caLogo

How MEC learned floor inefficiencies take a backseat to cutting lead-times.

When the Milwaukee Electronics Companies (MEC) decided to implement Lean manufacturing principles in 2003, its implementation team thought the major benefit would be increased throughput on the manufacturing floor. While those goals were achieved, the most significant lesson learned was the true bottleneck lay in supply-chain management.

There were a number of drivers for adoption of Lean principles, including:

  • More than 100 items were several weeks late.
  • Floor space was 100% utilized.
  • Improvements tended to be individual and project-based, with a low level of employee engagement.
  • Management was reactionary.
  • Legacy problems were not addressed.


MEC’s journey included improvement initiatives on the production floor, a focus on streamlining administrative procedures and development of an internal Lean culture. The first step was defining the state of the factory (Figure 1). Results of these efforts showed production time per piece was averaging 22.52 minutes, while the standard time assumption was 19.16 minutes. From a production standpoint, the production facility was underperforming. However, the most significant piece of data wasn’t the production time inefficiency. Instead, it was the six weeks of lead-time required for each product. The team labeled this area the “hidden factory” and focused improvement efforts there.


In identifying “muda” (waste), the team focused on the entire process and looked for:

  • Overproduction.
  • Material or work-in-process (WIP) in a wait state.
  • Unnecessary transport.
  • Overprocessing.
  • Excess inventory.
  • Unnecessary movement.
  • Defects.
  • Unused creativity.


In breaking down the hidden factory, the team did the math on the processing time and found that order processing and acknowledgment took three days; material acquisition time was typically 10 to 20 days, and factory lead time was seven days.

Improvement goals were set in three phases (Figure 2). In phase one, the goals were:

 

  • Reducing component lead-time.
  • Improving component availability by eliminating shortages.
  • Reducing component acquisition cost (i.e., purchasing transactional cost).
  • Reducing inventory carrying cost by improving the cash cycle of payables, receivables and inventory.


Process improvements in phase one included improved customer program management, partnering with suppliers, triggering component replenishment at point of use and replenishing inventory based on actual demand.

Customers with stable designs were managed differently under the new Demand Pull system. Initial replenishment bin size was set based on the customer’s data on historical demand and traditional lot size. Bin size is adjusted quarterly if demand trends change. In a stable program, bin size will typically need at least one adjustment during the first year, but usually the first adjustment of actual demand against assumptions will fine-tune the system. The goal has been to create an operations-to-operations-based system that eliminates the need for buyers at either the EMS provider or customer to generate discrete purchase orders, do inventory counts or expedite orders.

At the supplier level, there is also a Demand Pull program. Initially, a blanket purchase order with a non-binding estimated annual usage (EAU) may be used to set pricing. A standard release bucket size is developed for each component. Suppliers agree to identify non-cancellable, non-returnable (NCNR) inventory or other material that would fall outside the release bucket structure. The EMS provider supplies a non-binding weekly forecast with quarterly demand projections to support the suppliers’ needs for enough visibility to adequately maintain their pipelines.

Agreements may vary by program, but typically suppliers agree to supply release bucket quantities within three-to-five working days and also may agree to keep a second release bucket on hand to support variations in demand. The EMS provider and customer agree to accept liability of up to two releases, depending on stocking agreements within the program, as well as any special liability driven by unique component situations. This typically translates to a two-to-three week supply of raw material. Suppliers that support the Demand Pull program receive preferential consideration on new projects.

To reduce internal transactions at the EMS provider, bar-coded cards are created for each component. When raw materials bins are depleted, a production team leader scans the card, and the MRP system creates a new line on the purchase order to receive against, which triggers a replenishment release from the supplier.

In phase two, a supermarket for each customer is created. A mutually agreed on finished goods inventory kanban is stocked at the customer and replenished from an internal kanban at the EMS provider. In cases where there is a final system build, subassemblies also may be stocked in an internal supermarket for rapid assembly, in combination with a much smaller kanban of finished units.

Phase three replaces customer purchase orders with a replenishment trigger. In the new system, deliveries to the customer may be daily or weekly, depending on customer proximity to the facility and program requirements. With local customers, EMS provider employees may count the customer’s kanban inventory as part of their delivery activities and subsequently trigger replenishment. With more distant customers, consumption is monitored by the customer and replenishment is triggered by a customer email.

Lead-Time Reductions

The improved process cut order processing and acknowledgment to one day and factory lead-time to three days. However, the real improvement came in material replenishment time, which was reduced to five days. The result was a 75 to 83% reduction in overall lead-time. Production time was reduced to 16.8 minutes.

But the 120-plus Kaizen events generated additional benefits:

  • 40% more available floor space, while continuing to grow sales.
  • Floor WIP was reduced 80%.
  • WIP labor was reduced 80%.
  • Cycle time was reduced 80%.
  • On-time delivery typically exceeds 98%.
  • Annualized customer inventory turns when on the Demand Pull program typically average 12 to 15.
  • Customer purchasing and EMS customer service overhead was greatly reduced.
  • EMS finished goods inventory remains low.
  • Milk runs (deliveries to local customers) increase customer contact.
  • The SMT kanban increases the master scheduler’s flexibility.


Systemic Challenges

Leaning the supply chain is not without challenges. New suppliers often balk at adjusting their preferred order quantities to customer-preferred release quantities. Guarantees of preferential treatment on new projects tend to help this and as mentioned earlier, when truly necessary, exceptions can be made.

As products near end-of-life and release size shrinks, there is also some supplier concern. As products near end-of-life, they are removed from the kanban system and handled separately.

Eliminating muda offers both cost and productivity advantages. Schedule flexibility and teamwork are also enhanced. The result is a strong working relationship among suppliers, customers and the EMS provider, driving continued improvement across the board.

Don Sivilotti is quality manager at MEC Midwest (meccompanies.com); dsivilotti@mecmidwest.com.

Submit to FacebookSubmit to Google PlusSubmit to TwitterSubmit to LinkedInPrint Article
Don't have an account yet? Register Now!

Sign in to your account