Muda (waste), Mura (irregularity/fluctuation), and Muri (overburden/unreasonableness) are the three root causes of inefficiency in the Toyota Production System (TPS). The goal of TPS is to avoid Muri, eliminate Mura, and eliminate Muda – in precisely this order, because irregularity and overload typically cause visible waste in the first place.
Practical relevance
• Muda (waste): All activities that do not benefit the customer; traditionally categorized into seven types of muda: transport, inventory, motion, waiting, overproduction, overprocessing, defects/rework (often later supplemented by "unused talent" as an eighth category).
• Mura (irregularity): Fluctuations in quantity, mix, or cycle time—e.g., batching, volatile demand, inconsistent process times—that drive inventory, waiting times, and quality risks. Heijunka (leveling) smooths out mix and volume, thereby reducing mura.
• Muri (overload): Excessive load on people, machines, systems (e.g., unrealistic cycles, too tight resource buffers). Muri leads to errors, failures, and safety issues – and thus directly to muda.
Important: Toyota links muda, mura, and muri – mura/muri generate muda. That is why waste is not blindly "hunted down," but rather flow capability is built first (capacitively, tactically, architecturally).
Typical misunderstandings
• "Lean = eliminate 7 wastes" – incomplete. Just hunting Muda without addressing Mura/Muri shifts problems. Leveling (Heijunka), takt, standard work, and Jidoka are essential.
• "The eighth waste is Toyota canon" – historically not true; it originates from the Lean adaptation outside Toyota (often "unused talent"), technically useful, but not original TPS.
• "Mura = natural demand volatility, nothing can be done about it" – wrong. Smoothing, mix leveling, pull/kanban noticeably reduce real volatility in the system.
Relevance for organizations (beyond production)
The Three M logic is domain-agnostic: the same mechanisms apply in logistics, services, and IT/software – fluctuation (mura) and overload (muri) destroy flow, create queues, and make work unreliable; muda is the symptom. In knowledge work, this translates into task switching, handoffs, waiting for decisions, and overplanning – all of which are direct cost drivers for lead/cycle time. Lean software practitioners (Poppendieck) apply the 7 types of muda to partially completed work, extra features, handoffs, context switching, waiting, re-learning, defects, and more.
Real-world example
An e-commerce fulfillment center experiences seasonal peaks in mura. For years, the department's response was "more people, more overtime" (muri). The result: an increasing error rate (muda) and returns. After introducing Heijunka (mix smoothing, smaller batch sizes), pull signals via Kanban, and time slots for releases, returns decreased noticeably and the throughput time per order stabilized. The result was primarily due to the reduction of mura/muri, not "waste lists."
Strategies & countermeasures
A. First Mura/Muri, then Muda
• Heijunka (volume & mix smoothing), takt and capacity buffer before waste actions – otherwise teams will only produce waste more "evenly."
B. Flow Design & Pull
• Kanban/pull to decouple variable predecessor/successor steps; FIFO lanes against priority chaos; WIP limits against Muri. (Lean lexicon, TPS principles)
C. Jidoka & Standard Work
• Make errors visible at the source (Andon, Stop-the-Line), standard work as the temporarily best approach, PDCA/Kaizen for systematic adjustment.
D. Leveled releases & blackout windows (IT/product)
• Release calendar instead of "anytime deploys" with high coupling, blackout times during peak business periods, feature toggles for mix smoothing without a big bang. (Lean transfer)
E. Knowledge work: Three-M mapping
• Muda: Handoffs, waiting times, overprocessing in reports/slide battles.
• Mura: uneven ticket influx, management "sawtooth priorities."
• Muri: 120% utilization plans, constant context switching.
→ Remedies: WIP limits, classification by service classes, clear "last responsible moment," team APIs/ownership (lean software).
F. Measurement system
Predictability (P85/P95 cycle time), WIP/queue length, change failure rate, OEE/line balance in manufacturing; mura index (variance in arrival/mix), muri proxy (overtime, rework, escalations). (Lean practice/LEI)
How good coaches work
1. Gemba & value stream view: Observe where fluctuation/overload occurs; make mura/muri visible, don't just collect muda lists.
2. Heijunka planning: Weekly/PI grid with smoothing of mix/volume, slotting for variants.
3. WIP design: WIP limits per stage, pull rules, clear DoR/DoD.
4. Jidoka rules: Institutionalize stop-the-line; root cause analysis instead of "overtime."
5. Rhythm & cadences: Set takt/release cadences; mix leveling through smaller batches.
6. Leadership mission: Eliminate overload (muri no-go), manage variance (mura), reduce waste (muda) – in that order.
CALADE perspective
We prioritize system design over symptom hunting: Mura (fluctuation) and Muri (overload) are addressed first (Heijunka-capable cadences, pull, slotting, WIP architecture). We then eliminate Muda measurably along lead/cycle time and predictability. In knowledge work, we combine this with team topologies (team APIs, platform-as-a-product) and – if desired – with Living Strategy®/Living Transformation® to permanently anchor capacities, priorities, and smoothing in operations – without event hype. (Objective, without dogma.)
Related terms & sources
• Toyota Production System (TPS) – Goals: Design out Muri and Mura, eliminate Muda.
• Lean Lexicon / LEI – Classification of Muda/Mura/Muri, 7 wastes.
• Heijunka (leveling) – remedy for Mura/Muri, prerequisite for JIT.
• Kaizen/Jidoka/standard work – supporting pillars alongside JIT.
• Lean Software Development (Poppendieck) – 7 Muda equivalents in knowledge work.
• Toyota UK / Toyota Forklifts – Practical TPS representations with reference to the Three Ms.
← back to list