Optimal sequence-dependent job scheduling of packing process with multiple non-identical machines in tablet production
Conference proceedings article
ผู้เขียน/บรรณาธิการ
กลุ่มสาขาการวิจัยเชิงกลยุทธ์
ไม่พบข้อมูลที่เกี่ยวข้อง
รายละเอียดสำหรับงานพิมพ์
รายชื่อผู้แต่ง: Sodsangaroonngam A., Soontravanich C., Khompatraporn C.
ผู้เผยแพร่: Trans Tech Publications
ปีที่เผยแพร่ (ค.ศ.): 2014
Volume number: 548-549
หน้าแรก: 1046
หน้าสุดท้าย: 1050
จำนวนหน้า: 5
ISBN: 9783038350842
นอก: 1660-9336
eISSN: 1662-7482
ภาษา: English-Great Britain (EN-GB)
บทคัดย่อ
The Government Pharmaceutical Organization (GPO) is one of the largest pharmaceutical manufacturers in Thailand. GPO provides many medicinal products such as tablets and capsules to increase drug access to the Thai population through public hospitals and retail stores. Orders from these customers are accumulated and signal drug production. However, backorder shipments occur frequently and are estimated to worth about 180 million baths (or 6 million USD) a year. Initial study shows that the packing process behaves as a bottleneck, causing delays in the production and consequently to the drug shipments. The packing process utilizes multiple machines with different speeds and capacities. Moreover, instead of setup time in other manufacturing processes, drug production including the packing process requires a clearance process after each job-which is critical in pharmaceutical industry to reduce the risk of crosscontamination between drugs. The clearance time depends on the drug type of the next job. This paper presents a problem of job scheduling using parallel non-identical machines with sequence dependent clearance times. The objective is to minimize the total makespan. The mathematical model of the problem is a mixed integer/binary program. A numerical example from actual packing jobs is used to illustrate its application and solved by Excel Solver software. ฉ (2014) Trans Tech Publications, Switzerland.
คำสำคัญ
Parallel non-identical machines, Pharmaceutical industry, Scheduling problem, Sequence-dependent setup times