Industrial robotics does not need another controlled laboratory demonstration. It needs AI that can work through the unpredictable conditions of a real warehouse.
Stuttgart-based Sereact has raised $110 million in Series B funding to scale Cortex 2.0, the next generation of its physical AI platform, and enter the US market.
Headline led the round, with participation from Bullhound Capital, Daphni and Felix Capital. Existing investors Air Street Capital, Creandum and Point Nine also returned. Sereact has now raised more than $140 million.
A robotic brain that plans before moving
Founded in 2021 by Dr. Ralf Gulde and Marc Tuscher, Sereact develops AI software that can operate across single-arm picking cells, dual-arm systems, fixed robots and humanoid machines.
Its current Cortex platform uses vision and language models to interpret unfamiliar objects and complete warehouse tasks. Cortex 2.0 adds a world model designed to evaluate several possible outcomes before the robot moves.
The system generates potential trajectories, assesses them for stability, risk and efficiency, and selects the strongest option. This moves the robot from reacting after a failed attempt toward anticipating failure before making contact.
That capability is important for tasks such as packing fragile goods, arranging components in a precise orientation or handling items that can shift during movement.
Real deployments create the data advantage
Sereact says more than 200 of its systems are operating across Europe, completing over one billion production picks for customers including BMW, Mercedes-Benz, Daimler Truck, PepsiCo, Austrian Post and Rohlik Group.
According to the company, only one in approximately 53,000 picks requires remote human support. These figures are company-reported and have not been independently audited by Tech Lens Media.
Every successful pick, failure and recovery is recorded and used to improve the central model. Updated policies are tested and then deployed back across the fleet.
This creates a compounding data loop: more robots in production generate more difficult real-world examples, which can improve the software running across every deployment.
Why the US expansion matters
The new funding will support product development and the opening of Sereact’s first US office in Boston, where it plans to hire commercial, engineering and application teams.
Warehouses were the company’s first market because they generate high volumes of repetitive but varied physical interactions. Sereact now plans to move into more precise manufacturing tasks, including assembly, component placement and industrial kitting.
The strategic bet is that robot hardware will become increasingly interchangeable, while the intelligence controlling it becomes the more defensible layer.
Sereact must still prove that Cortex 2.0 can deliver the same reliability across new machines, industries and tasks. But its production footprint gives it something many robotics startups are still trying to build: a large stream of operational data generated outside the lab.
The larger signal is that physical AI is moving from robot demonstrations toward reusable intelligence that can run across entire industrial fleets.
Source : Official Announcement
