Enforce formal safety and Signal Temporal Logic (STL) constraints on robotics foundation models without retraining.
arXiv · March 19, 2026 · 2603.17969
The Takeaway
The framework shapes the action distribution of black-box foundation models at runtime to ensure they meet hard spatio-temporal requirements (like safety bounds). This bridges the gap between the flexibility of foundation models and the rigorous safety requirements of real-world robotics.
From the abstract
Robotics foundation models have demonstrated strong capabilities in executing natural language instructions across diverse tasks and environments. However, they remain largely data-driven and lack formal guarantees on safety and satisfaction of time-dependent specifications during deployment. In practice, robots often need to comply with operational constraints involving rich spatio-temporal requirements such as time-bounded goal visits, sequential objectives, and persistent safety conditions. I