mobile manipulation only makes sense if you have cognitive robotics on top. If it’s not precise, and you’re plus or minus 5mm, you have to program everything.
MiR and UR are having success by adding a camera, making their systems cognitive. This space will expand as more integrate vision on the mobile platform.
How do you handle specialization versus generalization?
Reger: The beauty of our go-to-market strategy is that we’re specialized in robotics, while our partners are specialized in markets. This is still mostly industrial uses cases like automotive and aerospace, and now retail is coming up a lot. Automotive is having issues, so you will see some things go up while that market actually drops, but we’re still just getting started.
China has said that by 2030, it wants to have more than 40 million robots, so you need 8 million robots a year to reach that number. This is the driver. We want to keep up as the rest of the world catches up.
Robots could help Germany be competitive
Neura moved production from China to Germany. You’ve also stated that cognitive robotics can make German automakers competitive again, after the VDMA Robotics + Automation Association expressed concern. How can robotics help German and European industry?
Reger: Germany has been a market leader, and we got comfortable. I see the need for this country to come up with new technology that could be an economic pillar, and it has to be robotics.
In Europe, we love to regulate and to do other things first before innovate, but I still have a lot of hope.
How is the government in Berlin at recognizing that?
Reger: Right now, on the state level, robotics is getting a lot of attention. There are plans for 2030, 2035 — this is a change that we’re super proud and happy about.
In our company, we have 38 nationalities, and this gives us a huge strength.
Tell Us What You Think!