Improving the throughput of automated holographic optical tweezers

The purpose of this work is to introduce three improvements to automated holographic-optical-tweezers systems that increase the number and speed of particles that can be manipulated simultaneously. First, we address path planning by solving a bottleneck assignment problem, which can reduce total move time by up to 30% when compared with traditional assignment problem solutions. Next, we demonstrate a new strategy to identify and remove undesired (e.g. misshapen or agglomerated) particles. Finally, we employ a controller that combines both closed- and open-loop automation steps, which can increase the overall loop rate and average particle speeds while also utilizing necessary process monitoring checks to ensure that particles reach their destinations. Using these improvements, we show fast reconfiguration of 100 microspheres simultaneously with a closed-loop control rate of 6Hz, and 10Hz by employing both closed- and open-loop steps. We also demonstrate the closed-loop assembly of a large pattern in a continuously-flowing microchannel-based particle-delivery system. The proposed approach provides a promising path toward automatic and scalable assembly of microgranular structures.