Approach the PE Industrial and Systems exam by separating model selection from calculation. For every practice item, name the decision type (design, evaluation, control, or comparison), list the candidate methods, check their assumptions against the given data, and only then compute. Log your model choice before solving so review sessions expose selection errors, not just arithmetic slips.
Takt Time, Cycle Time, and Lead Time Answer Different Questions
Takt time is customer demand rhythm; cycle time is the output rhythm of a station or process; lead time is elapsed time from order to delivery. Because data for all three can appear in one scenario, practice naming which one the question asks for before choosing a formula.
Worked scenario: a line operates 450 productive minutes per day and must deliver 90 units. Takt time is 450/90 = 5 minutes per unit. A station with a 4.6-minute cycle time meets demand because its cycle is below takt. A plausible mistake is treating 4.6 minutes as a problem because it differs from takt, then rebalancing unnecessarily. The better decision: verify cycle time is less than or equal to takt and leave the line alone; rebalancing only if a station exceeds takt.
The reverse situation deserves equal attention when you drill: if takt computes to 4 minutes and the slowest station runs 4.6 minutes, no amount of work-element shuffling at that station fixes demand; you need parallel stations, overtime, or a demand-side answer. Train yourself to state the comparison explicitly — cycle versus takt, in the same units, after subtracting planned downtime and breaks — before touching the precedence diagram.
- Takt time = net available production time divided by customer demand for that period; it has no dependence on how the line is designed.
- Station cycle time is the longest work-element time at a station after grouping; line cycle time is the maximum station time.
- Lead time includes queue, move, and wait, so it is almost never equal to the sum of processing times alone.
Layout Scenarios: Weight Flows by Distance, Not Just Counting Moves
Layout scenarios that ask to minimize material handling are naturally solved with the load-distance method: multiply each department pair's flow by the distance between candidate locations and sum. A plausible mistake is ranking layouts by the number of close pairs instead of the weighted score, which can invert the correct answer.
Worked scenario: two layouts place departments A and B either adjacent or across the plant. Layout 1 separates two pairs with heavy flow (50 loads, 40 loads) across a long rectilinear distance of 60 m, while keeping low-flow pairs close. Layout 2 separates one pair with flow of 10 over 60 m but puts the heavy flows adjacent. Counting close pairs favors Layout 1; the load-distance score favors Layout 2 because the 50-load pair now travels a short distance. Computing the weighted total resolves what pair-counting cannot.
When a scenario asks to minimize total material handling, build a from-to chart, confirm the distance metric (rectilinear versus Euclidean changes every number), and calculate the load-distance score for each candidate. If the item instead asks about fixed costs, process flexibility, or future expansion, the layout metric alone is not the decision criterion — read what the question rewards. Rehearse converting a trip-frequency table into a weighted score quickly, since fast conversion keeps layout items moving while you think about the decision.
Little's Law and Queueing: Define the System Boundary Before Substituting
Little's Law states average inventory (or WIP) equals throughput times flow time, valid for long-run averages in steady state. Queueing items add arrival and service distributions. The selection error is mixing WIP measured inside a boundary with flow time measured across a larger one.
Worked scenario: a shop reports 200 units of WIP on the floor and a throughput of 40 units per day. Little's Law gives a flow time of 5 days. If the item asks for time in queue only and also states that each unit spends 3.5 days in processing, queue time is 1.5 days. A plausible mistake is reporting 5 days as queue time because the formula was applied without separating value-added processing from waiting. The fix is to define which population the 200 units represent before multiplying or dividing anything.
For queueing questions, the given discipline matters less than the parameters: arrival rate, service rate, and how many servers are available. Check units first — arrivals per hour with service times in minutes is a unit mismatch that quietly inverts results. Also confirm steady-state language; if the scenario describes a system ramping up or draining down, average-based formulas describe a condition the scenario has not reached, and the defensible answer focuses on what the averages would be once steady state holds.
EOQ, Reorder Point, and Safety Stock Solve Three Different Problems
EOQ answers how much to order under constant demand and no shortages; the reorder point answers when to order; safety stock answers how much buffer is needed for a stated service level under variable demand. If a scenario varies demand or lead time, the data is pointing past the basic EOQ.
Worked scenario: annual demand is 12,000 units, ordering cost is 75 per order, holding cost is 2 per unit per year. EOQ is the square root of (2 × 12,000 × 75 / 2) = 949, about 949 units per order, ordered roughly 12.6 times a year. The plausible mistake is stopping there when the item also gives a demand standard deviation and asks for a 95 percent service level. That second part requires safety stock: z times the standard deviation of demand over lead time, which changes the reorder point but not the EOQ.
The distinction to internalize: the EOQ square-root formula balances ordering and holding costs only, and its assumptions — known constant demand, instantaneous or fixed replenishment — are exactly the assumptions that variable-demand scenarios violate. When the question mentions service level, stockout risk, or demand variability, treat EOQ as at most one component of the answer and reach for the distribution-free or normal-demand safety stock calculation the data supports. Writing the three question types in your practice log makes the cue visible.
Control Chart Selection: Match the Chart to the Data Type and Sample Size
Variables data (measurements) go to X-bar and R or X-bar and s charts; attributes data go to p, np, c, or u charts depending on whether you track defectives or defects and whether the sample size is constant. Misclassifying the data type sends you to the wrong chart every time.
Worked scenario: an inspector records the fraction of defective assemblies in samples of varying size each shift. The plausible mistake is computing control limits from an average sample size as if it were constant. The better decision: a p-chart with limits recalculated per sample (or clear justification for an average-size approximation), because the fraction defective is attribute data and the sample size varies. If instead the data were defect counts on units of differing opportunity, the u-chart is the match; constant sample size shifts the choice to np or c.
Worked scenario two: a machinist measures shaft diameters in subgroups of five. That is variables data, so an X-bar and R chart applies; a p-chart would discard the measurement information entirely. Train the decision order — data type first, defectives versus defects second, constant versus varying sample size third — as a verbal checklist. The table below compresses it. Alongside limit computation, rehearse naming the rule that a described pattern triggers (a run, a trend, a point beyond limits), so pattern interpretation becomes as automatic as the chart choice itself.
| Data you are given | Chart | Defining condition |
|---|---|---|
| Subgroup measurements (e.g., diameters) | X-bar and R (or s) | Variables data in rational subgroups |
| Fraction defective per sample | p-chart | Attributes; sample size may vary |
| Count defective per sample | np-chart | Attributes; constant sample size |
| Count of defects per unit | c-chart | Attributes; constant opportunity |
| Defects per unit, varying opportunity | u-chart | Attributes; sample size or area varies |
Mutually Exclusive Alternatives: Incremental Analysis Beats Simple Ranking
When alternatives are mutually exclusive and differ in cost or capacity, comparing internal rates of return can rank them incorrectly. Incremental analysis on the difference between alternatives, or equivalent annual worth for unequal lives, gives the defensible choice.
Worked scenario: Machine A costs 20,000 with annual savings of 6,000; Machine B costs 30,000 with annual savings of 8,500, both over five years. Ranking by IRR favors A, because the smaller investment earns a higher percentage. But if both clear the minimum attractive rate of return, the incremental investment of 10,000 generates an extra 2,500 per year, and whether that increment clears the MARR decides between them. The plausible mistake is choosing by the higher percentage return alone; the better decision runs the increment explicitly.
Unequal lives raise a second trap. A five-year and an eight-year alternative cannot be compared by raw present worth without a common study period or replacement assumption. Equivalent annual worth sidesteps this by putting both on a per-year basis, which is why it is the natural method whenever lives differ and the scenario permits repeated replacement. Rehearse all three moves — ranking check, increment computation, and annualization — and note in the margin which one the question's phrasing (least cost, best return, unequal lives) is pointing toward.
A Model-Selection Training Sequence and Readiness Checks
Structure preparation in five phases: map topics to your reference resources, drill single concepts, run mixed sets with a model-choice log, add timed conditions, and close with log-driven review. The log is the centerpiece because it separates selection errors from calculation errors.
Practical exercise: take ten mixed practice items from the free set. For each, write down before solving (1) the decision being asked, (2) the method you will use, and (3) one assumption that must hold. Score yourself with this rubric: three points if the named method was correct, two if the stated assumption matched the method's actual requirement, one if units were consistent throughout. A total of 24 or more out of 30 is a reasonable learning milestone for moving into timed sets; treat it as a self-check, not a prediction of exam performance.
Adaptable sequence: Phase one, list each syllabus topic and locate the supporting tables or formulas in your approved reference materials. Phase two, drill one topic at a time until the standard method and its assumptions are automatic. Phase three, mixed sets with the model-choice log above. Phase four, timed mixed sets where finding the reference entry is part of the clock. Phase five, reread your log and rework only the items where the method — not the arithmetic — was wrong. Readiness checks: you can state takt versus cycle time without notes, select a control chart from a data description, run an incremental analysis unprompted, and locate any needed table in your references quickly.
- Milestone rubric scores are self-check learning targets, not passing indicators.
- Keep a one-line reason next to every logged method choice; vague reasons mark topics to revisit.
- Rehearse calculator fluency on an approved model; NCEES restricts acceptable calculators and lists them on its site.
References and further reading
Use these references to explore the concepts and check the latest information from the relevant organizations.