I have experience in a multi-national organization that manages a supply chain that sources goods from all over the globe. The majority of our material supply filters to a main assembly location in the US where the components are configured into finished goods, then shipped out to customers worldwide. Our product line is highly configurable, which means we have a vast array of possible saleable configurations, which may include options such as functionality, colors, textures, etc. Over the past 10 years we have transformed from a batch manufacture-on-site operation to a lean model of assemble to order. We employ JIT strategies for major and minor components, and have been crafting our product designs and BOM’s to match this strategy.
While our organization is not huge in comparison to the ‘big guys’, we are a solid mid-cap operation. Quite a bit of effort is put into the coordination of the supply chain to keep the production line flowing, keeping in mind we are mainly using a single manufacturing location.
As our international product sales increase, it makes obvious sense expand final assembly locations to sites closer to our end customers. This will allow faster placement of product in international locations, as well as potentially decreased shipping costs, import fees and tax advantages. This provides a number of challenges to the operation when it has been developed around a single main warehouse and assembly location. In the past year our organization has taken steps through acquisition to open manufacturing locations on different continents to service local markets, which has presented challenges in a number of areas.
From a manufacturing philosophy, these satellite assembly locations do not, and will not maintain the same level of component stock needed to service all the potential product options. The satellite locations also do not maintain a local supply base to pull Kanban product from, which is causing batch shipments of goods to these locations. In the short term, these satellite locations are receiving goods shipments by shipping container, which means high quantities of components delivered in irregular and slow-to-react shipments. This is effectively driving batch assembly, where goods are received from shipping containers, built to semi-finished goods then stocked until being pulled for a customer order. I expect this operation to evolve over time into a more ‘lean’ model for these international facilities, but in the meantime the design cycle and release process is being asked to accommodate 2 different models.
What does this mean? Currently the company products and BOMs are designed for a lean manufacturing model. Finished goods are constructed from a large number of building blocks, each of these building blocks in managed through the procurement group based on a customer order. When a customer places an order for product XYZ, and it needs to be in colors ABC with a certain list of functionality and options, the procurement group orders or pulls those building blocks from the supply base and they arrive on site on a scheduled day. The product is assembled and tested, then shipped the same day. Lead time can be 2-6 weeks with minimal internal stock. I call this the Lego approach where engineering develops BOM’s of the building blocks that are easily configured and snap together to form the overall product.
With the addition of satellite assembly locations with limited local supply base and larger shipments by overseas container, this Lego approach falls apart. In order to be effective in these satellite locations to service customer demands, the incoming product shipments need to be pre-configured to minimize the number of components that need to ‘appear’ on the production line at the right time and place. From a design side, this means pre-configuring BOM’s for popular product configurations. This sounds simple, but in reality it means the product developers will be maintaining another level of BOM’s, one for the Lego blocks and another for the configured Lego sets. More redundant BOM’s means more opportunity for mistakes, among other issues.
We are currently looking for BOM design strategies to accommodate both flow and batch models in the short term, while allowing flexibility as the models converge over time. I know lots of firms out there have worked through these issues, would like to hear some thoughts.
Please leave a comment.