In order to increase customer satisfaction and maintain customer loyalty, logistics service providers must pay attention to the quality of service provided, one of which is effecive warehouse management, especially in scheduling the arrival and departure of products transporting vehicles. Therefore, this study discusses warehouse management in form of delivery and pickup scheduling at PT XYZ’s cross-docking warehouse. This study aims to obtain effective delivery and pickup scheduling and minimize operational costs. The Cross-docking Distribution Problem is an np-hard problem, so the Particle Swarm Optimization algorithm is used, which is a metaheuristic method in finding solutions. Based on the result, it was found that effective delivery and pickup scheduling was able to save inventory cost by 3.12% and reduce the percentage of delays from 73% to 0%. The scheduling process using Particle Swarm Optimization requires an average computation time of 26.2 seconds.

Waste collection and transportation processes are accounted for more than 50% of all total waste management costs then there is an urgency for a waste recycling company to minimize this activities-related total cost by any means. Route optimization in collecting those plastics wastes can be a good solution to addressing the problem. This research focuses on determining waste collection routes for multilayer plastic with the objective function of minimizing total transportation costs. The waste collection problem is modeled as Vehicle Routing Problem (VRP) with a certain capacity and constrained time windows but here the vehicle is allowed to travel multiple times, this model said as the Multi-Trip Multi-period Capacitated Vehicle Routing Problem with Time Windows (MCVRPTW). The model was solved by an exact method using mixed integer linear programming and a metaheuristic approach using the Genetic Algorithm (GA). The results showed that the exact method was able to solve problems on small data instances and required almost 100% higher computation time than of metaheuristic approach. The best GA solution with mutation and crossover rates of 0.75 and 0.1 provides savings of 30.2% compared to the total transportation cost for existing conditions.

COVID-19 has been a popular issue around 2019 until today. Recently, there has been a lot of research being conducted to utilize a big amount of data discussing about COVID-19. In this work, we conduct a closed domain question answering (CDQA) task in COVID-19 using transfer learning technique. The transfer learning technique is adopted because a large benchmark for question answering about COVID-19 is still unavailable. Therefore, rich knowledge learned from a large benchmark of open domain QA are utilized using transfer learning to improve the performance of our CDQA system. We use retriever-reader framework for our CDQA system, and propose to use Sequential Dependence Model (SDM) as our retriever component to enhance the effectiveness of the system.Our result shows that the use of SDM retriever can improve the F-1 score of the state-of-the-art baseline CDQA system using BM25 and TF-IDF+cosine similarity retriever by 3,26% and 32,62%, respectively. The optimal parameter settings for our CDQA system are found to be as follows: using 20 top-ranked documents as the retriever's output, five sentences as the passage length, and BERT-Large-Uncased model as the reader. In this optimal parameter setting, SDM retriever can improve the F-1 score of the state-of-the-art baseline CDQA system using BM25 by 5,06 % and TF-IDF+cosine similarity retriever by 24,94 %. Our last experiment then confirms the merit of using transfer learning, since our best-performing model (double fine-tune SQuAD and COVID-QA) is shown to gain eight times higher accuracy than the baseline method without using transfer learning. Further fine-tuning the transfer learning model using closed domain dataset (COVID-QA) can increase the accuracy of the transfer learning model that only fine-tuning with SQuAD by 27, 26%.

In a distribution problem, designing the right distribution route can minimize the total transportation costs. Therefore, this research aims to design a distribution route that produces a minimal distribution distance by clustering the demand points first. We generated the clustering method to cluster the demand points by considering the proximity among the demand points and the total vehicle capacity. In solving this problem, we are using p-median to determine the cluster and a genetic algorithm to determine the distribution route with the characteristics of the CVRPTW problem. CVRPTW or capacitated vehicle routing problem with time windows is a type of VRP problem where there is a limitation of the vehicle capacity and service time range of its demand point. This research concludes that clustering the demand points provides a better result in terms of total distribution costs by up to 16.26% compared to the existing delivery schedule. The performance of the genetic algorithm shows an average difference of 1.73%, compared to the exact or optimal method. The genetic algorithm is 89.68% faster than the exact method in the computational time.

No abstract