Lean Manufacturing (Toyota Production System) — Engineering Reference
1. At a glance
Lean manufacturing is the systematic elimination of waste (muda) while maximising flow and value to the customer. It originated as the Toyota Production System (TPS) developed inside Toyota Motor Corporation from roughly 1945 through the 1970s, principally under Taiichi Ohno (chief engineer, plant manager) with methodological contributions from Shigeo Shingo (external consultant, SMED + poka-yoke) and conceptual roots in Sakichi Toyoda’s automatic-loom jidoka (auto-stop on broken thread, 1924) and Kiichiro Toyoda’s “just-in-time” wartime parts-supply philosophy. The system was largely opaque to Western industry until Womack, Jones & Roos published The Machine That Changed the World (MIT IMVP study, 1990), which coined the term “lean” and benchmarked Japanese auto-plants at ~½ the labour-hours, ½ the defects, ⅓ the engineering hours, and 2–3× the inventory turns of mass-production peers.
Two pillars carry the TPS roof: Just-in-Time (JIT) — produce only what is needed, when needed, in the quantity needed — and Jidoka (autonomation, “automation with a human touch”) — build in the ability for the process or worker to detect abnormality and stop. Both rest on a foundation of standardised work, Heijunka (level loading), Kaizen (continuous improvement), and visual management. The central pillar of the house is people and teamwork — TPS is explicit that tools without a culture of respect, training, and learning is “lipstick on a pig”.
Outside automotive, the principles have been ported widely: lean services (back-office), lean healthcare (Virginia Mason, ThedaCare), lean software / kanban (Poppendieck 2003, David Anderson 2010), lean startup (Ries 2011), lean construction (Last Planner System, Ballard 2000). Many modern movements — Six Sigma’s DMAIC, agile retrospectives, DevOps value-stream metrics — are direct descendants or hybrids. Lean is now the dominant operations-improvement framework in industrial engineering.
2. Why it matters
The empirical case is durable. Across the 1985–2025 period, Toyota’s inventory turns ran 2–3× the Detroit Big Three; finished-goods days-on-hand under 10 vs 60+; first-time-through quality routinely >95 % on body-in-white. Toyota’s operating margin held positive through every global downturn except 2009. Independent IMVP plant-benchmark studies (1989–1994) and later Liker (The Toyota Way, 2004) documented 2× productivity, ½ defect rates, ⅓ engineering hours, smaller floorspace per unit.
Reported lean-transformation results outside Toyota are more variable but characteristic: 20–50 % lead-time reduction, 10–30 % labour-productivity gain, 40–80 % WIP reduction, 30–50 % floorspace reclamation, scrap/rework reduction 20–60 % when sustained 3+ years (LEI, Shingo Institute, McKinsey Lean Operations surveys). The failure mode is consistent: roughly two-thirds of lean programmes lose momentum within 5 years because the tooling was deployed without the management-system or cultural rewiring. Tools without thinking ≠ lean.
The economic mechanism is straightforward. Inventory ties up working capital + space + risk (obsolescence, damage). Long cycle times hide quality problems (defects discovered far downstream cost 10–100× more to fix). Push-based production decouples manufacturing from customer demand and overproduces — the worst waste because it generates all the others. Lean attacks all three simultaneously by replacing push schedules with pull signals, levelling demand to absorb variation, and tightening loop-time so problems surface at the workstation where the originator can fix them.
3. First principles — the eight wastes (TIMWOODS)
Ohno’s original seven wastes, with an eighth (unused human skill) added by Liker and the western lean community in the 1990s. Mnemonic: TIMWOODS.
| # | Waste (English / Japanese) | Definition | Typical example | Counter-measure |
|---|---|---|---|---|
| 1 | Transportation (運搬, unpan) | Moving material between operations | Forklift shuttling between aisle storage and line | Cellular layout, point-of-use storage |
| 2 | Inventory (在庫, zaiko) | Raw, WIP, or finished beyond pull-demand | 6-week buffer of subassemblies | Kanban, small-lot, JIT delivery |
| 3 | Motion (動作, dōsa) | Operator movement that adds no value | Reaching across bench for tool | 5S, ergonomic layout, shadow boards |
| 4 | Waiting (手待ち, temachi) | Idle time of person, machine, or material | Operator waiting on QC inspection | Line-balancing, SMED, level loading |
| 5 | Over-production (作りすぎ, tsukurisugi) | Making before pull or beyond demand | Building to forecast not to order | Pull system, kanban, takt-paced flow |
| 6 | Over-processing (加工, kakō) | More work / precision than spec requires | Polishing internal surface of hidden housing | Standard work to spec only |
| 7 | Defects (不良, furyō) | Scrap, rework, warranty | Wrong torque, blemished paint | Jidoka, poka-yoke, source inspection |
| 8 | Skills (unused, modern addition) | Human capability not engaged | Operator ideas ignored, no improvement role | Kaizen circles, suggestion systems, training |
Two additional waste-families complement muda:
- Mura (斑) — unevenness or variability in load, demand, or quality. Attacked by Heijunka.
- Muri (無理) — overburden of people or equipment. Attacked by takt-time discipline and standard work.
The three together — muda + mura + muri — are sometimes called the “3 M’s of waste”. Mura and muri generate muda; attack mura first, then muri, then residual muda.
4. The TPS House — canonical diagram
Toyota’s internal training material codified TPS as a “house” because the structure only stands if every element is in place — knock out one and the roof collapses. Variants of the diagram differ in detail but share the structure below.
┌─────────────────────────────────────────────────────┐
ROOF → │ Best quality · Lowest cost · Shortest lead-time │
│ · Safest workplace · High morale │
└─────────────────────────────────────────────────────┘
┃ ┃
PILLAR 1 ━━┫ JUST-IN-TIME ┃ PILLAR 2 ━━ JIDOKA
┃ · Takt time ┃ (autonomation)
┃ · Continuous flow ┃ · Stop on abnormality
┃ · Pull system (kanban) ┃ · Andon (visual signal)
┃ · Quick changeover (SMED) ┃ · Poka-yoke (mistake-proof)
┃ · Integrated logistics ┃ · Separate man-from-machine
┃ ┃
┃ ┌────────────────────┐ ┃
CENTRE → ┃ │ People & Teamwork │ ┃
┃ │ Genchi Genbutsu │ ┃
┃ │ Continuous Kaizen │ ┃
┃ └────────────────────┘ ┃
┃ ┃
┌─────────────────────────────────────────────────────┐
BASE → │ Heijunka (level loading) · Standardised Work │
│ Visual Management · Stable & reliable processes │
└─────────────────────────────────────────────────────┘
The base is stability — you cannot pull from an unreliable upstream; you cannot level demand to an unstable process. Most failed lean programmes start at the pillars (kanban, SMED) without the base (5S, standard work, TPM-style reliability), and the pillars fail in months.
5. Core lean techniques
5.1 Value Stream Mapping (VSM) — Rother & Shook 1998
A paper-and-pencil flowchart of the material + information flow for one product family from raw-material supplier to customer delivery. Drawn in two passes: current state (what is) and future state (what should be after improvement). Symbols are standard (factory, truck, push arrow, kanban, supermarket, FIFO lane, kaizen burst). Key metrics on each step: C/T (cycle time), C/O (changeover), uptime %, batch size, operators. Lead-time ladder along the bottom segregates value-added time (cumulative cycle times) from lead time (cumulative wait). Typical ratio: value-add ~0.5–5 % of lead time. The future-state map is a hypothesis — implementation is by kaizen events.
5.2 5S — workplace organisation
| Japanese | English | Purpose |
|---|---|---|
| Seiri (整理) | Sort | Remove what is not needed at the workstation |
| Seiton (整頓) | Set in order | A place for everything, everything in its place (shadow boards, taped outlines) |
| Seiso (清掃) | Shine | Clean = inspect; daily cleaning surfaces leaks/wear |
| Seiketsu (清潔) | Standardise | Visual rules so deviations are obvious |
| Shitsuke (躾) | Sustain | Audit, lead by example, build the habit |
5S is the visible foundation — the first thing visitors notice. It is necessary but not sufficient; many shops “do 5S” and stop.
5.3 Kanban — pull system (Ohno 1948)
A signal (originally a card, kanban = signboard/card) attached to a container of parts. When the downstream process consumes the container, the kanban is returned upstream as the authorisation to make / move the next container. Production stops when no kanban is present — the inventory in the loop is bounded by the number of kanban cards × container quantity.
Kanban quantity formula (Ohno’s original heuristic, widely cited):
N = (D · LT · (1 + SF)) / Q
| Variable | Meaning | Typical value |
|---|---|---|
| N | Number of kanban cards in the loop | Computed (round up) |
| D | Average demand rate (units / time) | From takt or forecast |
| LT | Replenishment lead time (same time unit) | Cycle + transport + queue |
| SF | Safety factor (fraction) | 0.10–0.30 |
| Q | Container quantity (units / kanban) | Set by ergonomics + logistics |
Variants: production kanban (signal to make), withdrawal kanban (signal to move), supplier kanban (extended to vendors — Toyota’s keiretsu), e-kanban (electronic — barcode or RFID scan triggers the next pull). Two-bin / three-bin is a simpler kanban form for low-value items: when bin 1 empties, it goes back as the order; bin 2 covers replenishment lead-time.
5.4 SMED — Single-Minute Exchange of Die (Shingo 1985)
Reduce changeover (setup) time to single digits of minutes (< 10 min). Method:
- Observe and time the current changeover (video helps).
- Separate internal from external setup. Internal = must be done with the machine stopped. External = can be done while machine is still running. Pre-staging dies, tools, and material is the single largest gain.
- Convert internal to external — pre-heat dies, use intermediate jigs, pre-set tool offsets, modular fixturing.
- Streamline what remains — quick-release clamps (cam, hydraulic, magnetic), eliminate adjustments by using stops + pins, parallel operations (two operators working both sides).
- Standardise + train + measure.
SMED enables small-batch production, which is the prerequisite for Heijunka and pull. Without SMED, every changeover penalises small batches and the system reverts to large batches and push.
5.5 Heijunka — level loading
Smoothing the production mix to dampen demand variability. Instead of running AAAAAAAAAA BBBBBBBBBB CCCCCCCCCC, run ABCABCABCABC. Same total output, dramatically lower upstream demand-variance, smaller kanban loops, smaller finished-goods buffer.
Implementation: a Heijunka box — a physical pigeonhole grid (rows = part types, columns = small time-slots, typically 15–60 min) into which production-kanban cards are sequenced for the day. The pacemaker process pulls cards from the box at takt.
5.6 Takt, cycle time, lead time
| Term | Symbol | Formula | Meaning |
|---|---|---|---|
| Takt time | T_takt | Available time / customer demand | Pace at which one unit must be produced to meet demand |
| Cycle time | C/T | Actual time / unit at a station | What the station actually achieves |
| Lead time | L/T | Order-to-delivery elapsed time | What the customer feels |
| Throughput time | T_p | Raw-in to finished-out | Internal manufacturing duration |
Line balanced when every station C/T ≤ T_takt and the slowest station (bottleneck) C/T ≈ T_takt. Over-balanced lines (every station << takt) overproduce; under-balanced lines (any station > takt) starve the customer.
5.7 Andon, poka-yoke, jidoka
- Andon (行灯, “paper lantern”) — visual signal of line status. Originally a coloured light + pull-cord at each station. Operator pulls cord on abnormality; team-leader responds within takt; line stops if not resolved. Andon board shows status of every station to the whole plant. The pull-cord embodies Toyota’s empowerment principle.
- Poka-yoke (ポカヨケ, “mistake-proofing”, Shingo) — devices that make defect physically impossible or immediately detectable. Examples: USB connector orientation (cannot insert backwards), the asymmetric pins on a 3-prong AC plug, fuel-tank caps tethered to the car, jigs that only accept the correct part orientation, sensors that confirm a part is present before the next operation begins. Two modes: prevention (stops the error from happening) and detection (stops production immediately if it does).
- Jidoka — autonomous detection + stop. Originated in Sakichi Toyoda’s 1924 loom that auto-stopped on a broken weft thread. Modern equivalents: torque-monitoring fastener tools that lock out on out-of-spec, vision-system go/no-go at the station, force-monitoring presses.
5.8 Kaizen — continuous improvement
Two scales:
- Daily kaizen — every operator suggests small improvements; team-leaders implement same-shift where possible. Cumulative gain compounds.
- Kaizen blitz / kaizen event — 3–5 day cross-functional intensive on a specific value-stream segment. Day 1: train + map current. Day 2: brainstorm future state. Day 3–4: implement physical changes. Day 5: standardise + present. Typical gains: 30–70 % cycle-time reduction, 50 %+ floorspace.
5.9 Hoshin Kanri — policy deployment
Strategic alignment mechanism. Top-level objectives (1–3 years) cascade through catchball (iterative top-down/bottom-up negotiation) to department, team, individual annual goals — each measured. The “X-matrix” is a one-page tool that ties strategic priorities → annual objectives → improvement priorities → metrics → ownership. Prevents “lean tools doing things nobody asked for”.
5.10 A3 thinking — one-page problem solving
A3 (the paper size, ~297 × 420 mm) report walks left-to-right through: background → current state (with data) → goal → root-cause analysis → countermeasures → implementation plan → follow-up + lessons. The constraint to one page forces clarity. The 5 Whys (Ohno) — repeatedly asking “why?” until a root cause emerges — and the Ishikawa / fishbone diagram (cause + effect, 6M: Man, Machine, Method, Material, Measurement, Mother Nature) are the most common root-cause tools used inside the A3.
5.11 Genchi Genbutsu — “go and see”
Toyota leadership principle: go to the actual place, look at the actual problem, talk to the actual people. Decisions made at desks from reports are at best second-hand and often wrong. The shop floor is called the gemba (現場, “real place”); the leadership routine of walking it daily is the gemba walk.
6. Lean math — three worked examples
Example A — Takt time and line balancing
A subassembly cell ships 480 units/day to the customer. Working day = 2 shifts × 8 h, but with two 10-min breaks and a 20-min meal per shift, available time = 2 × (8·60 − 40) min/day = 2 × 440 = 880 min/day ≈ 460 effective min/day after a 5 % planned-stop factor for line meetings and changeovers… taking the brief value of 460 min/day.
T_takt = (460 min × 60 s/min) / 480 units = 27 600 s / 480 units = 57.5 s/unit
Eight stations with measured cycle times (s): 45, 60, 50, 55, 40, 65, 50, 55. Station 6 at 65 s exceeds takt — the line cannot meet demand and is the bottleneck.
Total work content = 45+60+50+55+40+65+50+55 = 420 s. Theoretical minimum number of stations at takt 57.5 s/unit = ⌈420 / 57.5⌉ = ⌈7.30⌉ = 8 stations. The current 8 stations are sufficient in total work, but mis-balanced.
Rebalance options:
- Move ~10 s of work from station 6 → station 5. New times: 45, 60, 50, 55, 50, 55, 50, 55 — max = 60 s, still > takt. Move 3 s from station 2 → station 1 and 3 s from station 6 → station 7: 48, 57, 50, 55, 50, 52, 53, 55 — max = 57 s ≤ takt.
- Balance efficiency η = total work / (stations × max station time) = 420 / (8 × 57) = 92.1 %. Industry benchmark “good” = 85 %+, “excellent” = 92 %+.
Example B — Little’s Law for WIP
Little’s Law: WIP = Throughput × Cycle-Time (long-run average, any stable queue).
A machining line produces 100 units/h at steady state with lead time 24 h through the value stream. Average WIP:
WIP = 100 units/h × 24 h = 2 400 units
Lean kaizen reduces L/T to 8 h (smaller batches, FIFO lanes, SMED on the bottleneck):
WIP_new = 100 × 8 = 800 units — a 67 % reduction
Implications: ~1 600 units freed from floor → ~50 % less buffer floorspace; working capital freed = 1 600 × unit cost; defect-discovery loop shrinks 3× (defects found 8 h after creation, not 24 h). Little’s Law is the algebra behind every “smaller WIP = shorter L/T at constant throughput” claim.
Example C — SMED before/after on a stamping die
Initial die change observed and timed at 240 min (4 h). Step 1 — separate internal from external: pre-stage the new die, tools, and material on a cart while the press is still running the previous job → internal = 80 min, external = 160 min. The press only stops for the 80 min.
Step 2 — convert internal to external: pre-heat the new die in a soaking oven (was 30 min on the press to bring to temp), pre-set die-shoe height with intermediate adapters offline, use a single-bolt quick-die-change clamp instead of 8 hex bolts → internal = 25 min.
Step 3 — streamline: parallel operations (one operator working each side of the press), eliminate trial stampings by laser-aligned datums on the die-shoe → internal = 9 min.
Result: changeover 240 → 9 min, a 27× improvement. Cost: ~USD 35 k in clamps, oven, alignment fixtures, training. Pay-back at this press (3 changeovers/day × USD 200/min downtime cost × 230 days/yr × (240 − 9) min) = USD 31 M/yr in recovered capacity — typical SMED ROI is months, not years.
The downstream effect is more important: with 9-min changeovers, batch size can drop from 2 000 → 200 pieces, finished-goods inventory drops ~90 %, and Heijunka mix-levelling becomes feasible across the eight part numbers that share the press.
7. JIT / Pull system in detail
7.1 Kanban sizing under variability
Ohno’s heuristic (Section 5.3) assumes deterministic demand and lead time. Under variability the formula generalises to:
N = ⌈(D · LT + z_α · σ_LT · √D) / Q⌉
where z_α is the safety-stock service-level factor (1.65 for 95 %, 2.33 for 99 %) and σ_LT is the lead-time-demand standard deviation. When demand CoV > ~0.3, plain kanban breaks down (cards either pile up or starve) — consider DDMRP (Demand-Driven MRP, Ptak 2011) or CONWIP.
7.2 Supermarket + FIFO lane
- Supermarket — a controlled inventory between two processes that cannot be coupled by direct flow. Downstream pulls from supermarket; upstream replenishes. Kanban controls supermarket size.
- FIFO lane — when two processes can be coupled but with some buffer for variability. First-in-first-out by physical constraint (a roller conveyor sized for n parts). Bounded by lane length, not by kanban count.
7.3 CONWIP — constant WIP (Spearman, Woodruff, Hopp 1990)
Alternative pull bound: total WIP in the system is constant — when a unit exits, one is released at the head. Simpler than per-stage kanban; tolerates more product variety; widely adopted in high-mix low-volume manufacturing and software (it is the mathematical basis for the modern software-kanban WIP-limit). See Factory Physics (Hopp + Spearman, 3rd ed 2008).
7.4 Demand stability prerequisite
JIT is brittle if upstream supply is unreliable or demand is genuinely random. Pre-conditions: stable process (Cpk ≥ 1.33), stable supplier (on-time delivery >99 %), stable demand pattern (heijunka-able). Where these fail — exotic alloy with single-source supply, lumpy custom-engineered demand — pure JIT is inappropriate; use strategic decoupling buffers (DDMRP), make-to-order with quoted lead-time, or postponement instead.
8. Lean enterprise and lean beyond manufacturing
8.1 Lean office
VSM of an information flow (purchase-order processing, invoice approval, claims adjudication). The waste catalogue maps directly: transportation = handoffs between teams; waiting = sitting in an inbox; defects = rework; over-production = generating reports nobody reads; inventory = backlog of unprocessed tickets. 5S of digital workspaces (shared-drive folder structure, mailbox rules) is its own discipline.
8.2 Lean healthcare
Virginia Mason Medical Center (Seattle) adopted TPS post-2002 (then-CEO Gary Kaplan). ThedaCare (Wisconsin) under John Toussaint similarly. Documented: 50 %+ reduction in patient wait-time, ~40 % staff-walking-distance reduction, error-rate reductions in medication dispensing of 60 %+. Major US lean-healthcare framework: Lean Enterprise Institute + Catalysis (Toussaint’s nonprofit).
8.3 Lean software
Mary + Tom Poppendieck, Lean Software Development (2003) — translated Toyota principles to software (eliminate waste, amplify learning, decide as late as possible, deliver as fast as possible, empower the team, build integrity in, see the whole). Together with David J. Anderson (Kanban, 2010), it underpins the kanban-board / WIP-limit movement in software, which influenced DevOps (value-stream metrics) and agile (sprint retrospectives = kaizen events).
8.4 Lean startup
Eric Ries, The Lean Startup (2011) — applies the build-measure-learn loop and minimum-viable-product to product development under extreme uncertainty. The lean borrowing is loose (pull from validated learning) but the cultural lineage is direct: Ries worked with Steve Blank, who studied Toyota practice.
8.5 Lean construction
Glenn Ballard + Greg Howell, Last Planner System (LCI, 2000). Weekly + daily lookahead planning, percent-promises-complete (PPC) metric, makes-ready process. Now common on large commercial + infrastructure jobs (LCI member firms include Skanska, Turner, DPR).
9. Implementation roadmap
A typical sustained lean transformation, by month:
| Phase | Months | Activities |
|---|---|---|
| Stabilise | 1–3 | Leadership commitment; pilot value stream selection; current-state VSM; 5S baseline; reliability audit (OEE, Cpk); training |
| Flow | 3–9 | Cellular layout in pilot; kanban introduction; SMED on bottlenecks; standardised work; kaizen events (1–2 / month) |
| Pull | 6–12 | Supermarkets + FIFO; supplier kanban for top-N parts; daily-management cadence; andon discipline |
| Level | 12–18 | Heijunka boxes at pacemaker; production smoothing; expand across product families |
| Perfect | 12–24 | Hoshin kanri rollout; A3 culture; lean accounting (replace standard cost with value-stream costing); broad workforce engagement |
| Sustain | continuous | Layered audits; leader standard work; gemba walks; new-hire onboarding incorporates lean; supplier development |
Toyota internally treats this as a 30-year horizon; most external transformations claim “lean transformation” at year 3, when the real cultural work is still ahead. Liker (2004) and Rother (Toyota Kata, 2009) emphasise the management-routine (“improvement kata”, “coaching kata”) as the durable mechanism — tools fade, daily routines compound.
10. Edge cases and gotchas
| Pitfall | What goes wrong | Mitigation |
|---|---|---|
| Tools without thinking | 5S signs and kanban boards installed; no cultural shift; reverts within 18 mo | Hoshin kanri + leader standard work + gemba routines from day 1 |
| Push-to-pull cutover too fast | Stockouts, missed customer ship dates, executive panic, revert to push | Phase: reduce safety stock 20 % / quarter while improving Cpk + L/T |
| JIT supply-chain brittleness | 2011 Tōhoku earthquake (Renesas chip plant), COVID-19 (2020–2022 semi shortage), Suez Ever Given 2021 | Hybrid “just-in-case” buffers at strategic decoupling points; dual-sourcing; DDMRP |
| Lean = layoffs | Workforce learns kaizen → job loss; ideas stop | Toyota’s “no layoff from kaizen” commitment; absorb gains by growth / re-deployment |
| Mass-production / Ford confusion | ”We already do lean — we have a moving line” | Ford 1913 was flow + flexibility-zero; TPS adds flexibility + JIT + jidoka |
| Standard work as ceiling | Operators forbidden to deviate; improvements stop | Standard work is the current best = the baseline that kaizen raises |
| One-piece-flow dogma | Forced 1-piece flow in transactional / high-variability work; throughput drops | Small batches (5–25) often better in office / claims work; pure 1-piece is a manufacturing ideal |
| Lean vs Six Sigma silos | Two improvement programmes fighting for resources | Lean Six Sigma — DMAIC inside the value stream; lean reduces L/T, 6σ reduces variation |
| Cultural transfer | TPS practices implemented without Japanese work culture; resistance | 10+ year horizon (NUMMI joint venture 1984–2010 took ~5 years to converge) |
| Software cargo-cult | Kanban board with 30 “in-progress” columns and no WIP limit | Enforce WIP limits; measure flow time + throughput; weekly retro |
| Quality drift under lean pressure | Toyota’s 2009–2010 unintended-acceleration recalls; ~9 M vehicles | ”Lean must include reliability + quality engineering”; led to Toyota’s internal reorganisation + “Quality First” recommitment |
| High-mix low-volume | Kanban loops too many; Heijunka not feasible | CONWIP, DDMRP, postponement, modular product design |
11. Tools and software
11.1 VSM and process mapping
- eVSM (Strategos Inc) — purpose-built Visio plug-in, the lean-consultant standard.
- Microsoft Visio with VSM stencil sets.
- Lucidchart, Miro, Mural — online whiteboards with VSM template libraries.
- iGrafx — enterprise BPM with VSM module.
11.2 Kanban / project boards (digital pull)
- Trello, Jira (with Kanban template), Asana, ClickUp, Monday.com, Kanbanize, Planview LeanKit — software-kanban with WIP limits and flow analytics.
- Physical boards are still recommended on the shop floor: visibility + ritual outperform screens.
11.3 Scheduling, heijunka, APS
- SAP S/4HANA + APS / IBP, Siemens Opcenter APS (formerly Preactor), Plex Smart Manufacturing Platform (Rockwell), ASPROVA (high-mix scheduler popular in Japan), DELMIA Apriso.
11.4 MES / ERP for industrial lean
- Plex (cloud MES + ERP), Epicor Kinetic, SAP S/4HANA Manufacturing, Oracle Cloud Manufacturing, Infor M3 / CloudSuite Industrial, IFS Cloud, Tulip (low-code frontline-ops platform with integrated andon / standard-work / kaizen tooling).
11.5 Simulation
- Siemens Tecnomatix Plant Simulation, FlexSim, Simio, AnyLogic — discrete-event simulation for kanban-loop sizing, line-balancing what-ifs, supply-chain JIT stress-tests.
11.6 Andon and IIoT
- Tulip Andon, Litmus Edge, PTC ThingWorx, Falkonry, Splunk Industrial Asset Intelligence, Ignition (Inductive Automation) — modern IIoT andon platforms that replace hard-wired stack lights with networked dashboards + alerts.
11.7 Six Sigma + statistical
- Minitab + Minitab Workspace (formerly Companion by Minitab), JMP (SAS), Sigma Magic, EngineRoom (MoreSteam) — DOE, capability, MSA, hypothesis testing inside the lean-six-sigma DMAIC workflow.
12. Cross-references
[[Engineering/machining]]— subtractive manufacturing whose cells, SMED, and tool-life economics are core lean topics[[Engineering/casting-forging-forming]]— bulk-shaping processes; JIT and SMED on stamping dies[[Engineering/joining-welding]]— common lean cell within auto-body shops (BIW)[[Engineering/additive-manufacturing]]— JIT spare-parts and on-demand production complement to lean[[Engineering/reliability-engineering]]— TPM and OEE link lean to RAMS[[Engineering/ergonomics-human-factors]]— companion same batch; 5S motion + workstation design- planned
[[Engineering/six-sigma]]— Lean Six Sigma marriage (DMAIC inside the value stream) - planned
[[Engineering/supply-chain-management]]— supplier kanban, DDMRP, lean logistics [[Languages/Tier3/industrial-automation]]— PLC programming for andon, jidoka, and line-control logic
13. Citations
Canonical Toyota / TPS
- Taiichi Ohno, Toyota Production System: Beyond Large-Scale Production (Productivity Press, 1988 — English; original Japanese 1978).
- Shigeo Shingo, A Study of the Toyota Production System from an Industrial Engineering Viewpoint (2nd ed, Productivity Press, 1989).
- Shigeo Shingo, A Revolution in Manufacturing: The SMED System (Productivity Press, 1985).
- Shigeo Shingo, Zero Quality Control: Source Inspection and the Poka-Yoke System (Productivity Press, 1986).
- Toyota Motor Corporation, Toyota Production System: Internal Reference Manual (cited in Spear & Bowen 1999).
Western “lean” synthesis
- Womack, Jones & Roos, The Machine That Changed the World (Rawson Associates, 1990 — MIT IMVP study, coined “lean”).
- Womack & Jones, Lean Thinking (2nd ed, Free Press, 2003).
- Jeffrey Liker, The Toyota Way (2nd ed, McGraw-Hill, 2020) and The Toyota Way Fieldbook (with David Meier, 2006).
- Mike Rother, Toyota Kata (McGraw-Hill, 2009).
- Mike Rother & John Shook, Learning to See: Value Stream Mapping to Add Value and Eliminate Muda (LEI, 1998).
- Steven Spear & H. Kent Bowen, “Decoding the DNA of the Toyota Production System”, Harvard Business Review (Sep–Oct 1999).
Adjacent / derived movements
- Mary & Tom Poppendieck, Lean Software Development: An Agile Toolkit (Addison-Wesley, 2003).
- David J. Anderson, Kanban: Successful Evolutionary Change for Your Technology Business (Blue Hole Press, 2010).
- Eric Ries, The Lean Startup (Crown, 2011).
- Donald Reinertsen, The Principles of Product Development Flow (Celeritas, 2009).
- Wallace Hopp & Mark Spearman, Factory Physics (3rd ed, Waveland, 2008) — CONWIP, queueing theory.
- Glenn Ballard, The Last Planner System of Production Control (PhD thesis, U. Birmingham, 2000).
Standards and institutions
- SAE J4000 Identification and Measurement of Best Practice in Implementation of Lean Operation (1999) and SAE J4001 (companion user manual).
- AIAG-VDA FMEA Handbook (2019) and AIAG core quality manuals (APQP, PPAP, MSA, SPC).
- Lean Enterprise Institute (LEI) — publications and Planet Lean journal.
- Shingo Institute (Utah State University) — Shingo Model and Shingo Prize.
- Shingijutsu Co Ltd — consulting publications by former Toyota engineers (Iwata, Nakao).
- Toyota Motor Corporation, Sustainability Report and TPS pages at corporate.toyota.
Session log:
node ~/.claude/bin/obsidian-research.mjs log "Built Engineering/lean-manufacturing.md Tier 2 deep note"