ABPolicy: Asynchronous B-Spline Flow Policy for Real-Time and Smooth Robotic Manipulation

TL;DR

ABPolicy, an asynchronous flow-matching policy that operates in a B-spline control-point action space that introduces bidirectional action prediction coupled with refitting optimization to enforce inter-chunk continuity, is proposed.

Abstract

Robotic manipulation requires policies that are smooth and responsive to evolving observations. However, synchronous inference in the raw action space introduces several challenges, including intra-chunk jitter, inter-chunk discontinuities, and stop-and-go execution. These issues undermine a policy's smoothness and its responsiveness to environmental changes. We propose ABPolicy, an asynchronous flow-matching policy that operates in a B-spline control-point action space. First, the B-spline representation ensures intra-chunk smoothness. Second, we introduce bidirectional action prediction coupled with refitting optimization to enforce inter-chunk continuity. Finally, by leveraging asynchronous inference, ABPolicy delivers real-time, continuous updates. We evaluate ABPolicy across seven tasks encompassing both static settings and dynamic settings with moving objects. Empirical results indicate that ABPolicy reduces trajectory jerk, leading to smoother motion and improved performance. Project website: https://teee000.github.io/ABPolicy/.

Figures (6)

• Figure 1: Overview of ABPolicy. Our method enables real-time and smooth control via: (1) asynchronous inference to prevent execution stalls; (2) a flow-based trajectory generator predicting B-spline control points for intra-chunk smoothness; and (3) bidirectional prediction and refitting to ensure inter-chunk continuity.
• Figure 2: Overview of ABPolicy. A policy trained for Bidirectional Action Prediction (BiAP) asynchronously generates future B-spline control points at inference time. These are then optimized by our Continuity-Constrained Refitting (CCR) module to guarantee smooth continuity with the executed trajectory.
• Figure 3: Asynchronous inference overview. During inference delay, the robot executes the prior cycle’s actions. Gray shaded regions denote the inference-delay window, whereas orange shaded regions indicate the actions being executed.
• Figure 4: Manipulation tasks in static and dynamic settings, where dynamic tasks involve an object on a platform rotating at a constant velocity (approximately 10 seconds per revolution).
• Figure 5: Acceleration comparison between raw actions and our proposed method. The raw action representation produces large accelerations at the boundaries of action chunks, which leads to jitter. In contrast, our B-spline representation with refitting reduces boundary spikes and improves smoothness.