| 5-Why root cause analysis | Ask 'why?' again and again (about five times) until you reach the real cause of a problem, not the symptom | Recurrence count of the same problem per month; rejection % for the defect studied; downtime minutes for the stoppage studied | Rejection level, cylinder block castings (sand drop): rejection level from 37.17% to 16.3% | Case study |
| 5S (workplace organisation) | Put every tool, die and gauge in a fixed, labelled place so nobody hunts for anything and problems are visible | Minutes spent searching for tools/dies/gauges per shift; weekly 5S audit score; floor space freed (sq ft) | Tool search time; 5S audit score: 30 -> 5 min; audit score 7 -> 55 (week 1 -> week 20) (IN) | Peer-reviewed study |
| 8D problem solving | A team-based 8-step routine for a serious or customer problem: contain it, find the root cause, fix it permanently, prevent repeat | Rejection % / ppm for the problem part; number of repeat customer complaints on the same issue | Valve spring rejection: rejection from 17.07% (6-month average) to 4.91% (July 2014); target 5% | Peer-reviewed study |
| A3 problem solving | One page per problem: facts, root cause, actions, owner, follow-up check - so problems get closed, not just discussed | The KPI of the chosen problem (rejection %, ppm, scrap pieces/month, breakdowns/month, setup minutes); plus share of A3s closed with target met and on time | Share of A3 reports fully successful (target met, actions on time): 6 A3 reports in 3 plants over 1.5 years: 50% successful, 33% moderate, 17% low-performing (PL) | Peer-reviewed study |
| ABC/XYZ inventory classification | Sort every stocked item by value (ABC) and by how steady its usage is (XYZ), then set stock rules per class instead of one rule for all | Inventory value and days of stock; number of stock-outs per month (by class); obsolete/dead stock value; item count held | Average monthly stock-out frequency: decreased by 88.5% (P < 0.001), 6-month pilot (IN) | Peer-reviewed study |
| Assembly line balancing | Split work between stations so each takes about the same time, so no station waits and no single one holds up the line | Line efficiency %, balance delay %, output per shift, output per operator, number of stations/operators, idle time per station | Fuel tank assembly line output: productivity enhanced from 140 to 158 units per shift within 14 months (IN) | Peer-reviewed study |
| Automated visual inspection (machine vision / AI camera) | Cameras plus software check every part for visible defects at line speed instead of a tired inspector checking samples | Defect escape rate to customer (ppm), false rejects of good parts %, inspection seconds per part, inspectors per shift | Classification accuracy, response time: deployed CNN defect finder: accuracy over 99%, about 0.14 s per part (requirement 95% within 0.2 s) (KR) | Peer-reviewed study |
| Autonomous maintenance | Operators clean, inspect, lubricate and tighten their own machine every day with a simple checklist, catching faults before breakdowns | OEE and its parts (availability, performance, quality); unplanned stop hours per month; number of minor stops; abnormality tags raised/closed | OEE of one machine (Alfing): OEE April 37.65%, May 71.17%, June 77.38%, July 77.62% after JH steps 1-3 (IN) | Case study |
| Cellular manufacturing (one-piece flow cells) | Put the few machines that make one part family next to each other in a cell, so parts move one by one instead of in batches | Order throughput time (days/hours, release to dispatch); WIP pieces between operations; floor area used; operators per unit output | Production lead time; WIP; operators: lead time two days -> one hour; WIP 500-1500 -> max 100 pieces; operators -14%; rework -80% | Case study |
| CMMS / digital maintenance log | One digital record of every machine, breakdown, repair, spare part and planned job, so maintenance is planned and history is not lost | MTTR; MTBF; number of stoppages per critical asset; % planned vs emergency work orders; maintenance cost as % of asset value; spares stock-outs | Maintenance cost as % of replacement asset value: %RAV 3.74% with CMMS vs 5.67% without (benchmark 2.5%) (PK) | Peer-reviewed study |
| Compressed-air leak detection and repair | Find and tag hissing air leaks (often with an ultrasonic gun), fix them, and recheck every quarter so compressors run less | Compressor kWh per day or per unit of output; leak rate as % of compressor output (load/unload test on a non-production day); number of open leak tags | Savings vs repair cost per cfm of leakage: on average $156 savings per cfm repaired vs $30 repair cost per cfm; 740 leaks, >50% at hoses, threads, connectors, fittings (US) | Peer-reviewed study |
| CONWIP (constant work-in-process) | Put a fixed number of cards on the whole line: a new job starts only when a finished one leaves, so WIP stays capped and orders flow faster | WIP count (jobs or pieces) on the floor; order flow time from release to dispatch; throughput per week; on-time delivery % | WIP inventory: WIP down ~20,000 units (27.79%); kanban cards 48 to 38; inner-ring replenishment 6.94 to 6.53 days (IN) | Peer-reviewed study |
| Cost of Poor Quality (COPQ) measurement | Put a rupee figure on scrap, rework, returns and inspection each month so quality losses get the same attention as sales | COPQ as % of sales (internal failure: scrap+rework; external failure: returns, warranty, penalties; appraisal; prevention); rupees per month per defect type | Cost of internal failure: cost of internal failures reduced by about 40% after on-line gauge inspection (IN) | Case study |
| Daily management (tiered daily meetings) | Short fixed daily meetings at line, section and plant level review yesterday's numbers, raise problems and assign owners | Quality defect index or rejection %; output vs plan; OEE; open actions and % closed on time; inventory days | Productivity: productivity +17% in the first year (randomized; bundle of 38 practices incl. daily production/quality meetings and daily-updated visual boards) (IN) | Peer-reviewed study |
| Demand-Driven MRP (DDMRP) | Keep sized stock buffers at a few chosen points and reorder from actual consumption, not forecasts, so material is there without piling up | Inventory value or days on hand; stock-outs / shortages per month; on-time delivery %; purchase/part lead time | Inventory level: inventory reduced by 52.53% while material consumption rose 8.7%, service level maintained | Peer-reviewed study |
| Design of Experiments (DOE / Taguchi) | Change several process settings together in a planned set of trial runs to find the combination that gives fewest defects | rejection % for the target defect; yield; Cpk of the response; scrap cost per month | Casting rejection rate: rejection rate reduced from 16.98% to 6.07% (IN) | Peer-reviewed study |
| Employee suggestion system | A simple, fast route for every worker to propose small fixes, with quick yes/no, quick action and visible reward | Ideas submitted per employee per year; % of ideas implemented; days from idea to decision; savings (Rs) from implemented ideas; % of workers who submitted at least one idea | Estimated economic effect of suggestions per firm: 225.4 billion yen/year across 278 firms (avg 810 million yen per firm); 12.8 suggestions per employee per year (JP) | Case study |
| Energy management system and energy audits | Measure where power and fuel go per unit made, fix the big leaks of energy, and keep a monthly routine so savings stay | Specific energy consumption (kWh or toe per tonne / per piece); monthly energy bill per unit of output; share of energy cost in conversion cost | Annual energy savings after detailed energy audits and implementation support: 757 ECMs implemented, 1409 toe/year energy savings, CO2 reductions over 12,700 t; monetary benefit Rs 1130 lakh/year (table) (IN) | Case study |
| Finite-capacity scheduling (APS) | Plan jobs machine by machine against real hours available, so promised dates reflect what the shop can actually do | % orders dispatched on or before promised date; planner hours per week to make the plan; utilisation of key machines; schedule changes per week | Compliance with delivery deadlines: from 60 to 90 percent; planning time 4 days to 1 (BR) | Vendor-reported |
| Gemba walk | Managers walk the shop floor on a fixed schedule, watch the real work, ask operators and fix what blocks them | Share of planned walks actually done; issues raised per walk and % closed within 7 days; OEE or output per shift on the walked lines; MTTR | Overall factory performance (OEE-type efficiency): approximately 5% increase in overall factory performance; 5% package maker; 1.5% extruder, alongside higher adherence to the gemba walk schedule (BR) | Case study |
| Heijunka (production levelling) | Plan the same small mix of products every day or week instead of big batches of one item, so load, stock and staffing stay even | Finished-goods and inter-process inventory (days of stock), average lot size, number of setups per month, on-time delivery / service level, daily output variation | Actual inventory (monthly average): actual inventory decreased in 23% in average monthly values (BR) | Peer-reviewed study |
| Hoshin kanri (policy deployment) | Pick 3-5 yearly breakthrough goals, cascade them into each department's targets and projects, and review progress monthly | Labour productivity (output per man-hour); on-time delivery %; order lead time; % of improvement projects linked to an annual goal; specific energy use | Labour productivity: labour productivity +6% over two years | Peer-reviewed study |
| In-plant milk-run with supermarket | A small store (supermarket) near the line plus a fixed-route, fixed-time trolley that tops up each machine, so nobody waits for material | Minutes machines/operators wait for material per shift; milk-run lap time; number of forklift/crane trips; floor area taken by line-side stock; material shortages at line | Milk-run lap time: 15% reduction of milk-run lap time; fewer laps; one vehicle with four wagons serves all lines | Peer-reviewed study |
| IoT machine monitoring | Simple sensors on machines record run, stop and count automatically, so you see real uptime and stop reasons per machine on a screen | Machine utilisation % and OEE % per machine/shift (auto-logged); unplanned stop minutes and count by reason; pieces per hour | Average efficiency of work centres (real-time OEE): average efficiency improved by 12.3% in 7 months (BR) | Peer-reviewed study |
| Ishikawa (fishbone) cause-and-effect diagram | A team draws all possible causes of one problem on a 'fish skeleton' (man, machine, method, material...) and then checks the likeliest | Rejection % for the specific defect; number of hypotheses checked vs confirmed | Defect percentage: defect percentage decreased from 6.51% to 5.84%; annual cost reduction 199,800 baht (TH) | Peer-reviewed study |
| ISO 9001 quality management system | A written, audited system for how the plant controls quality: procedures, records, corrective actions, management review | Rejection/wastage %, customer complaints per month, repeat non-conformances, on-time corrective-action closure; sales growth | Wastage and rejection: average wastage & rejection fell from 9.34% to 6.39% after ISO 9000 (machine breakdown 1.60% to 0.51%) (IN) | Case study |
| Jidoka and andon (stop-the-line signalling) | Machine or operator stops at the first abnormality and lights a signal; help comes fast and the cause is fixed before more bad parts are made | Response time from signal to help arriving (minutes); unscheduled downtime hours; number of andon calls by reason; defects escaping to next stage; OEE | Unscheduled downtime: 55.55% reduction in unscheduled downtime and 4% increase in OEE (IoT andon + BI dashboard) | Peer-reviewed study |
| Just-in-Time (JIT) production and purchasing | Make and buy only what the next step needs, when it needs it, so stock, waiting and lead time shrink across the plant | Inventory days (RM, WIP, FG) or inventory turns; order lead time door-to-door; on-time delivery %; stock-outs per month | Inventory level: inventories came down by 47%; lead times down 28% (JIT-Kanban with digital twin, 12 SMEs) (SA) | Peer-reviewed study |
| Kaizen (continuous improvement) | Small, frequent improvements made by the people who do the work, measured every week | Cycle time per station; output per person-hour; number of kaizens implemented per month | Cycle time; productivity: cycle time 80 -> 75 s; productivity +6.7%; line efficiency +2.9% (IN) | Peer-reviewed study |
| Kanban (pull system) | Make or move parts only when the next step signals it needs them, so stock between machines stays small and fixed | WIP units/value between stations; raw-material days of stock; stockout count per month; replenishment time | WIP; kanban cards; replenishment time: WIP -27.79% (~20,000 units); cards 48 -> 38; replenishment 6.94 -> 6.53 days (IN) | Peer-reviewed study |
| Line-side kitting | Stores pre-pick all parts for one product or one job into a kit and deliver it to the station, so operators stop walking and searching | Operator walk/search time per unit; line stoppages due to material shortage (min/shift); line-side storage area; line-side inventory value; wrong-part assembly errors | Operator walk time; line side storage space; inventory value: operator walk time -78.3%; line side storage space -69%; inventory value -54.3%; line side replenishment -29.6% (IN) | Case study |
| Measurement System Analysis (Gauge R&R) | Check whether your gauges and inspectors give the same reading on the same part before trusting any quality data | %GRR (of study variation or tolerance); number of distinct categories (ndc); inspector agreement % for visual checks | Total Gage R&R: Total Gage R&R from 67.80% to 10.11% after adding a Go/No-Go pin (ID) | Case study |
| OEE measurement (loss tracking) | Measure how much of planned time each machine really makes good parts, and which losses eat the rest | OEE = availability x performance x quality per machine; top 5 loss reasons in minutes; share of unclassified stops | Typical OEE level: ~55% (about 20 cases); performance losses dominate (SE) | Peer-reviewed study |
| Overall Labour Effectiveness (OLE) | Measures how much of paid worker time turns into good output: availability x performance x quality, like OEE but for people | OLE % per section/shift (availability, performance, quality ratios); good pieces per paid labour-hour; overtime hours per good piece | Return on assets (ROA), average of 4 periods after vs 4 before introducing OLE: average ROA increased by 1.4% (CZ) | Peer-reviewed study |
| Pareto analysis (80/20) | Rank defects or stoppages by count or cost so you fix the few causes that make most of the loss first | Share of total rejection/downtime from the top 1-3 causes; total rejection % or downtime hours | Process rejection rate from dominant cause (inappropriate composition): rejection rate 2.54% -> 0.82%; dominant cause = 84% of total rejection; unit cost -2.50% | Peer-reviewed study |
| Plant layout redesign (Systematic Layout Planning) | Re-arrange machines, stores and aisles in the order the material flows, using a from-to chart, so parts travel less and space frees up | Total material travel distance per product (m); material handling cost or trips per day; floor area used (sq m); throughput time | Operator travel distance per standard task cycle: operator travel distance in the classification area from 25.99 m to 18.16 m (-30.13%) (PE) | Case study |
| Poka-yoke (mistake-proofing) | A simple jig, sensor or stop that makes the wrong part or wrong step physically impossible | Rejections per month (or ppm) for the targeted defect; customer complaints for that defect | Monthly rejections: 12-56 per month (Jun-Dec 2016) -> 0 (Jan-Jun 2017) (IN) | Case study |
| Predictive maintenance (condition monitoring) | Measure machine health (vibration, heat, oil) on a route and repair only when readings show a fault is starting | Unplanned downtime hours/month on critical machines; number of breakdowns; MTBF; maintenance cost per month; share of planned vs emergency jobs | Unplanned downtime: 30% unplanned downtime reduced; 12% total maintenance cost decreased | Peer-reviewed study |
| Preventive (planned) maintenance | Service machines on a fixed calendar (oil, filters, belts, wear parts) instead of waiting for them to break down | Breakdowns per month; breakdown hours; MTBF; MTTR; maintenance cost per machine; PM plan compliance % | MTBF: indicative tangible benefits: 80% breakdown reduction, 200% MTBF increase, 35% maintenance cost reduction, 31% MTTR reduction (IN) | Case study |
| Process FMEA (PFMEA) | List step by step how each process can go wrong, score the risk, and fix the worst ones before they reach the customer | RPN or action priority of top failure modes; defect rate / ppm; customer complaints per month; breakdown minutes | Overall defect rate: overall defect rate was reduced by 87.2% (MY) | Peer-reviewed study |
| Quality circles | Small groups of shop-floor workers meet weekly, pick a problem in their own area and solve it with simple QC tools | Rejection/rework % on the circle's chosen problem; inspection time; number of problems closed per circle per year; circle meeting attendance | No-load rework: no-load rework reduced from 16% to 2% within five months (IN) | Peer-reviewed study |
| Quick Response Manufacturing (QRM) with POLCA | Cut the calendar time an order spends in the plant: product-focused cells, spare capacity, and POLCA cards that release work only when the next cell can take it | Order lead time (calendar days from order release to dispatch, or MCT); % orders on time; WIP count/value between cells; late-delivery penalty cost | Process lead time (front shape, products P1+P2): 14.4 days reduced to 5.23 days (PK) | Peer-reviewed study |
| Reliability-centred maintenance (RCM) | For each critical machine, list how it can fail and pick the right maintenance task per failure: inspect, replace on schedule, monitor, or run to failure | Breakdown downtime minutes/month per critical machine; failure rate; MTBF; availability %; maintenance cost | Breakdown level (filling machine): breakdown level reduced 40-60% vs semester 1 of 2018 (ID) | Peer-reviewed study |
| Sales and Operations Planning (S&OP) | A monthly meeting where sales, production, purchase and finance agree one plan: what will sell, what we can make, what stock we hold | Forecast accuracy (% or MAPE by product family); finished/raw inventory days; on-time-in-full to customers; order lead time for new products | Order lead time for newly introduced products: 67 percent reduction | Peer-reviewed study |
| Six Sigma (DMAIC) | A data-driven project that finds the real cause of a chronic quality problem and locks the fix in | Defects per unit / ppm / first-time yield for one product; process capability (Cpk); cost of rejection per month | Defects per unit; first-time yield; OEE; savings: DPU 0.18 -> 0.0068; FTY 83.5% -> 99.32%; OEE 0.48 -> 0.83; ~$140,000/yr (IN) | Peer-reviewed study |
| Skill matrix and cross-training | A wall chart of who can run which machine/job at what level, used to plan training so absences and bottlenecks don't stop the line | No. of operators qualified per critical machine/job; output lost on days with absenteeism; line efficiency %; labour productivity (pcs per operator-hour) | Sewing line efficiency; labour productivity: Efficiency +21%, labour productivity +22% (balanced layout with multi-skilled manpower) | Peer-reviewed study |
| SMED (quick changeover) | Cut the time a machine stands still between the last good part of one job and the first good part of the next | Changeover time per change (last good part -> first good part), split internal/external steps; changes per shift; OEE availability | Die changeover time: -43% to -45%; OEE 70% after | Peer-reviewed study |
| Spaghetti diagram (travel mapping) | Draw on the floor plan every walk of one operator or one part for a shift; the tangle shows wasted travel to cut | Walking/travel distance per part or per shift (m); walking time share of cycle; number of trips | Total distance traveled: total distance traveled reduced by 61.2% (with lead time -23.66%, WIP -8.6%) (ET) | Peer-reviewed study |
| Standardised work | Write down the best known sequence, time and in-process stock for each job, post it at the station, and train everyone to it | Output per shift vs logged capacity, cycle time and its spread between operators/shifts, operator walking distance, line downtime, defects per shift | Output per 7 h shift (core making workstation): SW saved 31.6 s per cycle, production up to 58 pieces per 7 h shift (vs 45-50 actual, 54 capacity); productivity up to 6.5 percent | Peer-reviewed study |
| Statistical Process Control (SPC) | Plot key dimensions or defect counts on simple charts every shift so drift is caught before bad parts are made | Cpk of critical dimensions; rejection % or ppm per shift; number of out-of-control points per week | Cpk of drop impact resistance: Cpk index was increased from 0.48 to 1.79 (ID) | Peer-reviewed study |
| Supplier development | Buyer actively helps key suppliers fix their quality, delivery and cost (visits, training, joint kaizen) instead of only rating or switching them | Supplier incoming rejection PPM; supplier on-time-in-full %; line stoppages due to supplier parts; supplier lead time; cost of supplier-caused rework | Overall cost competitiveness profile (composite): Overall cost competitiveness profile of ACMA firms improved by 13.3% (IN) | Case study |
| Takt time planning | Work out how many seconds per piece the customer demand allows, and pace every station and staffing level to that beat | Line cycle time vs takt time, output per man-hour, overtime hours per month, schedule attainment % | Line cycle time vs takt: line CT was reduced by 16.0%, from 78.0 sec to 65.5 sec, managed to be lower than line takt time (MY) | Peer-reviewed study |
| Theory of Constraints (drum-buffer-rope) | Find the one machine that limits the whole plant, keep it running every minute, and pace everything else to it | Output of the bottleneck per shift; bottleneck idle minutes; on-time dispatch %; lead time; WIP | Lead time; inventory; due-date performance; revenue/throughput: lead time -69% (n=32); inventory -50% (n=28); due-date performance +60% (n=12); revenue/throughput +68% | Review of studies |
| Time and motion study | Watch and time each job step with a stopwatch, cut useless motions and waiting, and set a fair standard time per piece | Pieces per operator-hour; cycle time per operation; share of non-value-added time; standard vs actual time per piece | Overall productivity: 14% productivity increase; annual financial gain 21,84,000 (currency not stated, INR format) (IN) | Case study |
| TPM (Total Productive Maintenance) | Operators and maintenance together keep machines in shape so they stop breaking down and stop running slow | OEE split into availability x performance x quality; breakdown hours per month; MTBF/MTTR; rejection % | OEE; breakdown time; rejection; production: OEE +17.08%; breakdown time -23.14%; rejection -17.94%; production +15.63% (IN) | Peer-reviewed study |
| TWI Job Instruction | A 4-step way for supervisors to teach a job (prepare, show, try, follow up) so new operators get it right fast and the same way | Days for a new operator to reach standard output; first-pass yield / rejection per operator; output variation between operators and shifts; training hours per operator | Firm productivity vs non-trained applicants: +6% productivity within one year, rising to +27% after ten years (US) | Working paper |
| Value analysis / value engineering | Take your costliest parts, ask what each must do, and redesign material, shape or process to make it cheaper without losing function | Material + conversion cost per part (Rs/piece); weight per part; number of operations; margin on the product | Component cost reduction: rotor 30%, pump body 4%, stator housing 10% cost reduction (IN) | Peer-reviewed study |
| Value Stream Mapping | Draw how one order really moves through the plant, door to door, and see where it waits | Order lead time (days, door to door); WIP between each pair of stations; value-added time / lead time | Lead time; WIP; cost: lead time -27%; WIP -27% to -32%; INR 2,17,728 value-added cost saving (IN) | Peer-reviewed study |
| Vendor-managed inventory (VMI) | Supplier sees your stock and usage and refills to agreed min/max levels himself, instead of you raising every purchase order | Raw-material/packaging stock days; stock-outs or line stops for material per month; total inventory cost; supplier fill rate % | Downstream inventory level: VMI decision-transfer reduces inventory levels by 7% on average (US) | Peer-reviewed study |
| Visual management | Make the plan, actual output, problems and standards visible on boards and floor markings so anyone sees status at a glance | Hour-by-hour plan vs actual; OEE; number of problems logged and closed from boards; time to find tools/material | Improvement memos raised from visual control tables: 61 improvement memos, 39% fully implemented (24 months; bundle with SMED, 5S, TPM; OEE 87.9% to 92.3%) (RS) | Peer-reviewed study |
| Work sampling | Take many random snapshots of people and machines (working, waiting, idle) to see what share of the shift is really productive | % of observations productive for operators and machines; OEE; crew size per line/shift | OEE of production line: OEE 32% -> 56% (+24 points); workers productive 50.2% of time at baseline (2,200 observations) | Case study |
| Workstation ergonomics | Redesign work height, reach, jigs, lifting and layout so operators work faster with less strain, fatigue and injury | Pieces per operator-hour; cycle time per piece; rejection %; musculoskeletal complaints / injury days; ergonomic risk score (RULA/REBA/OCRA) | Productivity: +16% productivity and -22.7% upper-limb overload risk (OCRA) in about two years (IT) | Peer-reviewed study |