Stress-Free Trapping in a 3D Microfluidic Treadmill

Microscale particles experience a stress-free environment when the external field is uniform and exhibits perfect translational symmetry. By preserving such uniformity in three-dimensional flows, we create a microfluidic treadmill that manipulates microscale particles under stress-free conditions. Using 3D vision-based feedback control implemented with machine learning, we dynamically adjust the treadmill's velocity to cancel any motion of the target particle relative to the surrounding fluid, thereby realizing stress-free trapping. This experimental platform enables the isolation of boundary effects on motile particles and effectively generates an infinite observation space under a microscope, providing new insights into gradual and large-scale processes such as Brownian sedimentation and microbial transport in unbounded environments.