In this study, a system of automatic microparticle patterning that could enable the separation, trapping, and translation of single microbeads in liquid suspension using negative dielectrophoresis (DEP) tweezers was presented to form a single-bead pattern. A microchip with integrated electrodes was flipped and placed above the substrate through a micromanipulator. Microparticles laying on the substrate could be displaced to different positions relative to the electrodes on the microchip, and only the selected particles would be trapped by the electric fields generated from electrodes. Vision-based approaches were used to evaluate the necessary information, such as the gap distance and the positions of electrodes and microparticles in the image. A strategy for separating nearby particles was proposed to achieve single-bead patterning with high accuracy. A controller was used to guide the microparticles toward the position for trapping while avoiding flow disturbance. Different strategies were simulated to decrease the patterning time and find the minimum traveling distance and the best route of movement. The optimization problem is NP-hard. Hence, global optimization algorithms, such as genetic algorithm, particle swarm optimization, and ant colony optimization (ACO), were simulated, and the results were compared with those of the local optimization method. The comparison results showed that ACO obtained the best performance among the methods. The strategy for constructing high-quality microparticle patterns was also examined through experiments. Orange fluorescent polystyrene beads suspended in 6-aminohexanoic acid solution were considered and successfully patterned on a glass substrate by using the proposed system. Note to Practitioners—Micropatterning is an effective tool for pharmaceutical research and drug discovery. However, the reliability of results depends on the quality of patterns. Existing approaches, such as microfluidic devices, are limited to create a pattern from one chip for single use, and the entire process is sealed and isolated from the environment. In this study, a multielectrode microchip combined with a vision-based micromanipulator is introduced to create a novel noncontact approach for microparticle patterning, which offers high flexibility and guarantees the quality of the constructed patterns. The electrodes on the chip can be selectively energized to determine the shape of the final pattern. A real-time screening is performed so that the micromanipulator will only guide the particles in good condition for selection. An optimization algorithm is implemented to aid the particle selection with the electrodes, allowing high-quality microparticle patterns to be constructed in a short time for various applications. [ABSTRACT FROM AUTHOR]