Specification-Aware Distribution Shaping for Robotics Foundation Models
2026-03-18 • Robotics
RoboticsArtificial Intelligence
AI summaryⓘ
The authors focus on improving robots that follow natural language instructions by making sure they follow safety rules and time-based goals during tasks. They introduce a method that adjusts the robot's planned actions on the fly to meet complex time-related rules without changing the robot's original program. This approach checks if the robot can still meet these rules before each action by predicting future steps. The authors tested their method on a top robot model in different simulations and with complicated task rules.
Robotics foundation modelsNatural language instructionsSignal Temporal Logic (STL)Spatio-temporal constraintsAction distribution optimizationForward dynamics propagationSafety guaranteesTime-dependent specificationsPretrained modelsSimulation validation
Authors
Sadık Bera Yüksel, Derya Aksaray
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. In this work, we propose a specification-aware action distribution optimization framework that enforces a broad class of Signal Temporal Logic (STL) constraints during execution of a pretrained robotics foundation model without modifying its parameters. At each decision step, the method computes a minimally modified action distribution that satisfies a hard STL feasibility constraint by reasoning over the remaining horizon using forward dynamics propagation. We validate the proposed framework in simulation using a state-of-the-art robotics foundation model across multiple environments and complex specifications.