PENYELESAIAN VEHICLE ROUTING PROBLEM WITH SIMULTANEOUS PICK-UP AND DELIVERY MENGGUNAKAN GENETIC ALGORITHM

This research addresses the distribution route inefficiency problem faced by Pangkalan LPG 3kg Achmad Nasoha in Malang City, which has historically relied on driver intuition to determine visit sequences. This reliance on manual methods results in high total mileage and operational costs. The problem is specifically categorized as a Vehicle Routing Problem with Simultaneous Pickup and Delivery (VRPSPD), where the fleet delivers full cylinders while simultaneously picking up empty ones from 30 scattered customers. The main objective of this study is to design an optimal distribution route capable of minimizing the total distance traveled. To solve this combinatorial optimization problem, this study implements the Genetic Algorithm (GA) metaheuristic approach. The GA method was chosen for its robust ability to find near-optimal solutions for complex routing problems. Research data, including customer data and a symmetrical distance matrix obtained from Google Maps, were processed to identify the best solution. The results show a very significant performance difference between the initial and proposed conditions. The company's initial manual route recorded a total daily distance of 89.2 km. After the Genetic Algorithm implementation, the optimal proposed route was successfully generated with a total distance of only 68.27 km. The main conclusion is that the GA implementation provides an applicable and much more efficient route solution, evidenced by a distance saving of 20.93 km. This saving is equivalent to an efficiency increase of 23.46%, which has direct implications for a substantial reduction in fuel and vehicle maintenance costs, as well as an improvement in delivery time efficiency.