HKIE Professional Assessment - Manufacturing, Industrial & Systems Discipline common mistakes
HKIE Professional Assessment - Manufacturing, Industrial & Systems Discipline · Understand the traps and how to avoid them.
In the essay-style questions on production planning, the candidate calculates capacity, utilisation and takt time correctly but never converts the result into a recommendation, so the answer ends with a number and no decision.
Candidates learn the formulas in isolation during university and treat the exam as a calculation test; under time pressure they stop as soon as the arithmetic is finished, not realising the assessors are marking engineering judgement and the ability to commit to a course of action.
What to do instead
Before writing, note the decision the question is really asking for (e.g. add a shift, re-sequence, outsource, hold). Structure every long answer as: given data and assumptions, calculation, interpretation, recommendation with the trade-off stated. Reserve the final two or three sentences explicitly for the recommendation and the condition under which it would change.
When asked to design or critique a quality system, the candidate lists tools (Pareto, fishbone, SPC, FMEA) as a catalogue without stating which is applied at which stage, what data feeds it, or what action follows an out-of-control signal.
Tool names are easy to memorise and feel like a safe way to fill space, so candidates substitute recall for the systems thinking the question is testing; they have often seen the tools in separate modules and never had to sequence them into a closed-loop control plan.
What to do instead
Answer in the language of the control loop: measure (which characteristic, which gauge, sampling frequency), analyse (chart type and why, control limits basis), react (who responds, at what rule violation, what containment and corrective action), verify. Name the tool only as the mechanism at that step, and always state the reaction plan for an out-of-control point.
In metrology and reliability questions, the candidate quotes a tolerance or a reliability figure without stating the measurement uncertainty, the confidence level, or the traceability chain behind the number.
Textbook problems present tolerances as exact and reliability as a single MTBF, so candidates carry that habit into an assessment context where the assessor is checking whether the candidate understands decision risk; the uncertainty budget and confidence basis are the parts most often skipped in revision.
What to do instead
Adopt a fixed habit: every measured or predicted value is stated with its uncertainty and basis (e.g. expanded uncertainty at 95% coverage, k=2; MTBF with assumed exponential distribution and stated confidence). When judging conformity to a tolerance, apply the guard-band or shared-risk logic explicitly and say which decision rule you are using.
On automation and robotics questions, the candidate proposes a robot or an automated line without justifying the payload, reach, cycle time, end-effector, safety guarding or changeover implications, treating automation as a generic good.
Automation is often taught as a technology survey rather than as an engineering justification exercise, so candidates answer with capability claims ('a six-axis robot can do this') instead of a feasibility and economics argument; they also underestimate the safety and integration content the exam expects.
What to do instead
Justify the selection against a short checklist: required cycle time versus robot speed and acceleration, payload including end-effector, reach and workspace, repeatability versus part tolerance, gripper and part-presentation strategy, safeguarding per the applicable machinery safety approach, and payback period. State at least one option considered and rejected with the reason.
In supply chain and logistics questions, the candidate computes an EOQ or safety stock and presents it as the answer, ignoring demand variability, lead-time variability, service-level targets, and the mismatch between the model's assumptions and the actual operation.
Inventory models are examined as mathematics in coursework, and the simplifying assumptions (known constant demand, no quantity discounts, no capacity limits) are stated once and forgotten; candidates then apply the formula mechanically to a scenario the examiner has deliberately made non-ideal.
What to do instead
Always state the assumptions you are invoking and then test them against the scenario. Show safety stock as a function of both demand and lead-time variability at the stated service level, comment on lumpy or seasonal demand, and finish with the policy implication (review period, order-up-to level, supplier collaboration) rather than the raw number alone.
On engineering economics and project appraisal, the candidate discounts cash flows but mixes real and nominal rates, ignores inflation, tax, depreciation or working capital, and reports a single IRR or payback without a sensitivity check.
Examination questions are usually simplified to clean cash flows, so candidates never build the discipline of matching the discount rate to the cash-flow basis; under exam pressure the shortest calculation path is chosen and the qualitative risk discussion is dropped.
What to do instead
Write the basis explicitly at the top: currency, real or nominal, discount rate source, study period, salvage and tax treatment. Then present NPV as the primary criterion with IRR and payback as supporting, and add a short sensitivity or break-even statement (e.g. the capital cost increase or volume drop that would reverse the decision).
On risk and project management questions, the candidate produces a risk register of generic categories ('schedule risk, cost risk, safety risk') with no owner, no likelihood-consequence rating, no mitigation and no link to the project stage or contract strategy.
Risk is often revised as a list of headings rather than as a managed process, and candidates are reluctant to commit to specific numbers or named owners in a written answer; generic entries feel safer but score poorly because they demonstrate no engineering judgement.
What to do instead
Use a compact structure per risk: cause, event, consequence, likelihood and severity with the scale stated, owner, response type (avoid, reduce, transfer, accept), residual rating, and trigger for escalation. Anchor at least one risk to the specific project stage or contractual interface in the scenario, and show how the register feeds the contingency or float.
On safety, health, environment and sustainability questions, the candidate answers purely with technical controls and omits the statutory duty, the hierarchy of control, and the Hong Kong legal framework (e.g. the Factories and Industrial Undertakings Ordinance and its subsidiary regulations, or the relevant ordinance for the hazard in question).
Candidates prepare the engineering content thoroughly but treat law and management systems as peripheral, and they default to the most familiar control (usually PPE or a guard) without working down the hierarchy; the Hong Kong statutory layer is often studied last and thinly.
What to do instead
Answer in the order: legal duty and applicable Hong Kong legislation, hierarchy of control from elimination downwards with the chosen controls placed on it, then monitoring, training and review. Name the specific regulation or approved code of practice where you are confident; where you are not certain of the exact instrument, describe the duty accurately rather than inventing a citation.
On professional practice and ethics questions, the candidate describes what they personally would do without identifying the duty owed, the conflicting obligations, the applicable code or the correct escalation route.
Ethics scenarios invite a personal-voice answer, and candidates answer as an employee rather than as a professional accountable to a code and to the public; they also fear that escalating looks disloyal, so they under-report the formal steps.
What to do instead
Frame the answer around duties: to the public and safety, to the employer, to the profession and its code, and to the law. State the facts you would establish, the conflict, the options ranked by risk, the one you would take and why, the record you would keep, and the point at which you escalate to a senior engineer, the client or a regulator. Commit to a position rather than listing options neutrally.
Across the system integration and standards questions, the candidate writes long descriptive prose that never defines requirements, interfaces, verification methods or acceptance criteria, so the assessor cannot see whether the candidate can actually specify a system.
Descriptive writing is the default from academic reports and fills pages quickly, whereas structured specification writing is a distinct skill rarely practised; candidates also confuse describing a system with specifying one, and omit verification because it feels like an administrative afterthought.
What to do instead
Convert description into requirements: each stated need becomes a numbered requirement with a measurable threshold, a verification method (test, inspection, analysis, demonstration) and an owner. For interfaces, state the physical, energy and information flows and who is responsible on each side. Close with acceptance criteria and the standard or code invoked, so the answer reads as a specification rather than an essay.