Imaging-guided collision-free transport of multiple optically trapped beads

Automation of an optical tweezers system is critical to fully leverage its immense multiplexing capabilities, which allows for highly precise and reliable manipulation of microscopic dielectric objects. Although there has been substantial work in this context, automation has been limited to only transporting a few objects at a time, particularly while ensuring collision avoidances with other freely diffusing objects in the workspace. In this paper, we provide the first step in addressing this limitation by developing a new algorithm that combines a model predictive controller with a fast grid search-based path planning method. We validate the usefulness of the presented method through experiments with silica beads in water.