Perancangan Vehicle Routing Problem Pick-up And Delivery Menggunakan Algoritma Nearest Neighbor Dan Multi-start Adaptive Large Neighborhood Search (msalns) Pada Trayek Sekunder Pos Logistic Distribution Center Tambun Pt Pos Logistik Indonesia

Abstract

Abstrak PT.Pos Logistik Indonesia merupakan perusahaan distribusi pihak ketiga (3PL) yang memaksimalkan peluang bisnis logistik di Indonesia dengan mengoptimalkan jaringan PT.Pos Indonesia. Proses pickup and delivery problem (PDP) pada PT.Pos Logistik Indonesia menggunakan KBM yang beragam (heterogeneous), time window berdasarkan shift, dan permintaan muatan yang fluktuatif. Berdasarkan observasi dan wawancara dengan pihak perusahaan, penulis menemukan permasalahan yaitu utilitas KBM yang tidak optimal pada trayek sekunder. Berdasarkan target rata-rata utilitas PT.Pos Logistik pada trayek sekunder adalah sebesar 40% sedangkan persentase rata-rata utilitas kapasitas KBM yang dibandingkan dengan muatan masing-masing pada outgoing dan incoming adalah sebesar 18,702% dan 18,704%. Berangkat dari permasalahan tersebut, PT.Pos Logistik Indonesia membutuhkan perancangan rute yang dapat meningkatkan utilitas tiap kendaraan dengan mengunjungi lebih dari satu trayek yang akan selaras dengan tujuan meminimasi biaya operasional. Solusi yang diusulkan adalah perancangan rute menggunakan algoritma multi start adaptive large neighborhood search (MSALNS) yang didasari oleh dua fase, yaitu: fase I, mengelompokkan trayek berdasarkan jarak menggunakan algoritma nearest neighbors dan fase II menggunakan algoritma adaptive large neighborhood search (ALNS) sebagai solusi akhir dengan prosedur removal dan insertion heuristic. Dengan menugaskan kembali KBM pada tanggal 22-28 Desember 2019 berdasarkan rute eksisting dan membandingkan dengan solusi, peningkatan utilitas KBM sebesar 20% hingga 30%. Kata kunci : Vehicle Routing Problem (VRP), Pickup and Delivery Problem (PDP), Heterogeneous, Time Window, Multi Start Adaptive Large Neighborhood Search (MSALNS) Abstract PT.Pos Logistik Indonesia is a 3PL company that optimizing the PT.Pos Indonesia networks in Indonesia. The pickup-and-delivery problem (PDP) process at PT.Pos Logistik Indonesia uses heterogeneous KBM, time windows based on shift, and fluctuating cargo demand. Based on observations and interviews, the author found a problem, that the utility of KBM is not optimal on secondary routes. Based on the average of PT. Pos Logistik utility target on the secondary route is 40%, while the average percentage of KBM capacity utility for outgoing is 18.702% and incoming is 18.704%. Departing from these problems, PT.Pos Logistik Indonesia requires route design that can increase the utility of each vehicle by visiting more than one route which will be in line with the aim of minimizing operational costs. The proposed solution is to design a route using the multi start adaptive large neighborhood search (MSALNS) algorithm, which is based on two phases, namely: phase I, grouping routes based on distance using the nearest neighbors algorithm and phase II using the adaptive large neighborhood search (ALNS) algorithm as the final solution with a heuristic removal and insertion procedure. By reassigning the KBM on existing route on 22-28 December 2019 compared with the solution, the KBM utility increases by 20% to 30%. Keywords : Vehicle Routing Problem (VRP), Pickup and Delivery Problem (PDP), Heterogeneous, Time Window, Multi Start Adaptive Large Neighborhood Search (MSALNS)