OPTIMASI MODEL CVRPTW BERBASIS GOOGLE-OR TOOLS PADA DISTRIBUSI CATERING BOX MENGGUNAKAN SAVING MATRIX DAN GUIDED LOCAL SEARCH

Ahmad Ghozali Rifani, NIM.: 22106060047 (2026) OPTIMASI MODEL CVRPTW BERBASIS GOOGLE-OR TOOLS PADA DISTRIBUSI CATERING BOX MENGGUNAKAN SAVING MATRIX DAN GUIDED LOCAL SEARCH. Skripsi thesis, UIN SUNAN KALIJAGA YOGYAKARTA.

Preview	Text (OPTIMASI MODEL CVRPTW BERBASIS GOOGLE-OR TOOLS PADA DISTRIBUSI CATERING BOX MENGGUNAKAN SAVING MATRIX DAN GUIDED LOCAL SEARCH) 22106060047_BAB-I_IV-atau-V_DAFTAR-PUSTAKA.pdf - Published Version Download (2MB) | Preview
	Text (OPTIMASI MODEL CVRPTW BERBASIS GOOGLE-OR TOOLS PADA DISTRIBUSI CATERING BOX MENGGUNAKAN SAVING MATRIX DAN GUIDED LOCAL SEARCH) 22106060047_BAB-II_sampai_SEBELUM-BAB-TERAKHIR.pdf Restricted to Registered users only Download (6MB) | Request a copy

Abstract

Inzan Catering Bantul distributes daily catering boxes to dozens of customer locations across the Bantul, Sleman, and Yogyakarta City regions. To date, route planning has been conducted manually, resulting in unbalanced courier workloads, overlapping routes, and suboptimal travel distances, which subsequently increase operational costs and degrade the quality of perishable products. This study designs a prototype of a web application named "Catering Optimizer" that integrates Google OR-Tools and the GraphHopper routing engine to automatically solve the Capacitated Vehicle Routing Problem with Time Windows (CVRPTW) model. The optimization method combines the Saving Matrix constructive heuristic to generate initial solutions and the Guided Local Search metaheuristic for global route refinement. Distance and travel time matrices are retrieved from GraphHopper in parallel using 16 threads, incorporating a dynamic traffic congestion coefficient (γ = 1,0–1,6) across seven operational time periods. Testing conducted with historical data from April 24, 2026, involving 89 customer orders, yielded a 16.3% reduction in travel distance (from 382.8 Km down to 320.4 Km per day) and a 25% fleet efficiency improvement (from 4 kurirs to 3 kurirs per slot). All 6 daily routes were declared 100% feasible within a 180-minute Common Due Date (CDD) without any vehicle capacity violations, achieving an estimated financial saving of IDR 707.567 per week. The Streamlit-based prototype was successfully developed and functionally tested, with a computation time of 35–70 seconds.

Share this knowledge with your friends :

Actions (login required)

View Item