Sereact Scales Cortex 2.0 Across Humanoid Robots

Sereact just pulled in $110 million to power real robots in factories.

Sereact GmbH announced a Series B of $110 million to scale its Cortex 2.0 “robotic brain” and to push its first U.S. footprint, from Boston. The company says Cortex 2.0 is designed to work across embodiments and applications, threading a single AI core through single arm picking cells, dual arm returns stations, and humanoid robots, along with its 3D perception system called Lens. With the funding, Sereact is staffing up in the United States, opening in Boston to hire engineering, commercial, and applications talent as it trains a local go-to-market and services roster.

Engineering documentation shows that Cortex 2.0 is built to feed a data flywheel from live deployments rather than stand as a lab project. In practical terms, Sereact is betting that real-world operating data, how warehouses actually move, pick, and return items, will continuously refine the model without waiting for staged trials. In its press materials, the company highlights a simple but blunt metric: 200 deployed systems worldwide, one billion picks executed, and one human intervention per 53,000 picks. Demonstration footage shows the same brain reuse across a fleet of asset classes, including humanoid forms, which is intended to reduce the cost of expanding capabilities to new tasks.

The Cortex 2.0 momentum aligns with Sereact’s client roster already anchored in Europe. Daimler Truck, Mercedes-Benz, and BMW are named as customers, alongside logistics players like MS Direct, Active Ants, DeltiLog, Rohlik Group, and Austrian Post. This is more than a marketing line; it signals a move from very targeted pilots to production-scale deployment in complex environments where perception and manipulation must coexist with human coworkers. The company calls out that Cortex 2.0 can operate across embodiments, a feature meant to amortize software investments across hardware platforms. The message to investors and corporate buyers is that the same cognitive engine can be tuned for warehouse picking cells, returns stations, and humanoids, potentially reducing the time to onboard new automation assets.

Technology readiness, in practice, appears to sit in the field-ready camp. The company points to actual deployments rather than purely lab demonstrations, and the U.S. expansion is framed as the next step in broadening the installed base in real-world settings. The opening of a Boston office signals intent to recruit local talent to build, sell, and support systems in a market crowded with logistics automation vendors and startup AI firms alike.

Several practitioner-oriented questions follow every bold claim, even from a well-funded AI robotics play. The first is capability transparency: DOF counts and payload capacity for Sereact’s humanoids are not published in the current materials. That matters because real manipulation tasks hinge on how many joints the arms have, where the wrists can twist, and what grip strength the hands can sustain. Engineering documentation shows a focus on a cross-portfolio brain, but without published DOF and payload specs, engineers can only estimate whether a humanoid can lift, carry, or dexterously handle items in a warehouse aisle. The second concern is power and endurance. The company does not disclose power sources, runtime, or charging schemes for humanoids in their press materials, which makes it hard to assess uptime and maintenance requirements in continuous-operations facilities. The third caution is ecosystem risk. A single AI brain across robots sounds efficient, but it locks deployments to Cortex 2.0’s software stability, cyber hygiene, and update cadence. Any drift or regression in perception or planning could ripple across every embodiment.

Compared with Cortex 1.x lineage claims, Cortex 2.0’s emphasis on cross-embodiment operation and a data flywheel suggests a meaningful shift from siloed AI per product to a shared cognitive substrate. If these claims hold under independent benchmarking, the most tangible improvement would be reduced integration friction and faster task migration between cells, returns stations, and humanoids. But the article does not provide published benchmarks against the prior generation, so the reader must treat this as a strategic platform shift rather than a proven leap in a single metric.

Two to four industry takeaways emerge for R&D leaders and investors. First, the data flywheel model makes real deployments indispensable; the economics of continuous learning depend on scale and uptime, not a single successful pilot. Second, cross-embodiment AI can unlock faster expansion, but it raises the bar for system reliability, cybersecurity, and cross-vehicle maintenance planning. Third, the lack of disclosed DOF and payload for humanoids is a gating detail; without it, buyers cannot evaluate whether a given humanoid can perform the targeted manipulation tasks in their factories. Finally, the market timing is interesting: U.S. market entry coupled with a robust European footprint could pressure competitors to either accelerate or cede space in warehouse robotics where real-world learning drives the longest-term advantage.

Sereact’s funding round makes clear the industry’s appetite for AI-driven automation that can scale beyond a single robot model. Whether Cortex 2.0 becomes a broadly adopted backbone for humanoids and other assets will hinge on transparent performance, reliable hardware specs, and a rollout plan that keeps pace with the rapid cadence of industrial deployments.

Sources

Sereact gets Series B funding to expand Cortex 2.0 robot brain, enter U.S. market