Proposed Distribution Routes of Kao Products in CV. Abro Mandiri Branch Pariaman

Fauzan, Rahman (2020) Proposed Distribution Routes of Kao Products in CV. Abro Mandiri Branch Pariaman. Diploma thesis, Universitas Andalas.

Preview	Text (Cover & Abstract) Cover & Abstract.pdf - Published Version Download (181kB) | Preview
Preview	Text (Chapter I Introduction) CHAPTER I Introduction.pdf - Published Version Download (131kB) | Preview
Preview	Text (Chapter VI Closing) CHAPTER VI Closing.pdf - Published Version Download (37kB) | Preview
Preview	Text (References) References.pdf - Published Version Download (180kB) | Preview
	Text (Skripsi Full Text) Final Project Report Full Text.pdf - Published Version Restricted to Repository staff only Download (8MB)

Abstract

CV. Abro Mandiri is an authorized distributor of PT Kao Indonesia. CV. Abro Mandiri distributes Kao products with coverage of Pariaman City, Padang Pariaman Regency, and a small part of the Agam Regency. The company implements a next day delivery system, so the process of loading goods is usually done in the afternoon or evening after the driver delivers the goods ordered on the previous day. Demand from customers is generally dynamic, making it difficult for companies to split utility vehicle capacity and companies apply fixed territories that are not optimal. This can be known from the distribution process exceeds the working hours time. For the determination of the route order visited, the company does not provide the order of stores to visit and gives the driver the full distribution process, so the distance of the visited outlet is not efficient. The approach used in this problem is the Capacitated Vehicle Routing Problem with Time Windows (CVRPTW) and the solution method with the development of genetic algorithms. The Nearest Neighbor algorithm is used to provide a qualified population to genetic algorithms because by default, the population used is random, so it takes a very long computing time, especially on a considerable number of instances/nodes. The use of the Nearest Neighbor algorithm is expected to make the solution of the genetic algorithm more convergent with less computing time. Genetic algorithms use fitness to determine the best solution, so it does not require trying all possible solutions. CVRPTW program with the Genetic Algorithm method and Nearest Neighbor Algorithm as Population Initialization is designed in R Studio version 1.25 and R programming language version 3.4.3 codename "Holding the windsock". Based on the results of data processing using R Studio with the distribution case study from 6 July 2020 to 11 July 2020, the results obtained from the distribution time savings up to 23.78 hours with an average savings of distribution time reaching 19%. Distance savings of up to 575.04 km and average savings up to 27% shorter than the actual route. Transportation costs get a total cost savings of up to Rp. 592,291. Using the Nearest Neighbor Algorithm as a population initialization for Genetic Algorithm is up to 84% shorter distance with the same iteration as the initialization of the random population

Actions (login required)

View Item