Case-Based Questions
The skill being tested here is not "do you know the formula" — it's "can you recognise which OM concept a real, messy scenario is actually asking about." Each case walks through that reasoning process explicitly.
Case 1 — How Akshaya Patra Feeds 2.1 Million Children a Day From your course PPT (Unit 1)
The Akshaya Patra Foundation started in 2000 serving 1,500 children across 5 Bangalore schools under India's mid-day meal scheme. It has since grown to serve 2.1 million children daily by 2024, using large centralised kitchens with standardised cooking protocols and automation, distributing meals via a fleet of delivery vehicles to partner government schools. It now targets 3 million children by 2025 and is exploring a hub-and-spoke model — centralised kitchens supplying semi-cooked meals to smaller "spoke" kitchens nearer to new schools — to keep scaling.
What is the problem?
Akshaya Patra must keep increasing the number of children fed without a proportional increase in cost, kitchen space, or delivery time — classic operations questions of productivity and scalable process design, not a funding or menu-design problem.
Which concept applies?
Two concepts from Unit 1 & Unit 6: (1) Productivity — Akshaya Patra needs to keep raising output (meals served) without a proportional rise in input (kitchen labour, fuel, capital); standardised protocols and automation are exactly how it increases multifactor productivity. (2) The hub-and-spoke proposal is itself a production-system/process-choice decision (Unit 6): centralising the most standardised, highest-volume part of the work (bulk cooking) while pushing the more localised, lower-volume finishing step (final prep/delivery) to smaller spoke kitchens closer to demand.
How should I think about it?
Start from the numbers: input (kitchens, vehicles, labour) has grown much more slowly than output (meals/day), which is the definition of rising productivity — but growth targets (3 million by 2025) mean this ratio must keep improving, not just hold steady. Then ask what process choice supports further scaling without a proportional cost increase: a purely centralised model eventually hits transportation-time and kitchen-capacity limits, which is exactly why the hub-and-spoke model — separating a highly standardised bulk-cooking stage from a more flexible, localised finishing stage — is the natural next step.
Akshaya Patra's growth from 1,500 to 2.1 million children served daily illustrates a sustained rise in productivity: it achieves far greater output (meals served) without a proportional rise in input (kitchens, labour, vehicles), largely through standardised cooking protocols and automation in centralised kitchens — a direct application of Output/Input productivity improvement through process standardisation. As it targets further growth to 3 million children, however, a purely centralised model faces limits in kitchen capacity and delivery time to increasingly dispersed schools. Its proposed hub-and-spoke model addresses this using process-choice logic from the product-process matrix: the most volume-intensive, standardisable stage (bulk cooking) stays centralised for efficiency, while a more flexible, localised stage (finishing and distribution from spoke kitchens) is pushed closer to demand, allowing the organisation to keep growing without proportionally increasing central kitchen capacity or delivery distance.
Case 2 — Amazon India: Location Analysis for Fulfilment Centres From your course case study (Unit 4)
Amazon India must decide where to place fulfilment centres, sortation facilities, delivery stations, and transportation hubs across a geographically and economically diverse country. Key factors identified in the case include customer proximity, market demand, population density, transportation infrastructure, proximity to suppliers, labour availability, land/facility cost, industrial infrastructure, technology/digital infrastructure, government policy, risk/resilience, and last-mile delivery considerations. A purely centralised, single-warehouse model is ruled out because India's differences in road infrastructure, population density, and regional demand make it operationally infeasible.
Case question: Propose a location-selection model using the factor rating method, the centre-of-gravity method, and the load-distance method.
What is the problem?
This is a multi-layered network location decision, not a single-site problem — Amazon needs different facility types (fulfilment centres, sortation facilities, delivery stations, transportation hubs) positioned according to different priority factors, and it needs a systematic way to compare and select among candidate sites for each.
Which concept applies?
All three Unit 4 location-analysis techniques apply, but at different stages of the decision:
- Factor rating — to shortlist candidate regions/cities by scoring them on weighted factors like labour availability, infrastructure quality, government incentives, and risk, which are not purely distance-based.
- Centre-of-gravity — to identify an ideal coordinate for a new regional fulfilment centre relative to existing supplier hubs and major demand centres (weighted by order volume), as a starting reference point.
- Load-distance — once a shortlist of real candidate plots/industrial zones is identified near that ideal coordinate, to rank them by weighted shipment-distance and choose the one with the lowest total logistics burden.
How should I think about it?
Work in the same sequence the case itself implies: first narrow down broad regions using factor rating (since factors like regulation, labour, and infrastructure aren't reducible to coordinates), then use centre-of-gravity to mathematically pinpoint where — relative to supplier and demand locations — a facility would minimise weighted transportation distance, and finally use load-distance to choose between the small number of real, available sites near that computed point. This layered use of all three techniques mirrors exactly how the case frames Amazon's actual decision as "a network-level decision rather than a simple distance calculation."
Amazon India's facility-location decisions are best modelled as a three-stage process combining all three location-analysis techniques. First, the factor rating method should be used to shortlist broad candidate regions or cities, scoring each on weighted factors identified in the case — customer proximity, market demand, labour availability, transportation infrastructure, land cost, government policy, and risk — since these qualitative and quantitative factors cannot be reduced to distance alone. Second, within a shortlisted region, the centre-of-gravity technique should be applied using the coordinates and shipment volumes of existing suppliers and major demand centres to compute an ideal coordinate that minimises weighted transportation distance for a new fulfilment centre — serving as a mathematical starting reference rather than a final answer. Third, once a small number of real, developable sites near that ideal coordinate are identified, the load-distance technique should be used to rank them by computing each site's total load-distance value against existing suppliers and demand points, selecting the site with the lowest value as it implies the lowest transportation cost. Applying all three methods in sequence reflects the case's own conclusion that Amazon's location strategy is a network-level decision balancing cost, accessibility, speed, and customer service — not a single distance calculation.
Case 3 — The Slow Print Shop Illustrative case
A print shop processes customer orders through three stages: design layout (capacity: 30 jobs/hour), printing (capacity: 18 jobs/hour), and binding/packaging (capacity: 25 jobs/hour). Customers complain that although the design stage feels fast and the staff there always seem idle waiting for the next job, overall order turnaround has gotten worse as order volume has grown. The manager considers hiring another binding worker to speed things up.
What is the problem?
Overall turnaround time is limited by something in the three-stage process, and the manager is about to invest in the wrong fix.
Which concept applies?
Bottleneck analysis (Unit 3). Comparing capacities in the same unit: Design = 30/hr, Printing = 18/hr, Binding = 25/hr. Printing has the lowest capacity — it is the bottleneck, and it sets the capacity of the entire process at 18 jobs/hour, no matter how fast design or binding could go individually.
How should I think about it?
The design stage's staff being idle is not "wasted efficiency" to fix — it's a symptom of design being a non-bottleneck stage that has been starved because printing can't keep up (Unit 3's starving concept, applied downstream of design). Adding a binding worker would not help at all, since binding (25/hr) is already faster than the true bottleneck (printing, 18/hr); the manager should instead add capacity at the printing stage.
The print shop's overall capacity is limited by its bottleneck, which is the stage with the lowest processing capacity. Comparing all three stages on a common units/hour basis — design (30/hour), printing (18/hour), and binding (25/hour) — printing is the bottleneck, meaning the entire process can never process more than 18 jobs per hour regardless of how fast the other two stages run. The design stage appearing idle is consistent with it being starved: since it can produce work faster than the downstream printing stage can absorb it, work backs up before printing and design periodically runs out of new work to start. Hiring an additional binding worker would not improve overall turnaround at all, since binding's capacity (25/hour) already exceeds the bottleneck's capacity; the only way to increase the shop's true throughput is to add capacity at the printing stage itself.
Case 4 — Choosing a Layout for a Growing Bakery Illustrative case
A bakery currently makes a wide variety of custom-order cakes in small batches, with workstations for mixing, baking, and decorating arranged in separate areas that a cake order visits in different combinations depending on the design. The owner is now considering also launching one single, highly standardised "classic sponge cake" product to sell in bulk to a supermarket chain, expecting very high, stable weekly volume.
What is the problem?
The bakery's existing layout and process were designed for its custom-cake business; the question is whether that same setup is right for the new high-volume, standardised product too.
Which concept applies?
The product-process matrix (Unit 6) and layout choice (Unit 5) together. The custom-cake business sits at the low-volume/low-standardisation end of the matrix (batch production, process layout — appropriate given high variety). The new supermarket product sits at the high-volume/high-standardisation end (mass production), which calls for a product layout, not the existing process layout.
How should I think about it?
Running the high-volume, standardised sponge cake through the same process-layout setup used for custom cakes would put it off the diagonal of the product-process matrix — a mismatch that typically shows up as unnecessarily high cost and inefficiency for a product that doesn't need the process layout's flexibility. The bakery should set up a separate, dedicated product-layout (assembly-line-style) line for the standardised cake, sequenced and balanced (Unit 5's line-balancing) for its specific steps, while keeping the flexible process layout for the custom-order business.
The bakery's existing process layout — workstations for mixing, baking, and decorating arranged separately and visited in varying sequences — is well matched to its custom-cake business, which involves high variety and low, fluctuating volume, corresponding to batch production on the product-process matrix. The new standardised sponge cake for supermarket supply, however, involves high, stable volume and a single fixed recipe, corresponding to mass production, for which a product (assembly-line) layout is the appropriate match. Continuing to run the standardised product through the existing process layout would position it off the diagonal of the product-process matrix, resulting in unnecessary inefficiency for a product that does not need process-layout flexibility. The bakery should therefore establish a separate, line-balanced product layout dedicated to the standardised cake, while retaining its current process layout for the custom-order business, rather than trying to serve both product types through a single layout.
Case 5 — The Hardware Store's Inventory Headache Illustrative case
A hardware store stocks thousands of items, from expensive power tools (a few units sold per month, high unit cost) to cheap items like nails and screws (sold in huge quantities, very low unit cost). The owner currently checks stock of every item personally, every single day, and complains that this takes far too much time yet stockouts of popular tools still happen.
What is the problem?
The owner is applying the same tight, time-consuming control to every item regardless of its actual value to the business, while the items that matter most (expensive, higher-impact tools) still run out.
Which concept applies?
ABC classification / selective inventory control (Unit 7). The store's items should be split by annual dollar value (unit cost × usage) into A, B, and C classes, and control effort reallocated accordingly — potentially combined with a continuous (Q) system with a proper reorder point for the A items (Units 8–9), rather than an undifferentiated daily manual check for everything.
How should I think about it?
The expensive power tools likely represent a small fraction of units but a large fraction of total inventory value (Class A) and deserve tight, formal control — an EOQ-based order quantity and a calculated reorder point with safety stock. The nails and screws (Class C) are individually low-value; even holding a comparatively generous safety stock of them is cheap, so they don't need daily personal checks at all — periodic, infrequent review is enough. The owner's mistake is spending equal daily attention on both.
The hardware store's problem is that it applies uniform, high-effort control to all inventory items regardless of their actual value to the business, which is both inefficient and ineffective, since it still experiences stockouts of its most important items. Applying ABC classification, high-value, lower-volume items such as power tools would likely fall into Class A, representing a small percentage of items but a large percentage of total inventory value, and therefore warrant tight control — an economic order quantity with a properly calculated reorder point and safety stock. Low-cost, high-volume items such as nails and screws would fall into Class C, representing most of the item count but very little total value; since the cost of holding extra stock of these items is minimal, they can be managed with infrequent, periodic review rather than daily monitoring. Reallocating control effort in this way — tight, formal control on Class A items and minimal oversight on Class C items — would reduce the owner's daily workload while actually reducing stockouts of the items that matter most.
Case 6 — Which Job First? Illustrative case
A small tailoring shop has five customer orders waiting, each with a different promised delivery date and a different amount of remaining work. The tailor currently just works on whichever order arrived first, and has recently missed several delivery promises even though the shop isn't unusually busy. A regular customer suggests "just do the quickest ones first," while another suggests "do whichever is due soonest."
What is the problem?
The tailor's current sequencing rule (first-come, first-served) is producing poor due-date performance, and needs to choose a rule that will actually improve the outcome that matters — missed deliveries.
Which concept applies?
Sequencing rules (Unit 10) — specifically, the choice between SPT (shortest processing time, "quickest first") and DDATE (earliest due date, "soonest due first"), and the guideline for when each is appropriate.
How should I think about it?
Since the tailor's actual complaint is missed due dates (tardiness), not slow average turnaround, the textbook's own guideline applies directly: DDATE minimises mean and maximum tardiness, while SPT minimises mean flow time/number of jobs in the system but can let a few due-date-sensitive jobs run very late. Given the shop isn't in fact overloaded (FCFS's typical justification), and the presenting problem is specifically missed promises, DDATE is the better-matched rule here — not SPT, despite sounding intuitively efficient.
The tailoring shop's problem is poor due-date performance under a first-come-first-served sequencing rule, even though the shop is not operating at unusually high capacity. Since the specific complaint is missed delivery promises (tardiness) rather than generally slow turnaround, the earliest-due-date (DDATE) rule is the better-matched choice: DDATE is specifically guided to minimise mean and maximum tardiness by prioritising jobs closest to their due date. The suggestion to sequence by shortest processing time (SPT) would instead minimise average flow time and work-in-process, but can cause a small number of jobs with tight due dates to become significantly late, which is precisely the symptom already being reported. Because FCFS is only recommended when a shop is operating at low capacity — which does not appear to be the binding constraint here — switching specifically to DDATE, rather than SPT or continuing with FCFS, directly targets the shop's actual problem.