Simulation Based Modeling of Inventory Policies and Operating Procedures in Complex, Low-Volume Electronics Manufacturing
Bibliographic record
Abstract
This simulation study considers a low-volume manufacturing system, which produces complex, customized electronics. Modeling demand as a renewal-reward process, the simulation, inspired by the production system and available data from a Canadian company, examine the performance of alternative inventory policies and operating procedures. Performance indicators that measure the responsiveness and inventory on hand show trade-offs between them in order to supply relevant information to decision makers. Experiments compare make-to-order and make-to-stock scenarios with various inventory parameters as well as introducing variability to examine the modelâs robustness under uncertainty. \n\nThe system under consideration consists of three main processes to manufacture a finished product from raw materials. The first process fabricates metal and electrical components from raw materials. Second, a worker assembles components into a semi-finished product. The third requires information from the customer in order to customize the product according to their needs, and test the unit to ensure its quality. The company, known for their well-designed products and exceptional customer service, wants to improve the accuracy of their leadtime promising. The current MRP control system assumes a completely make-to-order environment where every piece of WIP has a customer order attached to it. However, a forecast of orders likely to materialize from the sales quotes allows production to initiate jobs before the actual order arrives.\n\nThe approach taken to analyzing this system involves studying the make-to-stock, make-to-order decision at two stock points, components and semi-finished units. The operating procedures examine four possible stocking strategies: holding no inventory, holding only component or semi-finished inventory, and holding both components and semi-finished units. Simulation experiments determine the trade-off between holding inventory and the responsiveness to the customer for each operating procedure. Sources of randomness introduced to processing time, capacity, and demand, show how they respond to added variability. \n\nThe simulation experiments indicate that holding no inventory, and waiting for a customer order to initiate jobs, results in unstable performance. In order to achieve a stable make-to-order system, it would be necessary to have a fifty percent reduction in demand or product cycle time, a capacity expansion, or forecasting method. In the absence of an accurate forecast model, holding inventory is necessary for an acceptable level of performance. Component inventory is useful as many components are common among a number of products. Suitable component inventory can lead to customer orders typically fulfilled within two weeks. Adding semi-finished inventory can reduce the customer lead-time to under a week though requires stocking at least a few of each semi-finished unit. Holding semi-finished inventory without component stock is possible. However, it is necessary that the replenishment quantity be three or more units ordered at a time. Otherwise, the setup time for components exceeds the allowable limits and resource queues become unstable, much like the completely make-to-order scenario. Using an order-up-to parameter for semi-finished stock can further decrease the setup time incurred per unit. \n\nThe model is robust to randomness in job times, though it is component stock, which provides an effective buffer to this variability. Machine breakdowns begin to affect responsiveness measures if the average time for repair is greater than a week. Reducing the capacity in the assembly and testing processes can provide the same level of service indicating the two resources are underutilized. The analysis of this system shows the current make-to-order model requires some forecast to function in steady state, which is difficult to model without information on the current forecasting processes. Expanding the simulation model to incorporate forecasting or some other means of analysis can improve its accuracy and credibility as a management decision tool.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".