One of the activities in the two-wheeled automotive industry is the process of sending components or parts from suppliers to motorcycle assembly plants. Delivery of parts using the fleet of each supplier. Along with the increase in market demand, the demand for components to suppliers is also increasing. This has an impact on increasing the number and frequency of fleets entering the factory area, besides that there is also an increase in transportation costs. This study aims to make the efficiency of the number, frequency and cost of fleet transportation by using the Milkrun approach and VRP (Vehicle Routing Problem). To get optimal results, the PSO (Particle Swarm Optimization) and GA (Genetic Algorithm) methods are used by utilizing the Matlab 2022 software. This research involved 40 suppliers, Logistic Partners and motorcycle assembly factories. The results of this study showed better results, namely before the research the number of fleets was 40 units and the frequency of 40 shipments with a total shipping cost of IDR 2,769,330, while after the research with the PSO optimization method the fleet number was 5 units and the frequency of 8 shipments with a total shipping cost of IDR 2,679,113. For GA optimization, the number of fleets is 4 units and the frequency is 8 times of delivery with a total shipping cost of IDR 2,000,058.