Grounding Generative Planners in Verifiable Logic: A Hybrid Architecture for Trustworthy Embodied AI
This work addresses the problem of trustworthy embodied AI for applications requiring strict safety guarantees, such as home safety tasks, providing a solution for developers and users of autonomous systems.
The authors tackled the problem of ensuring safety guarantees for physical deployment of embodied AI by introducing a neuro-symbolic architecture, achieving a 0 percent Hazardous Action Rate and a 77.3 percent Goal-Condition Rate. This was accomplished with an average of only 1.1 correction iterations.
Large Language Models (LLMs) show promise as planners for embodied AI, but their stochastic nature lacks formal reasoning, preventing strict safety guarantees for physical deployment. Current approaches often rely on unreliable LLMs for safety checks or simply reject unsafe plans without offering repairs. We introduce the Verifiable Iterative Refinement Framework (VIRF), a neuro-symbolic architecture that shifts the paradigm from passive safety gatekeeping to active collaboration. Our core contribution is a tutor-apprentice dialogue where a deterministic Logic Tutor, grounded in a formal safety ontology, provides causal and pedagogical feedback to an LLM planner. This enables intelligent plan repairs rather than mere avoidance. We also introduce a scalable knowledge acquisition pipeline that synthesizes safety knowledge bases from real-world documents, correcting blind spots in existing benchmarks. In challenging home safety tasks, VIRF achieves a perfect 0 percent Hazardous Action Rate (HAR) and a 77.3 percent Goal-Condition Rate (GCR), which is the highest among all baselines. It is highly efficient, requiring only 1.1 correction iterations on average. VIRF demonstrates a principled pathway toward building fundamentally trustworthy and verifiably safe embodied agents.