Scalabot Unveils HERON-CRA, Its Second Model Under HERON Tech Architecture
Scalabot on September 15 launched HERON-Context Reinforcement Action Model (CRA), the second model built on its HERON technical architecture, which features memory, reinforcement learning for error recovery, dexterous and deformable object manipulation, and cross-embodiment generalization to equip robots with continuously evolving real-world operational capabilities, shifting them from task execution to learning and iterative improvement.
the 2026 Bund Summit




