NEURA Robotics Unveils Neuron AI, a Multisensory Model for Robots
NEURA Robotics has introduced Neuron AI, a multimodal foundation model for Physical AI that combines 3D vision, touch, audio, robot state and language with memory. According to the company, the Perception-Language-Action (PLA) model lets robots act on information that cameras alone cannot capture and keep track of objects that leave their view.
- NEURA Robotics introduced Neuron AI, a Perception-Language-Action model combining 3D vision, touch, audio, robot state and language with memory.
- The model is trained on real-world robot data, simulation-amplified data and human demonstrations, and runs across five NEURA robot embodiments.
- Customers can fine-tune Neuron AI for their own tasks and deploy skills across fleets via the Neuraverse, with Time-to-Skill as the key measure.
- The humanoid 4NE1 will perform an eleven-minute autonomous laundry task at NVIDIA GTC Berlin, October 20 to 22.
- Neuron AI was developed and trained in-house in Metzingen, Germany.
Senses beyond vision, plus memory
NEURA says cameras capture geometry, color and position, but give no information about contact, force or sound. Neuron AI is designed around that gap. Rather than treating each sensor as a separate input, the model encodes 3D vision, tactile and force data, audio, robot state and language, aligns them in one shared space over time, and turns them into a single stream of action.
The company points to tasks where this matters. When a robot opens a lock, only sound and touch confirm that it is really open. With touch, a robot can grasp objects it cannot see. In welding, the quality of a seam can be judged from the sound of the process.
The model also has a memory function. NEURA describes object permanence, the ability to remember objects when they are out of sight, as something humans acquire in their first two years of life. Because a robot acts in a continuous physical world, it must track objects and the consequences of its own actions over time, often without seeing them. According to NEURA, it adapted the model's architecture specifically to introduce memory and trains it on dedicated datasets that use it.
Three data sources, five robot bodies
Neuron AI is trained on three kinds of physical experience:
- Real-world robot data: robots across different embodiments perform tasks in NEURA Gyms, the company's training infrastructure, and in real-world environments, generating synchronized vision, touch, force, audio and movement data.
- Simulation-amplified data: NEURA's Data Amplifier and simulation stack expand the training data through simulated variations of tasks and environments.
- Human demonstration data: people wearing sensor-equipped data suits capture movements and task demonstrations.
Together these sources train a shared foundation across five NEURA robot embodiments, with different grippers and hands. The company also names data partnerships with industrial companies including Schaeffler and Bosch, which complement training in the Gyms.
From a single skill to a robot fleet
The NEURA Gyms connect data collection, skill training and benchmarking before deployment, and the resulting data and skills inform later model releases. The Neuraverse is the platform for deploying and updating these capabilities across robot fleets. It also gives customers tools to fine-tune the model for their own tasks, turning process expertise into robot skills.
Customers can keep their skills private. Where they choose to share skills or deployment data, those contributions can support the wider ecosystem and future model development, according to NEURA.
The company names Time-to-Skill as a central measure of progress: how quickly a robot can learn a new task and perform it reliably. For industrial customers, NEURA says, shortening that time is essential to bringing Physical AI into more processes and adapting it as requirements change.
Laundry demo at NVIDIA GTC Berlin
At NVIDIA GTC Berlin 2026, NEURA will show its humanoid robot 4NE1 ("For Anyone") running an eleven-minute laundry-handling sequence with the current version of Neuron AI. The robot sorts laundry, loads a washing machine, closes and reopens it, unloads it and transfers the laundry to a dryer, without human intervention.
NEURA says the demonstration focuses on handling and task coordination, and that the challenge extends beyond any individual movement: the robot must carry context through the sequence and move from one step to the next. Visitors can see it at Station-Berlin, Hall 3, Booth #3031, from October 20 to 22.
What it means
Neuron AI was developed and trained in-house in Metzingen, Germany. David Reger, founder and CEO of NEURA Robotics, said: "With Neuron AI, we are building the brain of the physical world." For industrial users, the pitch is a single model that spans several robot types and can be tuned to their own processes via the Neuraverse. The source gives no pricing, availability dates or benchmark figures, so how the model performs outside NEURA's own demonstration remains to be seen.
NEURA Robotics: NEURA Robotics Introduces Neuron AI: The Brain of the Physical World
