PENGEMBANGAN MODEL INTEGRASI PENJADWALAN DAN PENGADAAN DENGAN MEMPERTIMBANGKAN PROSES NEGOSIASI MENGGUNAKAN ALGORITMA OPTIMASI INTERAKTIF MULTI-OBJEKTIF

• Main Author: Brian Prasetya, Alexander
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/71054
• Institution: Institut Teknologi Bandung

The Indonesian fashion industry has a significant contribution towards the economy of the country, with a total revenue of up to US$ 13.323 trillion in 2021. However, the complexity of the industry requires collaboration and cooperation among stakeholders in the supply chain. One of the solutions for this problem is the fourth-party logistics (4PL) service providers, companies which provide integrated supply chain management for industries. PT. A is a 4PL company that provides services for local fashion MSMEs (Micro, Small, and Medium Enterprises) in Indonesia. Recently, there were problems regarding the order completion delay by up to 75% longer from its due date in one of its clients, PT. X. The tardiness phenomenon is caused by production scheduling and the delay in material arrivals from suppliers. Hence, integration between production scheduling, material order schedule, as well as reliable suppliers will be needed to fulfil the quantity, quality, and delivery due date requirements. Suppliers' preference for material delivery might differ from the clients' preferred production schedule. Hence, a mechanism to bridge this difference will also be needed. To solve this problem, a multi-objective optimization model that integrates production scheduling, material arrival planning, as well as supplier selection is constructed. Then, Interactive Weighted Tchebycheff (IWT) algorithm coupled with the forward filtering algorithm is deployed to solve the model. The algorithm is designed to generate multiple solutions as an alternative to the decision maker, so that strategic as well as subjective factors which are hard to be mathematically modeled can be considered. Based on the solution generation and negotiation simulation scenario, the model can slash order completion time by around 43%-56% compared to the existing policy, and also allow the incorporation of preferences of each stakeholder involved in the negotiation. The negotiation algorithm can reach the local optimum status under 1 hour of computation time.