Abstract: Abstract: Plot breeding is a very important link of seed breeding, providing for the field experimental data and breeder seed or original seed. Seed harvested from the plot is commonly processed with manual tools or cleaned by simple machines. A batch seed cleaner is a piece of special equipment to meet the specific processing of germplasm materials in the experimental area. However, there are multiple operation procedures, high labor intensity, and low efficiency in seed batch processing. Manual operation with independent electric control systems has also caused uncontrollable clean-up quality, inefficient and time consuming. This study aims to develop a novel batch seed cleaner for the specific requirements of high efficiency, no residue and easy cleaning. An automatic control system was introduced to implement automatic operation using STM32 embed programming. A variety of functional modules were included, such as feeding margin detection, variable frequency vibration, high-frequency vibration cleaning, screen tilting adjustment, and terminal display with interactive interface and STM32 main control system. Hardware and software of the control system were designed for a sequence working flow, including the feeding, winnowing, and sieving. Three groups of photoelectric sensors were installed to monitor the location and status of the seed. Specifically, the No.1 photoelectric sensor was installed in the lower part of the hopper to monitor and control the feeding and air separation system. The No.2 and the No.3 photoelectric sensors were mounted on the two ends of the lower screen to control the screening device and vibration motor. An angle sensor was used to detect the angle of the screen. The key working parameters and functions were optimized, including the feeding speed, sieving vibration frequency, clean-up vibration, and tilt angle adjustment. The priority flag bit was used to optimize the sequential operation. An optimal control strategy was achieved for a stable and reliable system. A prototype performance test was carried out for high robustness of the automatic control system, thereby systematically investigating the influence on the quality and efficiency of seed batch cleaner. The cleaning test for corn seed showed that the coefficient of variation for the seed purity, the percentage of chosen seed, batch working time and working efficiency were 0.15%, 0.26%, 2.2% and 2.19% respectively, under the set working conditions. An optimal combination was gained for the quality requirements of plot seeding, where the cleaning process was smooth, while the automatic control system operated normally, and the batch seed cleaner worked reliably. This finding can provide a sound reference for the automatic control system in the intelligent batch cleaner for seed or granular materials.