Master production scheduling

S&OP agrees volumes by product family and month, for example 1 lakh units of shampoo in December. The plant can’t build “shampoo”; it builds specific items: the 200 ml anti-dandruff pack, the 650 ml family pack. The master production schedule (MPS) turns the family plan into a plan for each finished item, week by week.

A simple MPS takes demand (forecasts plus confirmed orders) and stock on hand, then plans production so stock doesn’t run out:

Three ideas come up in interviews:

• Time fences: the next week or two are usually frozen, because materials are already bought and lines are booked. Changes there need approval; further out the plan is open to change.
• Feasibility: before release, a rough-cut capacity check tests the MPS against bottleneck lines.
• Promising: the MPS shows what customers can be promised, its available-to-promise quantity.

The MPS then drives MRP, which works out every component and raw material needed to build it.

The MPS record

For each finished item, the MPS keeps a weekly table:

• Forecast and customer orders already booked.
• Projected available balance: last week’s balance plus MPS production minus demand. Near term, demand is usually the larger of forecast and orders.
• MPS quantity: what is planned to be produced, in sensible lot sizes.
• Available to promise (ATP): the part of stock and planned production not yet promised to customers.

The projected balance plans production against the forecast; ATP protects promises against orders. Both are needed.

Two time fences

• Demand time fence: inside it, only customer orders count as demand; the remaining forecast is assumed not to materialize.
• Planning time fence: inside it, the system doesn’t create or move MPS quantities automatically; the master scheduler decides.

Checking feasibility

Before the MPS is released, a rough-cut capacity check tests it against the bottleneck resources, and the master scheduler resolves overloads by moving quantities, adding capacity or escalating to S&OP.

What to master-schedule depends on the business

• Make to stock: schedule finished items, as in the shampoo example. There are relatively few end items and many customers.
• Assemble to order: a few modules combine into thousands of end products (laptops, cars, industrial pumps). Scheduling every combination is hopeless, so the MPS plans the modules and options, and a separate final assembly schedule builds each specific product once an order arrives.
• Make to order: each order is planned when it arrives; the MPS manages capacity and long-lead materials rather than end items.

Planning bills and hedging the mix

In assemble-to-order, a planning bill describes an average product as percentages of options. If laptop demand is 10,000 a month and 60% of buyers choose 16 GB memory and 40% choose 8 GB, the MPS plans 6,000 and 4,000 memory modules. Because the mix itself is uncertain, many companies overplan the options, for example by 10%: 6,600 and 4,400, a total of 11,000 modules for 10,000 laptops. The extra 1,000 is a buffer against the mix, not against the total, and it is much cheaper than holding finished laptops in every configuration.

Stability against responsiveness

Each change to the MPS ripples down through MRP to suppliers and the shop floor. Common practices:

• Freeze the near term and measure schedule adherence: how much of what was planned was actually built.
• Allow changes beyond the fence only through the master scheduler, who checks capacity and materials.
• Review the MPS weekly, not daily, unless the business truly needs it.

The bow wave

When an MPS is overloaded, unfinished quantities roll into the next week, adding to an already full plan. Week after week, past-due quantities pile up at the front of the schedule like a bow wave in front of a ship. The plan loses credibility, and people start working from their own lists. The cure is an honest MPS that fits capacity, with the overload resolved by decisions rather than carried forward.