Operations Management and Production Strategy
What this domain tests
You are tested on the arithmetic of flow, capacity and quality: how work-in-process relates to throughput and time, where capacity is actually lost, how much to order, and whether a process is capable of its specification.
Core rules
- Little's Law. WIP = throughput × flow time. Fix two variables and the third follows; cutting flow time at constant throughput requires cutting WIP.
- Bottleneck and TOC. Five focusing steps: identify, exploit, subordinate, elevate, repeat. Utilising a non-bottleneck more only builds inventory.
- EOQ. Q* = √(2DS/H). It assumes stable demand, no quantity discounts and no setup reduction.
- JIT/kanban. Requires levelled demand, reliable quality and fast setup; kanban without those preconditions just relocates inventory.
- Process capability. Cp = (USL − LSL)/6σ measures spread only; Cpk = min[(USL − μ)/3σ, (μ − LSL)/3σ] also penalises off-centring. Cpk ≥ 1.33 is the usual industrial target.
- Product-process matrix. Job shop, batch, line, continuous - misalignment between volume/variety and process type destroys cost position.
Worked example
A machined part is specified at 10 ± 2 mm. The process mean is 10.6 mm with σ = 0.6. Cpk = min[(12 − 10.6)/(3 × 0.6), (10.6 − 8)/(3 × 0.6)] = min(0.78, 1.44) = 0.78. The process is not capable, and the low side is not the problem - the mean is shifted. Centring the mean at 10.0 gives Cpk = 1.11, still below 1.33, so spread must also be reduced (to about σ = 0.50). Two distinct actions: adjust the setting, then reduce variation.
Traps that cost marks
- Reporting Cp when the question asks about an off-centre process.
- Adding capacity at a non-bottleneck.
- Applying EOQ when a quantity-discount schedule is given.
- Assuming 100% utilisation at the bottleneck is desirable - it lengthens lead time.