A Machine-Learning-Assisted Simulation Approach for Incorporating Predictive Maintenance in Dynamic Flow-Shop Scheduling

In this paper, a machine-learning-assisted simulation approach for dynamic flow-shop production scheduling is proposed. This is achieved by introducing a novel framework to include predictive maintenance constraints in the scheduling process while a discrete event simulation tool is used to generate the dynamic schedule. A case study for a pharmaceutical company by the name of Factory X is investigated to validate the proposed framework while taking into consideration the change in forecast demand. The proposed approach uses Microsoft Azure to calculate the predictive maintenance slots and include it in the scheduling process to simplify the process of applying machine-learning techniques with no need for hard coding. Several machine-learning algorithms are tested and compared to see which one provides the highest accuracy. To gather the required dataset, multiple sensors were designed and deployed across machines to collect their vitals that allow the prediction of whether and when they require maintenance. The proposed framework with discrete event simulation generates optimized schedule with minimum makespan while taking into consideration predictive maintenance parameters. Boosted Decision Tree and Neural Network algorithms showed the best results in estimating the predictive maintenance slots. Furthermore, the Earliest Due Date (EDD) model produced the minimum makespan with 76.82 h while scheduling 25 products using 18 machines.

Penjadwalan Mesin dengan Metode Tabusearch Di PT. AAA

Penelitiaan ini dilakukan pada PT.AAA terdapat 7 stasiun kerja dengan posisi yang berbeda dan memproduksi jenis produk yang berbeda, perusahan ini memproduksi daun pintu untuk dipasarkan di seluruh bagian Indonesia terutama pulau Jawa. Pada perusahaan ini sering terjadi keterlambatan dalam penyelesaian pesanan dan waktu antar pesanan menyebabkan banyak konsumen yang merasa kecewa karena pesanan tidak sampai tepat waktu. Tabu search dengan inisial awal SPT karena susunan mesin pada perusahaan ini secara pararel. Pada metode Tabusearch total makespan sebesar 20.55 jam dengan efisiensi sebesar 1.1143. Urutan job yang terdapat pada iterasi ke-2 yang menjadi makespan minimum dengan urutan job 4-2-1-6-5-7-3. Dapat disimpulkan bahwa metode penjadwalan dengan algoritma Tabusearch memiliki makespan yang lebih kecil. This research was carried out at PT.AAA with 7 work stations with different positions and producing different types of products, the company produced leaf doors to be marketed in all parts of Indonesia, especially Java. In this company, there are often delays in completing orders and the time between orders causing many consumers to feel disappointed because the order does not arrive on time. The tabu search with the initial initial SPT is due to the parallel arrangement of the machines in this company. In Tabusearch method the total makespan is 20.55 hours with an efficiency of 1.1143. The sequence of jobs found in the 2nd iteration becomes the minimum makespan with the sequence of jobs 4-2-1-6-5-7-3. It can be concluded that the scheduling method with the Tabusearch algorithm has a smaller makespan.

Production Time Optimization using Campbell Dudek Smith (CDS) Algorithm for Production Scheduling

The production time optimization study used the Campbell Dudek smith (CDS) algorithm in the production process scheduling aimed at makespan optimization for engine operation to produce 12-size pan products, 14-size griddle, 16-size griddle, 18-size griddle, and 20-size griddle. The method applied by the Campbell Dudek and Smith (CDS) algorithm, CDS is a method used in flowshop-type scheduling developed from Johnson's rule that is able to minimize makespan 2 machines arranged in series. The CDS method is very suitable for production characters who apply the machine sequence to the production process. CDS produces several iterations that have makespan values, from the few iterations the most minimum makespan value is obtained to determine the order of products to be produced. This research produces an application that can schedule products to be produced by the machine automatically. From the results of testing with a total production of 12 pieces on each product with repetitions of 6 times, the minimum makespan value is 210.12 minutes with a work order of 20, grid 18, griddle 16, griddle 14, and griddle 12. Accuracy of results Application testing showed 99.99% for the first time and 99.96% for the second time when compared to manual calculations.