Swedish robot "dog" Luna has received a real artificial "nervous system" IntuiCell
IntuiCell (Sweden), a startup launched in 2020 from Lund University, has introduced a robot "dog" named Luna, which is actually the first robot to have a functional digital "nervous system". This technology will allow robots to acquire their own "artificial nervous" perception. It is not only about responding quickly to commands, but also about the ability to work adaptively and autonomously, taking into account environmental conditions and even the "feelings" of the mechanical body.
IntuiCell's approach is based on imitating the biological nervous system and transferring its architecture to software. Luna doesn't need to be trained on massive databases or offline simulations - it "learns" directly in the real world, analysing its actions and mistakes in real time. As Viktor Luthman, CEO of the company, explained: "This is the first software that allows a machine to learn in the same way as a human or animal."
The essence of the technology is the implementation of a system that simulates the work of the peripheral and central components of the human nervous system. This means that robots will be able not only to detect touch, pressure or temperature, but also to make decisions on their own regarding reactions, such as pulling the manipulator away in case of excessive load or overheating.
In the demo videos, Luna is shown trying to stand and adapt like a newborn puppy or giraffe, gradually improving its balance without human intervention.
IntuiCell's goal is to extend the technology to more complex platforms, such as glumanoid robots, deep sea exploration, or interplanetary missions where pre-training data is not available. The use of such machines could be useful in extreme environments, such as on the Moon, Mars, or in natural disaster zones.
IntuiCell has received funding from Navigare Ventures (Sweden), SNÖ Ventures (Norway), and European Union funds. The company is now preparing to scale the system and apply it to various types of robots, from quadrupedal platforms to future mobile humanoid systems.
Swedish developers use an advanced artificial intelligence architecture that provides tactile and spatial feedback in real time. This allows the robot not only to sense external signals but also to learn from the experience it has gained, just as living organisms do. The technology is focused on applications in complex manufacturing, medicine, rehabilitation equipment and care.
According to the developers, the goal is not to imitate human intelligence as such, but to create an effective bodily interface for machines - capable of analysing, adapting and acting without the need for direct remote control.
In a global context, this could be a key step towards the emergence of more autonomous and safer robots that can operate in environments where human presence is limited or undesirable, such as hazardous materials facilities, deep-sea exploration, or natural disaster zones. It is expected that the first prototypes will be presented in 2026.
Source: 1e9.community
