Tricky Industrial interview questions

75 questions with answers (IND-426 to IND-500). Try each one aloud before reading the answer.

Work Study & Ergonomics

IND-426 · "A worker faster than normal should be rated below 100%." True?

Trap: This reverses the scale. A faster-than-normal worker is rated above 100%, for example 120%. Normal time = observed time × rating, so the short observed time is scaled up to what a normal worker would take.

IND-427 · "Standard time is simply the average observed time." Correct?

Trap: This ignores rating and allowances. Observed time must first be normalized by the performance rating and then increased by PF&D allowances: ST = OT × rating × (1 + allowance).

IND-428 · "Ergonomics is just about comfortable chairs and posture." Is that the whole field?

Trap: That covers only physical ergonomics. The field also includes cognitive ergonomics (workload, displays, human error) and organizational ergonomics (shift design, teamwork). It also covers the environment: noise, heat and lighting.

IND-429 · "Design for the average person and it will fit most people." Right?

Trap: Very few people are average in all dimensions at once. A 50th-percentile reach is too long for half of users. Design for extremes (5th percentile for reach, 95th for clearance) or provide adjustability.

IND-430 · "Work sampling needs continuous stopwatch observation of the worker." True?

Trap: It confuses work sampling with time study. Work sampling uses many random, instantaneous snapshots, recording only the activity seen at each moment. Proportions of time are inferred statistically.

IND-431 · "15% of normal time and 15% of the shift are the same allowance." Are they?

Trap: The base differs. 15% of normal time gives ST = NT × 1.15. 15% of the shift gives ST = NT/0.85 = NT × 1.176, which is a larger standard.

Production Planning & Control

IND-432 · "Speeding up a non-bottleneck machine will raise plant output." Will it?

Trap: Throughput is set by the bottleneck. Speeding up a non-bottleneck only builds WIP in front of the constraint or creates idle time. Output rises only if the bottleneck's capacity rises.

IND-433 · "An hour lost at the bottleneck can be made up later by the other machines." True?

Trap: The other machines already have spare capacity; the bottleneck does not. An hour lost at the bottleneck is an hour of throughput lost for the whole system. An hour saved at a non-bottleneck is a mirage.

IND-434 · "Use MRP to plan finished-goods demand." Is that its purpose?

Trap: MRP calculates dependent demand for components from the MPS. Finished goods have independent demand, which must be forecast or taken from customer orders and entered through the MPS.

IND-435 · "Keeping every machine 100% busy maximizes plant efficiency." Does it?

Trap: Local efficiency is not system efficiency. Running non-bottlenecks flat out overproduces, inflates WIP, lengthens lead times and hides problems. Non-bottlenecks should run at the bottleneck's pace.

IND-436 · "SPT is the best dispatching rule for every objective." True?

Trap: SPT minimizes mean flow time and average WIP, but long jobs can wait indefinitely and miss due dates badly. EDD minimizes maximum lateness, and Moore–Hodgson minimizes the number of tardy jobs.

IND-437 · "Doubling the bottleneck's capacity doubles plant throughput." Correct?

Trap: Once the bottleneck is elevated, the next-slowest resource or market demand becomes the constraint. Throughput rises only up to the new constraint's capacity.

IND-438 · "A level production strategy avoids inventory cost." Does it?

Trap: The opposite is true. A level plan keeps output constant and uses inventory or backorders to absorb demand swings, so it carries inventory cost. It is the chase strategy that avoids inventory but pays for hiring, layoffs and overtime.

Inventory Management

IND-439 · "Zero inventory is always best." Is it?

Trap: Zero inventory ignores variability. Supply disruptions, demand uncertainty, setup costs and seasonality all justify some cycle stock and safety stock. The goal is the right inventory at minimum total cost, not none.

IND-440 · "If annual demand doubles, the EOQ doubles." True?

Trap: EOQ ∝ √D, so doubling D increases EOQ by √2, about 41%. Orders become more frequent, not just larger.

IND-441 · "If lead time doubles, safety stock doubles." Right?

Trap: With independent daily demand and fixed lead time, SS = zσ_d√L, so doubling L increases SS by √2, about 41%, not 100%. Lead-time variability, if present, adds a separate term.

IND-442 · "Ordering the EOQ minimizes purchase cost." Does it?

Trap: EOQ minimizes only ordering plus holding cost. Purchase cost D × C is constant in the basic model. With quantity discounts, the lowest total cost may be at a price-break quantity above EOQ.

IND-443 · "Going from 95% to 99.9% service level is a small change in safety stock." Is it?

Trap: Safety stock rises non-linearly. z goes from 1.645 to 3.09, which nearly doubles safety stock for a 4.9-point service gain. The last few percent of service are very expensive.

IND-444 · "A items in ABC analysis are the most expensive items." True?

Trap: The classification is by annual usage value (demand × unit cost), not unit price. A cheap item used in huge volumes can be an A item, and an expensive item rarely used can be a C item.

IND-445 · "Place a new order when on-hand stock falls to the reorder point." Correct?

Trap: Comparing on-hand stock ignores orders already placed. Compare inventory position (on-hand + on-order − backorders) with the ROP, otherwise duplicate orders are triggered when the lead time exceeds the order cycle.

Operations Research

IND-446 · "Round the LP solution to get the integer optimum." Is that valid?

Trap: Rounding can produce infeasible or clearly suboptimal points, because the integer optimum may be far from the LP optimum. Use branch and bound or cutting planes. The LP value only bounds the integer optimum.

IND-447 · "An LP always has a unique optimal solution." True?

Trap: If the objective is parallel to a binding constraint, alternative optima exist along an edge or face, all with the same Z. LPs can also be infeasible or unbounded. If a finite optimum exists, at least one vertex is optimal.

IND-448 · "Adding a constraint can improve the optimal objective." Can it?

Trap: Adding a constraint can only shrink or keep the feasible region, so the optimum stays the same or gets worse. Relaxing a constraint can never make it worse.

IND-449 · "A shadow price of $15 means each extra unit of resource is always worth $15." True?

Trap: The shadow price holds only within the RHS range of feasibility. Beyond the allowable increase the basis changes, and the shadow price usually falls, often to zero, as the constraint stops binding.

IND-450 · "The assignment problem is best solved by the general simplex method." Is it?

Trap: The assignment problem is highly degenerate in simplex, with n of the 2n − 1 basic variables at zero, so simplex makes many unproductive pivots. The Hungarian method exploits the structure and is far more efficient.

IND-451 · "The least-cost method gives the optimal transportation solution." Does it?

Trap: Least-cost, NWC and VAM give only initial basic feasible solutions. Optimality must be tested with MODI or stepping stone. VAM often comes close but is not guaranteed optimal.

Queueing Theory & Simulation

IND-452 · "100% server utilization is the ideal." Is it?

Trap: With any variability, waiting time grows as ρ/(1 − ρ). As ρ → 1, queue length and waiting time tend to infinity. Systems need spare capacity, typically running well below 100%.

IND-453 · "Doubling both λ and μ in an M/M/1 queue leaves waiting time unchanged." True?

Trap: ρ, L and Lq are unchanged, but W = 1/(μ − λ) halves because service is twice as fast. Little's Law confirms this: L is the same while λ doubles, so W halves.

IND-454 · "If the mean service time is less than the mean interarrival time, there will be no queue." Correct?

Trap: That holds only for perfectly regular arrivals and service. With random arrivals or service times, queues form whenever arrivals bunch, even at moderate utilization. Variability, not just the average load, creates waiting.

IND-455 · "Two servers with separate lines perform as well as one shared line." Do they?

Trap: With separate lines, one server can be idle while customers wait in the other line. A pooled single queue (M/M/2) gives much lower average waiting at the same utilization.

IND-456 · "One long simulation run gives the exact answer." Does it?

Trap: Simulation output is a random variable. You need independent replications or batch means, a warm-up deletion for steady state, and confidence intervals. A single run gives one sample, not the true value.

IND-457 · "Little's Law applies only to Poisson arrivals and exponential service." True?

Trap: L = λW holds for any stable system in steady state, whatever the arrival and service distributions, number of servers or queue discipline. It even applies to factories and supply chains (WIP = throughput × lead time).

IND-458 · "Raising utilization from 80% to 90% raises waiting by about 12%." Right?

Trap: Waiting scales with ρ/(1 − ρ), not with ρ. That factor goes from 4 to 9 in M/M/1, so the average queue more than doubles (Lq from 3.2 to 8.1).

Statistical Quality Control

IND-459 · "All points within control limits means the product meets specifications." True?

Trap: Control limits reflect process variation; specification limits reflect customer requirements. A process can be in control yet produce out-of-spec parts if Cpk < 1.

IND-460 · "Set the control limits at the specification limits." Should you?

Trap: Control limits on X̄ concern subgroup averages and are computed from process data (±3σ/√n). Placing them at the spec limits makes the chart blind to shifts and confuses individual parts with averages.

IND-461 · "Cp = 2 means the process is excellent." Always?

Trap: Cp ignores centring. A process with Cp = 2 whose mean sits near a specification limit can have Cpk ≈ 0 and make many defects. Always check Cpk.

IND-462 · "100% inspection guarantees zero defects reach the customer." Does it?

Trap: Human inspection is not perfectly effective, because fatigue, monotony and ambiguity let defects escape. Quality must be built in through capable processes and poka-yoke, not inspected in.

IND-463 · "Narrowing control limits to ±2σ catches more problems, so it is better." True?

Trap: The false-alarm rate rises from about 0.27% to about 4.55% per point, so ARL₀ falls from about 370 to about 22. This causes over-adjustment (tampering), which increases variation. Use run rules, EWMA or CUSUM for sensitivity instead.

IND-464 · "A lot ten times larger needs a sample ten times larger for the same protection." Correct?

Trap: When the lot is large relative to the sample, the OC curve depends mainly on n and c, not on lot size. Protection is set by the sample size, not by the sampling fraction.

Six Sigma

IND-465 · "Six Sigma means 99.9999% quality." Is that right?

Trap: Six Sigma means 3.4 DPMO, about 99.99966% yield, and this already includes the conventional 1.5σ long-term mean shift. A centred ±6σ process would give about 0.002 ppm.

IND-466 · "A 3σ process produces about 2,700 defects per million." True in Six Sigma terms?

Trap: 2,700 ppm (0.27%) is for a perfectly centred, short-term ±3σ process. Six Sigma tables apply the 1.5σ shift, so a 3σ process is quoted at about 66,807 DPMO.

IND-467 · "Counting more opportunities per unit shows the process improved." Does it?

Trap: Adding trivial opportunities inflates the denominator and lowers DPMO without any real change. Count only opportunities that are meaningful to the customer, and also track DPU or yield.

IND-468 · "A strong correlation found in Analyze proves the root cause." Does it?

Trap: Correlation is not causation. A lurking variable may drive both factors. Confirm causes with process knowledge, designed experiments or pilot changes before implementing solutions.

IND-469 · "p > 0.05, so the two machines are proven to perform the same." Correct?

Trap: Failing to reject H₀ only means there is insufficient evidence of a difference. The test may lack power because the sample is small. Equivalence must be shown with an equivalence test or a confidence interval within a practical margin.

IND-470 · "An FMEA item with RPN 120 is always more urgent than one with RPN 100." Is it?

Trap: RPN multiplies ordinal scales, so equal RPNs can mean very different risks. A failure with severity 9–10, which affects safety or regulatory compliance, needs action regardless of its RPN. Many practitioners prioritize by severity first, or use action-priority tables.

Lean Manufacturing & TPM

IND-471 · "Smaller batches reduce throughput because there are more setups." Do they?

Trap: Often the opposite. Smaller batches cut WIP and lead time and expose problems. As long as setups do not consume bottleneck capacity, throughput is unchanged or rises. SMED makes small batches economical.

IND-472 · "Lean is mainly a headcount reduction programme." Is it?

Trap: Lean targets waste in processes, not people. Using it for layoffs destroys the employee engagement that kaizen depends on. Freed capacity is normally redeployed to growth or improvement work.

IND-473 · "Kanban removes the need for forecasting." True?

Trap: Kanban controls short-term replenishment, but forecasts are still needed to size the number of kanbans, plan capacity and staffing, and order long-lead materials. Kanban counts must be recalculated when demand changes.

IND-474 · "OEE should always be pushed to 100%." Is that realistic?

Trap: 100% OEE means no stops, full ideal speed and zero defects. That is a theoretical limit, not a practical target. About 85% is often cited as a world-class benchmark for discrete manufacturing, and OEE is best used to find losses.

IND-475 · "5S is just housekeeping." Is it?

Trap: Cleaning (Shine) is one step. 5S creates visual standards so abnormalities, such as leaks, missing tools or excess stock, become obvious. Standardize and Sustain make it a discipline. It is the foundation for TPM, standard work and flow.

IND-476 · "In TPM, the maintenance department does all the maintenance." True?

Trap: TPM relies on autonomous maintenance by operators: cleaning, lubrication, inspection and minor adjustments. This frees maintenance specialists for planned maintenance, improvements and complex repairs.

Supply Chain & Logistics

IND-477 · "Choose the supplier with the lowest unit price to minimize cost." Correct?

Trap: Unit price ignores total cost of ownership: freight, duties, quality failures, inventory from long lead times, delivery unreliability, payment terms and risk. The cheapest quote can be the most expensive supplier overall.

IND-478 · "The bullwhip effect is caused by erratic customers." Is it?

Trap: Bullwhip appears even when end demand is stable. It comes from the supply chain's own structure: forecast updating on orders, order batching, price promotions, shortage gaming and long lead times.

IND-479 · "More warehouses always lower total logistics cost." True?

Trap: More warehouses reduce outbound transport cost and delivery time but increase facility cost and total safety stock, roughly by the square root of the number of sites. Total cost is U-shaped, so there is an optimal number of warehouses.

IND-480 · "Air freight always increases total cost compared with sea freight." Does it?

Trap: For high-value, perishable or volatile-demand goods, air's shorter transit cuts pipeline inventory, safety stock, obsolescence and lost sales. These savings can outweigh the higher freight rate.

IND-481 · "Sharing POS data eliminates the bullwhip effect." Does it?

Trap: Information sharing removes demand-signal distortion but not the effects of batching, lead time, promotions or rationing games. Bullwhip is reduced, not eliminated, unless those causes are also addressed.

IND-482 · "Cross-docking works for every product." True?

Trap: Cross-docking needs predictable, high-volume flows, tight inbound–outbound synchronization, reliable suppliers and good information systems. Slow, erratic items usually need conventional storage.

Facility Layout & Material Handling

IND-483 · "Adding more stations to a line always improves line efficiency." Does it?

Trap: At a given cycle time, extra stations add idle time and lower efficiency (Σt/(nC)). More stations help only if they allow a shorter cycle time to meet higher demand.

IND-484 · "Cycle time can be set lower than the longest task time." Can it?

Trap: In a simple line, the longest indivisible task sets the minimum cycle time. It can be beaten only by duplicating parallel stations for that task, splitting the task or speeding it up.

IND-485 · "A product layout is the most efficient for any factory." True?

Trap: A product layout is efficient only for high-volume, stable, standardized products. With high variety and low volume it becomes inflexible and underutilized. Process or cellular layouts suit those conditions better.

IND-486 · "Use straight-line distances to evaluate a factory layout." Appropriate?

Trap: Material moves along aisles at right angles, so rectilinear distance |x₁ − x₂| + |y₁ − y₂| is usually more realistic. Euclidean distance underestimates the actual travel.

IND-487 · "The centre-of-gravity method gives the minimum-transport-cost location." Does it?

Trap: The centre of gravity minimizes weighted squared Euclidean distance, not the weighted distance itself. The true minimum needs iterative methods (Weiszfeld) for Euclidean distance or weighted medians for rectilinear distance. It also ignores roads, land cost and other factors.

IND-488 · "Automating material handling always cuts cost." True?

Trap: Automation has high capital cost and needs stable, high-volume flows. Automating a poor layout or process locks in waste. First reduce handling by improving the layout and flow, then automate where volume justifies it.

Project Management

IND-489 · "The critical path is the one with the most activities." True?

Trap: The critical path is the longest path in total duration, with zero total float. A path with fewer but longer activities can be critical.

IND-490 · "Crashing any activity shortens the project." Does it?

Trap: Only critical activities affect the project duration. Crashing a non-critical activity just adds cost and float. Also, once paths become parallel-critical, they all must be crashed together.

IND-491 · "If a task is late, add more people and it will catch up." Will it?

Trap: Duration seldom falls in proportion to resources. Ramp-up, communication overhead and space limits give diminishing or even negative returns, as Brooks's law states for software. Crash cost slopes rise steeply.

IND-492 · "PERT's probability calculated from the critical path alone is accurate." Is it?

Trap: It ignores near-critical paths, which may finish later (merge bias), so it is optimistic. It also assumes independent activities and an approximately normal path sum. Monte Carlo simulation gives a more realistic estimate.

IND-493 · "Total float and free float are the same." True?

Trap: Total float is the delay possible without delaying the project. Free float is the delay possible without delaying any successor's early start. Free float ≤ total float, and total float is shared along a path.

IND-494 · "SPI = 1.0 at the end shows the project was on schedule." Does it?

Trap: At completion EV = PV = BAC, so SPI always returns to 1.0 even if the project finished late. Use earned schedule (SPI(t)) or actual dates to judge schedule performance near the end.

Engineering Economics & Forecasting

IND-495 · "For mutually exclusive projects, pick the one with the higher IRR." Correct?

Trap: IRR ignores scale and the timing of reinvestment, so it can rank projects in conflict with NPV. Use NPV at the MARR, or incremental IRR on the difference between the alternatives.

IND-496 · "Depreciation is a cash outflow." Is it?

Trap: Depreciation is a non-cash accounting expense. The cash outflow happens when the asset is bought. Depreciation affects cash flow only by reducing taxable income: tax shield = depreciation × tax rate.

IND-497 · "Money already spent on the old machine must be included in the replacement decision." True?

Trap: Sunk costs are irrelevant to future decisions. Compare only future cash flows: the current market value of the old asset (its opportunity cost) and the future costs of keeping versus replacing it.

IND-498 · "A low MAD proves the forecast is unbiased." Does it?

Trap: MAD measures error size, not direction. A forecast can have small errors that are always on the same side. Check bias with the mean error (RSFE) and the tracking signal.

IND-499 · "A shorter payback period means a more profitable project." True?

Trap: Payback ignores cash flows after the payback point and, in simple form, the time value of money. A slow-payback project can have a much higher NPV. Payback indicates liquidity and risk, not profitability.

IND-500 · "A higher α in exponential smoothing always gives better forecasts." Does it?

Trap: A high α reacts quickly to real changes but also chases random noise, which raises error for stable demand. Choose α by minimizing MAD or MSE on historical data. If demand has a trend or seasonality, use Holt or Winters instead.

Other levels: Basic · Intermediate · Advanced · All Industrial questions