Constraining Streaming Flow Models for Adapting Learned Robot Trajectory Distributions

TL;DR

This paper introduces CASF, a framework that enhances streaming flow policies for robot motion by enabling real-time adaptation to safety and task constraints, maintaining smoothness and multi-modality.

Contribution

We propose a novel method to incorporate differentiable constraint metrics into streaming flow policies, allowing real-time trajectory adaptation during execution.

Findings

CASF effectively enforces safety constraints in real-time.

Trajectories remain smooth and feasible under constraints.

Outperforms standard post-hoc projection methods in experiments.

Abstract

Robot motion distributions often exhibit multi-modality and require flexible generative models for accurate representation. Streaming Flow Policies (SFPs) have recently emerged as a powerful paradigm for generating robot trajectories by integrating learned velocity fields directly in action space, enabling smooth and reactive control. However, existing formulations lack mechanisms for adapting trajectories post-training to enforce safety and task-specific constraints. We propose Constraint-Aware Streaming Flow (CASF), a framework that augments streaming flow policies with constraint-dependent metrics that reshape the learned velocity field during execution. CASF models each constraint, defined in either the robot's workspace or configuration space, as a differentiable distance function that is converted into a local metric and pulled back into the robot's control space. Far from restricted regions, the resulting metric reduces to the identity; near constraint boundaries, it smoothly attenuates or redirects motion, effectively deforming the underlying flow to maintain safety. This allows trajectories to be adapted in real time, ensuring that robot actions respect joint limits, avoid collisions, and remain within feasible workspaces, while preserving the multi-modal and reactive properties of streaming flow policies. We demonstrate CASF in simulated and real-world manipulation tasks, showing that it produces constraint-satisfying trajectories that remain smooth, feasible, and dynamically consistent, outperforming standard post-hoc projection baselines.