Lean Certification Bronze Silver Gold Study Guide · Industrial/Systems · 10 min read

Lean Certification: Bronze / Silver / Gold — Advanced Study Guide

Lean Certification: Bronze / Silver / Gold — Advanced Study Guide

Issuing body: SME (Society of Manufacturing Engineers) in alliance with AME (Association for Manufacturing Excellence), the Shingo Institute, and ASQ Jurisdiction: Global (US-origin) Level: Bronze — tactical / individual contributor; Silver — integrative / value-stream leader; Gold — strategic / enterprise transformation leader Prerequisites: Bronze — lean fundamentals (no formal prerequisite); Silver — current Bronze certification (and documented leading of a full value-stream transformation, 5 lean events, and mentoring another practitioner); Gold — current Silver certification (and documented enterprise-level transformation leadership across demand, delivery, development, and support value streams) Renewal: 3-year certification cycles; renewal via continuing education and portfolio updates

1. How the credential is earned

This is the only lean credential jointly issued by four major bodies, and it is the only one that combines an exam with a judged portfolio. That matters: passing the exam alone does not make you Bronze-certified in the full sense — the certification is exam plus portfolio.

Step-by-step pathway

  1. Read the recommended reading. The exam is written directly from a published book list (Toyota Production System texts, Lean Thinking, Learning to See, Shingo model materials). Buy the book package, read it, tab it — the exam is open-book and open-notes (per SME exam-preparation guidance), so your library is your allowed reference.
  2. (Optional) Take a review course. A 2-day Bronze review is commonly offered through AME/GBMP-type providers; cost on the order of $500 (varies). Not required.
  3. Sit the exam. Bronze: 3 hours, 170 multiple-choice questions, open book. Silver: 3 hours, 150 questions, open book. Gold: 3 hours, 150 questions, open book. (Question counts from provider-documented program materials; confirm current counts on sme.org before booking.)
  4. Submit the portfolio. Bronze: document tactical projects/events where you applied lean tools, with results, plus ~80 hours of documented training/education and reflection. Silver: document leading a lean transformation of a complete value stream plus 5 lean events and proof of mentoring another practitioner. Gold: document leadership of an ongoing enterprise transformation across all four value streams (Demand, Delivery, Development, Support) plus development of a training strategy.
  5. Gold only: panel interview. A peer panel interviews you on your accomplishment record.
  6. Maintain. 3-year cycle; renewal through continuing education and portfolio updates.

Exam format and strategy

  • 170 questions in 180 minutes ≈ 63 seconds per question. You cannot look up everything — the exam is open-book but time-boxed.
  • Question types (per SME): majority knowledge (definitions: "the first S of 5S is…"), some application (calculate takt time from given data), a few judgment (best choice among plausible countermeasures).
  • Strategy: tab your books by BOK section; memorize the formulas (takt, kanban count, Little's law, OEE, line-balance efficiency) so you never burn lookup time on arithmetic; budget one pass with no revisits on judgment questions — your first lean instinct is usually the exam's intended answer.

2. Knowledge domains

The certification is built on the SME/AME/Shingo Lean Certification Competency and Behavior Model, organized around the Shingo model's guiding principles: Cultural Enablers, Continuous Improvement, Enterprise Alignment, Results. The BOK then drills into the tool set. Below is the theory that actually carries points.

2.1 Value, waste, and value stream mapping

Value is defined by the customer: any activity the customer would willingly pay for. Everything else is waste (muda), and waste divides into Type I (necessary but non-value-added, e.g., regulatory inspection — minimize) and Type II (pure waste — eliminate). The classic wastes: TIMWOOD + unused talent (Transportation, Inventory, Motion, Waiting, Overproduction, Over-processing, Defects, underused people).

Value stream mapping is the diagnostic. Draw the current state before designing the future state — the number one trap on the exam is proposing countermeasures before the current state is quantified. Lead time = sum of process times + all waiting/inventory time. The ratio (value-added time / total lead time) for a typical operation is 1–5%; the exam loves asking you to compute it.

Worked sanity check: A line has 5 operations at 45 s each, with 3 days of WIP between each operation (2 shifts/day, 8 h/shift). Value-added time = 225 s ≈ 3.75 min. Total lead time ≈ 4 × 3 days = 12 working days ≈ 192 h. VA ratio = 0.0625/192 ≈ 0.03%. The point: inventory, not process time, is the lead-time problem. That insight is half of lean.

2.2 Takt time, flow, and pull

Takt time = available production time / customer demand rate. It is a requirement, not a target — the pace the process must hold to satisfy demand.

Derivation example: 2 shifts × 8 h = 16 h; subtract breaks/lunch (2 × 45 min = 1.5 h) → 14.5 h available = 52,200 s. Demand = 580 units/day. Takt = 52,200 / 580 = 90 s/unit. Every operator's work content must be ≤ 90 s or the line cannot meet demand.

Little's law governs all flow: L = λW, where L = WIP (units in system), λ = throughput (units/time), W = flow time. Why it matters: it proves that reducing WIP is the only way to cut lead time at constant throughput. It holds for any stable system regardless of distributions — that generality is why the exam tests it.

Pull vs push: In push, upstream produces to schedule regardless of downstream need (WIP explodes under variation). In pull, the downstream process withdraws only what it needs (kanban signal); production is capped. The kanban count formula:

N = D × LT × (1 + S) / C

where N = number of kanbans (round UP — never down), D = demand rate (units/time), LT = replenishment lead time (same time units), S = safety factor (typically 0.1–0.3), C = container capacity (units). Derivation: D×LT is the demand during replenishment; the (1+S) buffer covers variation; dividing by C converts units to containers. Rounding up is the trap: rounding down starves the line.

Line balancing: Assign work elements to stations so each station's time ≤ takt. Balance efficiency = total work content / (number of stations × takt). Aim > 85%. Yamazumi charts visualize operator work vs takt.

2.3 Setup reduction (SMED) and TPM

SMED (Single-Minute Exchange of Die) converts internal setup (machine stopped) to external setup (done while running). The four stages: (1) separate internal/external, (2) convert internal to external, (3) streamline remaining internal (parallel operations, functional clamps, eliminate adjustments), (4) eliminate setup entirely where economic. Exam trap: "single minute" means single-digit minutes (<10), not one minute.

TPM and OEE: OEE = Availability × Performance × Quality. Availability = run time / planned production time; Performance = (ideal cycle time × actual output) / run time; Quality = good units / total units. World-class OEE ≈ 85% (0.9 × 0.95 × 0.999). The six big losses map to the three factors: breakdowns and setup/adjustments hit availability; idling/minor stops and reduced speed hit performance; defects/rework and startup rejects hit quality.

2.4 Quality at the source

Jidoka (autonomation): machines detect abnormalities and stop — andon cords, fixed-position stops. Poka-yoke (error-proofing): design so the error is impossible (shape-coded connectors) rather than detectable. Exam distinction: jidoka stops the process when a defect is found; poka-yoke prevents the error from becoming a defect. 5 Whys for root cause — the trap is stopping at symptoms; the "greatest contribution of 5-Why is identifying the root cause, not the solution."

2.5 Heijunka, standard work, 5S

Heijunka (leveling): sequence production to smooth volume and mix — e.g., produce A-B-A-C-A-B rather than all A's then all B's. It reduces the bullwhip amplification upstream. Standard work = the documented, repeatable method: takt time, work sequence, standard WIP. 5S in order: Sort, Set in order, Shine, Standardize, Sustain — sustain is the one that fails in practice and the one exam questions probe.

3. Key equations

Equation Units / notes Assumptions
Takt = available time / demand seconds (or min) per unit Demand known and stable over the planning period
N = D·LT·(1+S) / C, round up kanbans (count) Constant average demand; LT includes all replenishment time
L = λW (Little's law) units = (units/time)(time) System stable over the measurement window
Balance efficiency = Σ work / (stations × takt) dimensionless Work content measured at standard pace
OEE = A × P × Q dimensionless Planned time excludes scheduled downtime
Process lead time ≈ Σ cycle times + Σ queue times time Queues dominate — verify with VSM data
VA ratio = VA time / lead time dimensionless VA strictly customer-defined

4. Worked examples

Example 1 — Takt and staffing. Customer demand 1,200 units/day. Available: 2 shifts × 7.5 h net = 15 h = 54,000 s. Total manual work content per unit = 270 s. Takt = 54,000 / 1,200 = 45 s/unit. Minimum operators = 270 / 45 = 6.0 → 6 operators at 100% balance; plan 7 with a floater for variation, or rebalance to keep everyone under 45 s. If demand rises 10% to 1,320: takt = 40.9 s, operators = 270/40.9 = 6.6 → 7 minimum.

Example 2 — Kanban sizing. D = 400 units/day, LT = 0.5 day, S = 0.2, C = 50 units. N = 400 × 0.5 × 1.2 / 50 = 4.8 → 5 kanbans. A distractor answer of 4 starves the line 20% of the time — round up.

Example 3 — Little's law. Throughput λ = 60 units/h, WIP L = 480 units. Flow time W = 480/60 = 8 h. To hit a 2-h lead-time target without adding capacity, WIP must fall to L = 60 × 2 = 120 units — a 75% WIP cut. This is why lean attacks inventory.

Example 4 — OEE. Planned time 480 min, downtime 60 min → A = 420/480 = 0.875. Ideal cycle 1 min, output 380 → P = 380/420 = 0.905. Good units 361 → Q = 361/380 = 0.950. OEE = 0.875 × 0.905 × 0.950 = 0.752 (75.2%). Biggest lever: availability (downtime). Don't chase quality when A is the constraint.

Example 5 — SMED economics. Changeover 45 min, 2/day, run rate 100 units/h, margin $2/unit. Daily lost production = 1.5 h × 100 = 150 units = $300/day margin ≈ $75k/year. SMED to 9 min recovers 1.2 h/day → $60k/year. That payback math is how Silver/Gold portfolios justify projects.

5. Standards cross-reference

  • ISO 9001 (quality management) — lean's quality-at-the-source tools (jidoka, poka-yoke, standard work) are how the process-control clauses are actually satisfied on the floor.
  • ISO 14001 / ISO 45001 — 5S and visual management directly support environmental and safety management; ergonomics improvements from standard-work kaizen map to ISO 45001 hazard controls.
  • Shingo Model — the cultural-enabler and enterprise-alignment dimensions of the BOK come straight from Shingo guiding principles; Gold portfolios are judged against them.
  • OSHA 29 CFR 1910 — setup-reduction and standard-work changes must not create new machine-guarding or lockout/tagout exposures; every SMED kaizen needs a safety review.

6. Pitfalls and trap patterns

  • Current state before future state. Any answer proposing tools before the value stream is mapped is wrong.
  • Pull ≠ kanban cards everywhere. Pull applies where continuous flow is impossible; where flow is possible, flow beats pull.
  • Heijunka vs batching. Leveling is the opposite of running economic batch quantities — exam distractors will offer "larger batches" as the leveling answer. Reject.
  • OEE gaming. Raising OEE by overproducing inflates inventory — the classic suboptimization trap; OEE is a loss-visibility tool, not a production target.
  • Portfolio, not just exam. Candidates who pass the exam then submit thin portfolios fail. Start documenting projects, hours, and results from day one.

7. Practice problems (answers at end)

  1. Demand 900 units/day, net available time 7.5 h/shift, 2 shifts. Total work content 300 s/unit. Compute takt and minimum operators.
  2. D = 1,000/day, LT = 2 days, S = 0.25, C = 100. How many kanbans?
  3. A cell has WIP of 150 units and throughput 25 units/h. What is the flow time? If the target is 3 h, what WIP is allowed?
  4. Planned time 600 min, unplanned downtime 90 min, ideal cycle 2 min, actual output 230, rejects 12. Compute OEE and name the biggest loss factor.
  5. Changeover takes 30 min, performed 3×/day. SMED cuts it to 6 min. At 200 units/h and $1.50/unit margin, what is the annual (250 days) margin recovered?
  6. A 4-station line has work content 200 s/unit and takt 55 s. Compute balance efficiency and the minimum stations theoretically needed.

Answers: 1. Takt = 54,000/900 = 60 s; operators = 300/60 = 5. 2. N = 1000×2×1.25/100 = 25 kanbans. 3. W = 150/25 = 6 h; allowed WIP = 25×3 = 75 units. 4. A = 510/600 = 0.85; P = (2×230)/510 = 0.902; Q = 218/230 = 0.948; OEE = 0.727; biggest loss = availability (downtime). 5. Saved = 24 min × 3 = 72 min/day = 1.2 h × 200 = 240 units/day × $1.50 = $360/day × 250 = $90,000/year. 6. Efficiency = 200/(4×55) = 90.9%; theoretical minimum = ceil(200/55) = 4 stations.


Original ETS reference writing, verified against public sources on 2026-09-30. Thin sourcing is flagged inline. Not legal advice — confirm current editions with the issuing body.

Next step

Credentials prove you studied. ETS proves the part works.

The Academy maps to the codes your auditors ask about — PPAP, APQP, FMEA, GD&T, ISO 13485 — taught by engineers who have defended designs in production and in court.

  • NDA on request
  • Senior engineer replies in one business day
  • Fixed price where applicable
  • Your data stays yours