Case study: Buffer Stock Analysis for a Leading Networking Equipment Manufacturer
About the Client
The client is an American multinational corporation that develops and sells networking products. Its products include routers, switches, network management software, network security products, and software-defined networking technology.
The client manages the abrupt customer demand by having buffer stock at the SKU level. The buffer decisions at SKU level depends on the individual demand planners business experience which generates unnecessary inventory pileups. It also creates inflexibility to manufacture and sell alternate SKUs due to non-scientific buffer allocation to SKU mix. The objective of the project was to develop and demonstrate an algorithm to optimally allocate safety stock at the lowest component level.
BRIDGEi2i engaged with the client to deploy an optimizer solution that ran a quarterly check to recalculate bump-up factors for each segment. MRP was processed using the buffer calibration by the solution.
- BRIDGEi2i Data Experts carried out data cleansing and quality management.
- Data Experts created Master Data and creation of train set & test set.
- They also created an exploded view of the SKU level plans and consumption into lowest level parts.
- Data Engineers carried out the segmentation of components into actionable groups based on its commodity group and lead time.
BRIDGEi2i’s AI Labs followed a modern simulation based approach in contrast to traditional EOQ method. Material drive and shortages were calculated after applying optimal BF to BDP. ML algorithms were created to validate with out of sample SKUs.
BRIDGEi2i’s consulting team and AI Labs deployed the Optimizer solution which calculated the optimal ‘Bump-up’ factor based on marginal cost utility logic. The buffer settings were performed using simulation. The AI Lab experts additionally built an application to calculate bump-up factor. MRP was processed using the buffer calibrated by the application.