QERRA-v2: Hybrid Quantum-Classical Ethical Safety Layer for Robot Deliberation – Recent Updates

Solo developer Marussa Metocharaki has released QERRA-v2, an ambitious open-source project aiming to embed ethical reasoning into autonomous robots. The system acts as a 'safety layer' for high-level robot deliberation—the mission control or planning component—by checking proposed actions against 12 ethical dimensions defined by the SEMEV-12 framework. It applies region-specific safety rules for the EU, USA, and UAE, uses the Detoxify library to screen for toxic or manipulative inputs, and generates a full audit trail for transparency. The long-term vision incorporates a quantum-inspired 'W-state' layer for exploring decision possibilities, with an 8-qubit test already run on IBM hardware.

Recent April 2026 updates have polished the core classical analysis engine, making it more stable and publicly accessible via a documented live API with a demo key. While the full hybrid quantum-classical engine is still simulated due to resource constraints, the functional classical safety kernel is ready for testing and integration. Metocharaki has also provided a basic ROS2 stub, creating a starting point for developers to connect QERRA-v2 to existing robot deliberation frameworks like Behavior Trees and planners. The project, licensed under AGPL-3.0, represents an experimental but concrete step toward making autonomous robot decision-making ethically auditable and regionally compliant.