The main objective of this project involves the production analysis for the Greensboro (GSO) Prototype Assembly plant of Qorvo. This plant assembles products based on orders obtained globally. When a customer needs a semiconductor prototype produced, they contact the Greensboro (GSO) Prototype Assembly plant with specifics of the job and the GSO selects a Factory Commit Date (FCD) for this job . The main problem that Qorvo has run into with creating their products is being able to determine an accurate FCD. Historically, the FCD has been determined arbitrarily in monthly meetings using historical data and an intimate knowledge of the manufacturing processes. The goal of our project is to create a simulation that can accurately produce a factory commit date that is more quantitative, accurate, and supported. The Qorvo production plant manufactures many different mechanical parts for semiconductors, defense and infrastructure technologies, and various other prototype technologies that are used around the world. Ensuring that the factory commit day is accurate will not only help the Qorvo be viewed as a more competitive manufacturing plant but it will also give the consumers to have a more reliable date for when incoming shipments can come in allowing them to be more efficient as well.
The scope that is covered during this timeline focuses on providing an accurate Factory Commit Date to GSO customers by creating a systematic way to schedule assembly jobs. To accomplish this task, we used Simio to create a model that will be used to develop a scheduling tool that takes arrival and processing times into account. The assumptions that were required in order to develop the best functioning model with the data we were provided are we simulated for historic production paths and not for new paths, time study data and other values sent by Qorvo are accurate and Qorvo will continue upkeep of the schedule once we have submitted the project To effectively model the Qorvo production floor we have created a Simio file that represents the machines and various routes that the actual manufacturing floor is operating with. In addition to this, we have compiled an Excel file that updates the processing time tables in the Simio model. Our sponsor, who works at Qorvo, was able to send us a time study that he Executive Summary for Qorvo conducted so we were able to put accurate processing times and routing information into the Simio model. Simio is the best way to model the facility’s system because it can compute the processing time data and any other stochastic or deterministic values in real time. It can also compute better estimates of the factory commit date after running experiments while accounting for the various data streams that were inputted into the model.
Given the short time span that we had, we had employed a limited scope for the values and information that we are able to put into the model. In the future, the employees at Qorvo should be able to add more routes that can resemble the paths that different orders take while being able to predict an even more accurate factory commit date. The model that we have configured is a basic map of the manufacturing floor that includes the various machines that products can be run through. We have accurate processing and wait times for the paths that our sponsor provided us but we were not able to include the entire list of routes that every product can go through simply because of the time constraint that we were put under. In order to keep the Simio model we have created in full effect, we recommended that Qorvo buy a business license to use and operate the Simio software. This model is only a relatively basic layout of the manufacturing floor and does not include every possible route that a product can go through which means that the Qorvo employees can definitely improve upon the model to fit the company’s needs better. Without a license to operate the Simio software this will be impossible to accomplish rendering the model we have created not as effective over time.
Although the COVID-19 virus caused our group to face some challenges obtaining the information and in-person data that was needed, we were able to come up with new ideas for constructing models and meeting the deliverable dates. The use of Zoom video conferencing, increased quantity of meetings, and a strong communication among the group and sponsors allowed us to complete the project and produce a simulation that Qorvo can use to help predict a more accurate FCD.
