This thesis investigates production-inventory systems where replenishments are received every period (for example every day or shift), but where production plans are determined less frequently (weekly, fortnightly, or monthly). Such systems are said to use staggered deliveries. This practice is common in industry, but the theoretical knowledge is limited to a small set of inventory models, none of which include capacity costs. This thesis uses time series analysis to expand our understanding of staggered deliveries from the perspectives of inventory and production-inventory control. The contribution to inventory theory consists in the development of an optimal policy for autocorrelated demand and linear inventory costs, including exact expressions for costs, availability, and fill rate. In addition the thesis identifies a procedure for finding the optimal order cycle length, when a onceper- cycle audit cost is present. Notably, constant safety stocks are suboptimal, and cause both availability and fill rate to fluctuate over the cycle. Instead, the safety stocks should vary over time, causing the availability, but not the fill rate, to be constant. The contribution to production-inventory theory comes from two perspectives: First, an optimal policy is derived for quadratic inventory and capacity costs; second, four pragmatic policies are tested, each affording a different approach to production smoothing and the allocation of overtime work (once per cycle, or an equal amount of overtime every period). Assuming independent and identically distributed demand, these models reveal that all overtime or idling should be allocated to the first period of each cycle. Furthermore, it is shown that the order cycle length provides a crude production smoothing mechanism. Should a company with long reorder cycles decide to plan more often, the capacity costs may increase. Therefore, supply chains should implement a replenishment policy capable of production smoothing before the order cycle length is reduced.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:681320 |
| Date | January 2016 |
| Creators | Hedenstierna, Carl Philip |
| Publisher | Cardiff University |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Source | http://orca.cf.ac.uk/87502/ |
Page generated in 0.0021 seconds