Abstract: Coptis chinensis is one of the root and rhizome herbal medicines that contains multiple alkaloid components in its rhizomes. The ever-increasing market demand is often required in the conventional Chinese medicine industry. However, manual harvesting cannot fully meet the large-scale production in recent years, due to the high labor intensity and low efficiency. In this study, an alternating digging device was proposed for Coptis chinensis harvesting using multi-objective optimization. The alternating entry of the digging shovel into the soil reduced the power consumption of a single digging operation. Firstly, the kinematic modeling was performed on the crank-rocker rod mechanism in the digging device. A genetic algorithm was also selected to optimize the lengths of the rods in the crank-rocker mechanism. The maximum thrown-out velocity of the root-soil composite was taken as the objective function. While the rod length conditions, minimum transmission angle, digging depth, minimum soil cutting thickness, digging angle, digging depth, and the throwing position of the root-soil composite were taken as the constraint conditions. An optimal set of the rod lengths was obtained. Secondly, a three-dimensional model of the rhizome was established using image reconstruction. A discrete element composite model of the digging device-rhizome-soil was constructed using the discrete element method (DEM). The coupled simulation experiments of the digging were conducted with the multibody dynamic software RecurDyn. The movement and state of the root-soil composite indicated that the digging performance of the alternating digging device fully met the design requirements. The digging device successfully threw the root-soil composite onto the vibrating screen. Finally, a prototype of an alternating digging device was developed to verify the performance. In the digging phase, the rapidly cutting digging shovel was scraped against the surface of some rhizomes. As such, there was minor skin breakage in a small number of rhizomes. Nevertheless, there was no outstanding failure, because the rhizomes were enveloped by a dense network of fibrous roots and the resulting root-soil composite structure. In the throwing phase, the smaller rhizomes were occasionally slipped through the gaps between the shovel teeth. But the majority of the root-soil composites were larger than these gaps. Moreover, the rotational speed of the crank was relatively lower than before. Once the crank speed was too high, the acceleration of the digging shovel increased accordingly. There was a deformation in the fibrous root structures. Thus, some smaller root-soil composites were then slipped through the gaps between the shovel teeth to reduce the throwing qualification rate. Due to the spiral arrangement of the crank, the four sets of the digging shovels were achieved in the alternating digging operations. The efficient digging requires only a low-power motor to drive the device. Additionally, the better performance was achieved in effectively avoiding the entanglement of the fibrous roots with the digging shovels. The experiment showed that the digging device achieved the optimal stability and efficiency when the forward speed was 0.5 m/s and the crank speed was 300 r/min. The throwing qualification rate, soil loosening rate, and damage rate were 93.15%, 99.28%, and 2.95%, respectively. In summary, the digging performance of the alternating digging device fully met the requirements for the digging rhizomes. The findings can also provide a strong reference for the mechanized harvesting equipment for Coptis chinensis.