PIER-Flow: Physics-Informed Efficient Rectified Flow for Real-Time Mobile Robot Navigation
2026-07-11 • Robotics
Robotics
AI summaryⓘ
The authors developed PIER-Flow, a method that helps robots move safely and quickly in crowded, changing spaces. They took a complex planning technique (MPC) and made it much faster and easier to use by turning it into a simple continuous process. Their approach keeps the robot's movements physically accurate and runs very fast, even on small computers, which makes it practical for real-world use. Tests showed their method works almost perfectly with almost no crashes and is much quicker than traditional methods.
Autonomous NavigationModel Predictive Control (MPC)Physics-Informed LearningOrdinary Differential Equation (ODE)Diffusion ModelsBehavior CloningReal-Time InferenceSim-to-Real DeploymentMobile RobotsLatent Sampling
Authors
Shibo Li, Zhongcheng Wang, Jiahe Cao, Jianhua Yang, Ke Wu
Abstract
Autonomous navigation in dense and highly dynamic environments requires both physically feasible control and low-latency replanning. Optimization-based methods such as Model Predictive Control (MPC) explicitly handle robot kinematics and safety constraints, but repeated nonlinear optimization can limit real-time responsiveness. Deterministic behavior-cloning policies enable efficient inference but may fail to represent multimodal avoidance behaviors, whereas diffusion policies capture multimodality at the cost of time-consuming iterative denoising. We propose PIER-Flow (Physics-Informed Efficient Rectified Flow), a lightweight navigation policy for mobile robots. By distilling an MPC expert into a continuous-time Ordinary Differential Equation (ODE), PIER-Flow achieves single-step action generation through parallel latent sampling and lightweight feasibility selection. We introduce a physics-informed training objective to enforce kinematic consistency, paired with an asynchronous action chunking architecture for robust sim-to-real deployment. Extensive simulations demonstrate that PIER-Flow achieves a 98.85\% success rate and zero collisions, with an average inference of $\sim$1.29 ms, which accelerates planning by 37.2$\times$ compared to MPC and over 800$\times$ against standard diffusion models. Crucially, real-world deployment on a resource-constrained edge computer further achieves an approximately stable inference latency of $\sim$5.3 ms, avoiding the latency spikes and freezing events observed with planning baselines.