Production systems and the planning hierarchy

Why it matters

Before any forecast, schedule or inventory model is applied, a plant has to know what kind of production system it runs and at which level a decision belongs. A job shop and a continuous chemical plant need very different planning tools, and a decision taken at the wrong level (for example, trying to fix a capacity shortage with daily dispatching) wastes effort. This topic is the map that the rest of production planning and control (PPC) fits into.

Key ideas

Production system. A production system converts inputs (materials, labour, machines, energy, information) into outputs (goods or services) through a transformation process, with feedback and control. Systems are classified mainly by volume (how many of each product) and variety (how many different products).

Types of production system (high variety → high volume)

• Job (job-shop / project) production – one-off or very small quantities made to customer order. General-purpose machines, highly skilled labour, process (functional) layout, long and variable throughput time, high work-in-process (WIP), high unit cost. Examples: tool rooms, special machinery, ship building (project).
• Batch production – moderate quantities of several products, made in lots that are repeated at intervals. Machines are changed over (set up) between batches, so set-up time and batch size become key decisions. Process layout or cellular layout. Examples: pumps, machine tools, pharmaceuticals.
• Mass (flow-line) production – very large quantities of one or a few standardised discrete products on dedicated lines balanced to a cycle time. Product layout, special-purpose machines, low skill per station, low unit cost, low flexibility. Examples: automobiles, two-wheelers, bulbs.
• Continuous (process) production – very large volume of a homogeneous product that flows without interruption, often 24 h a day. Highly automated, very high fixed cost, very low variable cost. Examples: refineries, cement, fertiliser, steel rolling.

The product–process matrix puts these on a diagonal: as volume rises and variety falls, the process moves from job shop to batch to line to continuous. Operating off the diagonal (for example, a dedicated line for a product sold in tens per year) usually costs money.

Make-to-stock vs make-to-order. Make-to-stock (MTS) produces against forecasts and serves customers from finished-goods inventory (typical of mass and continuous). Make-to-order (MTO) and engineer-to-order (ETO) start work after the order arrives (typical of job shops). Assemble-to-order (ATO) stocks modules and assembles to order. The choice decides where inventory sits and what the MPS is stated in.

Planning hierarchy. Planning decisions are nested; each level works on a longer horizon and with more aggregate data than the one below it, and sets limits for it.

1. Strategic (long range, 2–10 years) – business plan, product lines, plant location, facility capacity, process technology, make-or-buy. Data in money and product families.
2. Tactical (medium range, about 3–18 months) – sales and operations planning, aggregate production planning (workforce, overtime, inventory, subcontracting by month for product families), resource planning.
3. Operational (short range, days to weeks) – master production schedule (end items by week), material requirements planning (components and purchases), capacity requirements planning, then scheduling, sequencing, dispatching and shop-floor control (jobs by hour or shift).

Information flows down as plans and limits (disaggregation) and up as feedback (actual output, capacity problems, inventory). A later topic covers each level in detail.

Basic production-system measures (batch and flow systems)

• Batch time = set-up time + batch size × cycle time per unit.
• Production rate = units per batch ÷ batch time. Set-ups lower the average rate; bigger batches raise it but raise inventory.
• Production capacity = number of work centres × shifts per period × hours per shift × production rate. Capacity is an upper limit; utilisation = actual output ÷ capacity.

Formulas

T_b = T_su + Q · T_c

• T_b = time to process one batch (h), T_su = set-up time (h), Q = batch size (units), T_c = cycle time per unit (h/unit). Applies to one work centre processing one batch.

R_p = Q / T_b

• R_p = average production rate (units/h). For a flow line with no set-ups, R_p = 1 / T_c.

PC = n · S_w · H_s · R_p

• PC = production capacity (units per period), n = number of identical work centres, S_w = shifts per period, H_s = hours per shift (h). Use when all work centres make the same kind of item.

U = actual output / PC

• U = utilisation (fraction or %).

Q_BE = (F₂ − F₁) / (v₁ − v₂)

• Break-even volume between process 1 and process 2 (units per year), F = fixed cost per year (₹), v = variable cost per unit (₹/unit), with F₂ > F₁ and v₂ < v₁. Above Q_BE the high-fixed-cost process is cheaper.

Worked examples

Example 1 (standard) – batch production capacity. Given: set-up time T_su = 2 h, batch size Q = 200 units, cycle time T_c = 3 min/unit. The machine works 5 days a week, 2 shifts of 8 h.

1. Convert cycle time: T_c = 3/60 = 0.05 h/unit.
2. Batch time: T_b = T_su + Q · T_c = 2 + 200 × 0.05 = 2 + 10 = 12 h.
3. Production rate: R_p = Q / T_b = 200 / 12 = 16.67 units/h (without set-up it would be 20 units/h).
4. Available hours per week = 5 × 2 × 8 = 80 h.
5. Complete batches per week = 80 / 12 = 6.67, so 6 full batches.
6. Weekly output from complete batches = 6 × 200 = 1200 units per week.

Notice that the set-up consumes 2 of every 12 hours (16.7 %), which is why batch plants work hard on set-up reduction.

Example 2 (GATE level) – choosing a production system by break-even analysis. Three process options for a new part:

• General-purpose machines (job shop): F₁ = ₹2,00,000/year, v₁ = ₹150/unit.
• CNC batch cell: F₂ = ₹5,00,000/year, v₂ = ₹90/unit.
• Dedicated transfer line (mass): F₃ = ₹12,00,000/year, v₃ = ₹40/unit. Expected demand = 9000 units/year. Find the break-even volumes and the cheapest process.

1. Total cost TC = F + v · Q.
2. Job shop vs CNC: Q = (5,00,000 − 2,00,000)/(150 − 90) = 3,00,000/60 = 5000 units.
3. CNC vs transfer line: Q = (12,00,000 − 5,00,000)/(90 − 40) = 7,00,000/50 = 14,000 units.
4. Job shop vs transfer line: Q = 10,00,000/110 = 9091 units (not needed for the ranges, since CNC is cheaper than both between 5000 and 14,000).
5. Decision ranges: job shop below 5000 units, CNC cell from 5000 to 14,000 units, transfer line above 14,000 units.
6. Check at Q = 9000: job shop 2,00,000 + 150 × 9000 = ₹15,50,000; CNC 5,00,000 + 90 × 9000 = ₹13,10,000; line 12,00,000 + 40 × 9000 = ₹15,60,000.
7. Choose the CNC batch cell; minimum total cost = ₹13,10,000 per year.

Common mistakes

• Treating "mass" and "continuous" as the same: mass production makes discrete countable items on a line; continuous production processes a homogeneous material in uninterrupted flow.
• Forgetting the set-up time when computing the production rate of a batch process, which overstates capacity.
• Mixing units: cycle time in minutes with set-up time in hours.
• Comparing only two options in a three-way break-even problem; always check which process is cheapest in each range, because one pairwise break-even point may lie in a range where a third process is cheaper.
• Placing MPS or MRP at the strategic level. Aggregate planning and MPS work with given capacity; plant capacity and location are strategic.
• Thinking higher utilisation is always better: in job shops very high utilisation causes long queues and late orders.

For GATE PI

Expect conceptual questions matching production systems with their layout, volume–variety, skill level and examples, and ordering of planning activities (business plan → aggregate plan → MPS → MRP → scheduling). Numerical questions come through break-even comparison of processes, production rate with set-up time, capacity and utilisation. Practise three-option break-even charts and drawing the cost lines quickly to identify ranges.

Quick check

1. Which production system uses a product layout and dedicated machines for one standardised discrete product?
2. Arrange in order of decreasing planning horizon: MPS, aggregate plan, facility capacity decision, dispatching.
3. Set-up 1 h, batch of 100 units, cycle time 1.2 min/unit. What is the average production rate?
4. Process A: F = ₹1,00,000, v = ₹60; process B: F = ₹2,50,000, v = ₹30. What is the break-even volume?

Answers: 1. Mass (flow-line) production. 2. Facility capacity decision, aggregate plan, MPS, dispatching. 3. T_b = 1 + 100 × 0.02 = 3 h, R_p = 33.3 units/h. 4. 1,50,000/30 = 5000 units.