6. Lean Management and Flow Improvement

Introduction and context

Lean management is a disciplined approach to creating customer value with fewer resources, less delay, lower variation, and minimal waste. It is often associated with manufacturing, particularly the Toyota Production System, but its underlying logic is equally relevant to hospitals, banks, logistics, software development, public administration, education, and professional services. Lean is not simply a collection of tools for reducing cost. It is a management system that links purpose, process design, employee involvement, problem solving, and continuous improvement.

The central operational problem addressed by Lean is the separation between value-creating work and the many activities surrounding it. Customers generally pay for an outcome, not for waiting, movement, rework, excess inventory, repeated data entry, or internal approvals. Lean therefore asks a fundamental question: what must happen to deliver the required product or service, and how can that work flow smoothly at the rate of customer demand?

Flow improvement is particularly important in Total Quality Management because quality and flow are mutually reinforcing. Poor flow hides defects in queues and batches; excessive work-in-process increases the time before problems are discovered; and unstable processes encourage expediting rather than prevention. Conversely, a stable, visible, and well-designed flow makes abnormal conditions easier to detect and correct.

Lean should not be confused with indiscriminate downsizing. Removing people or equipment without redesigning the process can increase workload, risk, and defects. Genuine Lean improvement removes non-value-adding work while strengthening process capability, respect for people, and customer outcomes.

Learning objectives

After studying this chapter, the learner should be able to:

1. Explain the principles, assumptions, and evolution of Lean management.

2. Identify and analyse the eight forms of waste in manufacturing and service processes.

3. Calculate and interpret takt time, cycle time, lead time, and work-in-process.

4. Design appropriate pull, Kanban, cellular, and standard-work arrangements.

5. Apply 5S, visual management, SMED, poka-yoke, and gemba practices.

6. Evaluate Lean interventions in relation to quality, capacity, people, technology, and risk.

7. Develop an integrated flow-improvement plan for a real or illustrative organisation.

From production efficiency to a management system

Lean thinking is commonly expressed through five principles: define value from the customer’s perspective; identify the value stream; create flow; establish pull; and pursue perfection through continuous improvement. These principles are connected rather than sequential checkboxes. A process cannot be made to flow effectively if value is poorly defined, and pull cannot function reliably if the process is unstable.

Value is the capability or outcome for which the customer is willing to exchange resources. In a hospital, value may include accurate diagnosis and timely treatment; in an insurance office, correct and prompt claim settlement; in a factory, conformity to specification and dependable delivery. Some activities are necessary but not directly value-adding, such as regulatory checks or preventive maintenance. They should not automatically be eliminated; instead, they should be simplified, reduced, or integrated into the flow.

The value stream includes all actions, information, decisions, and material movements required to deliver a product or service. Value-stream mapping makes the relationship between processing time, waiting time, inventory, information delays, and quality problems visible. A common finding is that the proportion of elapsed time spent on actual value-adding work is small. The managerial response should be process redesign, not merely pressure on employees to work faster.

Flow means that work moves through a process with minimal interruption, backtracking, batching, and waiting. One-piece flow is an ideal in which items are processed and transferred individually or in small economical quantities. It is not always practical, especially where changeovers, transport distances, or process economics make larger batches necessary. The relevant question is whether batch size is justified by a real constraint or is simply an inherited convention.

Pull means that upstream activities respond to actual downstream consumption rather than producing speculatively. A pull system limits work-in-process and exposes instability. It requires reliable information, clear replenishment rules, and agreed capacity. A system that is called pull but is constantly overridden by urgent requests is not functioning as a disciplined pull system.

Perfection is pursued through kaizen: structured, repeated improvement involving the people who understand the work. Kaizen may involve a small workplace adjustment, a cross-functional redesign, or a strategic investment. The scale differs, but the logic is the same: understand the current condition, identify the gap, test a countermeasure, verify results, and standardise what works.

Seeing waste in the value stream

The traditional Lean vocabulary identifies seven wastes; many contemporary treatments add an eighth, unused human capability. The mnemonic TIMWOODS is often used: transportation, inventory, motion, waiting, overproduction, overprocessing, defects, and skills or unused talent.

Transportation is unnecessary movement of materials, documents, patients, information, or products. It does not change the item’s value and can create damage, loss, or delay. In a service process, transferring a customer file between departments is transportation in informational form.

Inventory includes raw materials, work-in-process, finished goods, queues, unresolved cases, and excess information awaiting review. Inventory may protect a process from disruption, but it also conceals defects, consumes space, increases obsolescence, and extends lead time. Lean does not treat every stockholding as waste; it asks whether its quantity and location are justified by demand and process capability.

Motion is unnecessary physical or mental movement by employees. Searching for tools, repeated login procedures, poor workstation design, and walking to obtain approvals are examples. Motion differs from transportation because transportation concerns the movement of the item being processed, while motion concerns the movement of the worker or user.

Waiting occurs when people, materials, decisions, information, or equipment remain idle. Typical causes include unbalanced work, machine downtime, missing information, approval layers, and large batches. Waiting is often the largest contributor to service lead time.

Overproduction occurs when work is completed earlier, faster, or in greater quantity than required. It is particularly harmful because it generates other wastes, including inventory, handling, storage, and defects. Printing reports that no one reads and producing components against an unreliable forecast are examples.

Overprocessing involves effort beyond customer or regulatory requirements. Examples include duplicate data entry, excessive precision, unnecessary signatures, repeated inspections caused by poor process control, and reports containing unused detail. Overprocessing should not be confused with essential quality assurance.

Defects include incorrect products, inaccurate data, missed appointments, billing errors, unsafe service, and failures to meet specifications. Defects generate rework, complaints, returns, investigation, and loss of trust. Poka-yoke and built-in quality seek to prevent or immediately detect such errors.

Unused human capability occurs when employees’ knowledge, creativity, judgement, and improvement ideas are ignored. It may also arise when people are placed in roles that do not use their competence or when fear prevents them from reporting problems. This waste is strategically serious because process improvement depends on distributed knowledge.

Waste identification must be evidence-based. A process walk, direct observation, queue analysis, defect data, and interviews are more reliable than assumptions. Lean practitioners distinguish symptoms from causes. For example, a large queue is visible waste, but its causes may include uneven demand, an approval bottleneck, unreliable equipment, or poorly defined priorities.

Measuring demand and flow

Flow improvement requires a common operational language. Takt time is the available production time divided by customer demand during that period:

Takt time = Net available operating time / Customer demand

If a service centre has 420 productive minutes per day and receives 70 cases, takt time is six minutes per case. Takt is a demand rhythm, not a target for forcing employees to work unsafely. It must be adjusted for breaks, planned maintenance, realistic attendance, and the mix and complexity of work.

Cycle time is the time required to complete one unit of work at a particular process step. If cycle time is consistently greater than takt time, the step cannot meet demand without additional capacity, work redistribution, simplification, or demand adjustment. If cycle time is much shorter than takt, the resource may be capable of supporting other work, although variability and changeovers must be considered.

Lead time is the elapsed time from request to delivery. It includes processing, transport, waiting, rework, and queues. Throughput is the rate at which completed units leave the process. Work-in-process (WIP) is work started but not yet completed. Little’s Law describes the relationship:

WIP = Throughput × Lead time

The relationship is not a promise that reducing WIP automatically improves every process; it is a diagnostic identity under stable conditions. If throughput remains constant, reducing WIP should reduce lead time. If WIP is reduced without addressing a bottleneck, throughput may fall and customer service may deteriorate.

Capacity analysis should consider effective rather than theoretical capacity. Availability, performance, quality losses, changeovers, absenteeism, skill constraints, and demand variation all affect actual output. Lean therefore combines time measures with quality and reliability measures rather than optimising speed alone.

Designing pull and Kanban systems

A pull system uses a signal to authorise production, movement, or replenishment. Kanban is one method of implementing pull. A Kanban may be a card, container, electronic signal, barcode event, or status indicator. Its function is to communicate what is needed, where it is needed, in what quantity, and under what replenishment rule.

A basic Kanban loop consists of a downstream point consuming material, returning a signal, and an upstream point replenishing the consumed quantity. The number of Kanban units can be estimated using:

Number of Kanban containers = Demand during replenishment lead time × (1 + safety factor) / Container quantity

The safety factor should reflect variability and risk, not become an excuse for uncontrolled stock. Kanban parameters must be reviewed when demand, lead time, container size, or process reliability changes.

Kanban is most effective when processes are reasonably stable, quality is assured at the source, replenishment routes are dependable, and rules are visible. It is not a substitute for solving chronic downtime, inaccurate inventory records, or unstable suppliers. In service settings, virtual Kanban boards can limit the number of open cases, visualise status, and prevent overloading specialists.

A push system schedules work based primarily on forecasts or upstream plans. Push planning remains useful for long-lead materials, capacity planning, and environments with predictable production constraints. Lean practice often combines push and pull: a forecast may guide procurement, while actual consumption triggers final assembly or service completion.

Stabilising work through 5S, standard work, and visual management

5S is a workplace-organisation method: sort, set in order, shine, standardise, and sustain. Sorting removes unnecessary items. Setting in order locates required items so that they can be found and returned easily. Shining includes cleaning and inspection, allowing abnormal conditions such as leaks, wear, or contamination to be noticed. Standardising defines the expected arrangement and routine. Sustaining involves discipline, leadership, audits, and improvement rather than occasional clean-up campaigns.

5S is often weakened when it is treated as housekeeping alone. Its deeper purpose is to make the normal condition visible and abnormalities immediate. In a digital office, 5S may involve eliminating obsolete files, defining naming conventions, organising shared folders, reducing duplicate applications, and establishing clear ownership of records.

Standard work specifies the current best-known method for performing a task safely, correctly, and consistently. It usually includes takt time or expected pace, work sequence, and standard in-process stock. Standard work is not intended to suppress professional judgement. It provides a baseline from which deviations and improvements can be understood. A standard that is impossible to follow, excessively detailed, or disconnected from actual conditions will not create control.

Visual management places critical information at the point of use. Examples include production boards, colour-coded locations, response lights, escalation signals, service queues, and simple performance charts. Effective visual management answers three questions: What should be happening? What is actually happening? What action is required? Displays that show data but do not trigger ownership or action are decoration rather than management.

Quick changeover and flexible flow

Single-Minute Exchange of Die (SMED) is an approach to reducing changeover time. Its logic is broader than a literal one-minute target. First, observe and document the changeover. Next, separate internal activities, which require the equipment to be stopped, from external activities, which can be completed while it is running. Convert internal activities to external where possible, then simplify, parallelise, and standardise the remaining work.

Examples include preparing tools in advance, using quick-release fasteners, presetting settings, employing carts, and assigning parallel tasks. Reduced changeover time permits smaller batches, which in turn reduce inventory and improve responsiveness. However, changeover reduction must not compromise safety, verification, or first-off quality approval.

Flexible flow also depends on cellular design. A functional layout groups similar equipment or expertise together; a cellular layout arranges resources needed for a product family or service family in a sequence that supports flow. U-shaped cells can reduce walking and make staffing flexible, but layout alone does not create flow. Product or case similarity, balanced work, reliable equipment, standard work, and clear replenishment rules are necessary.

In services, a cell may be a cross-trained team responsible for a complete customer segment, such as mortgage applications or discharge planning. The benefits include fewer handoffs, clearer accountability, and faster feedback. The risks include underutilisation if demand is highly uneven and loss of specialist depth if cross-training is poorly planned.

Building quality into the process

Lean flow without quality is merely fast failure. Poka-yoke, or mistake-proofing, prevents an error or makes it immediately apparent. A fixture that permits assembly in only one orientation, a software field that validates a required format, and a barcode match between a patient and medication are examples. Stronger designs prevent the error; weaker designs merely warn after it occurs.

Source inspection verifies quality where the work is performed rather than relying exclusively on end-of-line inspection. When an abnormality occurs, the process should be stopped or contained according to risk, the condition made visible, and the cause investigated. The objective is not to blame the operator but to learn why the system permitted the error.

Lean problem solving commonly uses the Plan–Do–Check–Act cycle. The team defines the problem and current condition, analyses causes using observation and evidence, tests countermeasures, checks results against a baseline, and standardises successful changes. The five-whys technique can support causal inquiry, but it should not be used mechanically or as a substitute for data. Complex problems may require process mapping, Pareto analysis, statistical methods, or designed experiments.

Gemba and the social side of Lean

Gemba means the actual place where work and value creation occur. A gemba walk is a structured visit to observe the process, ask respectful questions, and understand conditions directly. It is not a management inspection intended to catch employees doing something wrong. Useful questions include: What is the standard? What makes the work difficult? Where does work wait? What abnormality occurred recently? What support is needed?

Leadership is essential because Lean changes decision rights, measures, routines, and sometimes organisational boundaries. Leaders must protect time for improvement, respond to reported problems, and avoid rewarding overproduction or heroics. Employees need training in problem solving and assurance that identifying a problem will not result in punishment. Respect for people is therefore both an ethical principle and a practical requirement for obtaining accurate process knowledge.

Lean metrics should be balanced. Useful measures may include first-pass yield, lead time, on-time delivery, WIP, changeover time, safety incidents, employee suggestions implemented, and customer satisfaction. A single productivity metric can create dysfunctional behaviour. For example, measuring only cases completed per hour may encourage premature closure, omission of quality checks, or shifting difficult work to another department.

Illustrative case: improving outpatient diagnostic flow

Background

The following is a fictional illustrative case. A regional hospital’s outpatient diagnostic unit performed blood tests and imaging referrals for several clinics. Patients reported long and unpredictable waits even though the average diagnostic processing time was short.

Situation and analysis

A cross-functional team mapped the patient and information flow. Patients registered at one desk, waited for clinical verification, moved to a specimen station, waited again for transport, and later returned to a separate desk for results clarification. The current-state map showed that actual processing occupied approximately 25 minutes, while the average elapsed time exceeded two hours. The main wastes were waiting, transportation, overprocessing through duplicate data entry, and defects involving incomplete requisitions.

Demand data showed a morning peak. Registration and specimen collection had different staffing patterns, creating an early bottleneck. A Pareto analysis found that a small number of missing fields caused a large share of rework. Observation at the gemba also revealed that frequently used labels and supplies were stored away from the collection point.

Approach

The team established a future-state design. A standard electronic requisition with mandatory fields reduced incomplete requests. Registration and clinical verification were combined for low-risk tests. Supplies were reorganised using 5S, and a visual replenishment signal was introduced. A small cross-trained cell handled registration, verification, and specimen preparation during the morning peak. A Kanban limit was placed on pending clarification cases so that unresolved requests did not accumulate invisibly.

The team defined standard work for specimen identification and introduced barcode verification as a poka-yoke. A daily visual board displayed waiting time, rejected specimens, pending clarifications, and staffing status. Supervisors conducted short gemba reviews rather than relying only on end-of-day reports.

Outcome

In the illustrative scenario, average patient lead time fell substantially, incomplete requisitions declined, and staff reported fewer interruptions caused by searching and repeated clarification. The improvement was not attributed to working faster. It resulted from removing duplicate work, balancing capacity with demand, preventing errors at source, and making abnormal flow visible. The team continued to monitor safety, patient experience, and staff workload to ensure that speed did not displace care quality.

Learning points

The case demonstrates that Lean in services must follow the entire customer journey, not merely the activity of one department. Takt-based staffing can address demand peaks, but it must be combined with standard work and cross-training. Visual controls are useful only when linked to action. Finally, technology produced value because it simplified and mistake-proofed the process; digitising the old sequence without redesign would have automated waste.

Risks, limitations, and professional considerations

Lean initiatives fail when leaders launch isolated tools without a problem definition. A 5S audit cannot solve an unstable supplier, and a Kanban card cannot solve an unreliable process. Other common mistakes include reducing inventory before improving quality, setting takt below safe human capability, treating standard work as rigid surveillance, and using kaizen events to compensate for absent strategic investment.

Local optimisation is another risk. A department may improve its own output while increasing queues or rework elsewhere. Value-stream measures should therefore cross organisational boundaries. Similarly, Lean may be misused as a euphemism for staff reduction. Such use undermines trust and deprives the organisation of the knowledge needed for improvement.

Ethically, flow redesign must protect safety, dignity, privacy, accessibility, and fair treatment. In healthcare and other regulated settings, compliance checks cannot be eliminated merely because they appear non-value-adding. Personal data used in digital Kanban or performance systems must be governed appropriately. Automated prioritisation should be explainable, tested for bias, and subject to human oversight when decisions affect rights or wellbeing.

Technology can strengthen Lean through electronic Kanban, real-time process mining, barcode traceability, digital work instructions, sensors, and predictive maintenance. Process mining can reveal hidden queues and rework loops from event logs. Artificial intelligence may support demand forecasting, anomaly detection, scheduling, and knowledge retrieval. Nevertheless, poor process definitions and inaccurate data will produce misleading digital insights. Automation should follow process understanding, not replace it.

Current developments and best practices

Contemporary Lean practice increasingly addresses end-to-end ecosystems rather than isolated factories. Lean principles are being applied to omnichannel fulfilment, remote services, software delivery, sustainability, and administrative processes. Environmental waste, energy use, emissions, and resource circularity are being incorporated into value-stream analysis. Remote and hybrid work also require visualising digital queues, handoffs, interruptions, and knowledge work.

Best practice is to begin with a clear customer and organisational problem, observe the current condition, involve process users, and establish a measurable baseline. Improvements should be tested at small scale where appropriate, checked for unintended effects, and built into standard work. Leaders should develop capability rather than merely demand results. A mature Lean system combines daily management, structured problem solving, respect for people, and strategic alignment.

Chapter Summary

Lean management is a system for creating customer value through stable, visible, responsive, and continually improving processes. Its five principles are value, value-stream understanding, flow, pull, and perfection. The eight wastes—transportation, inventory, motion, waiting, overproduction, overprocessing, defects, and unused human capability—provide a lens for locating non-value-adding activity.

Takt time connects capacity to customer demand, while cycle time, lead time, throughput, and WIP help diagnose flow. Pull and Kanban regulate replenishment; 5S and visual management expose abnormal conditions; standard work establishes a reliable baseline; SMED supports smaller batches; cellular design reduces handoffs; poka-yoke builds quality into the process; and gemba connects leadership with actual work. Successful Lean application requires employee involvement, balanced measures, ethical governance, and careful use of technology. Lean is not a cost-cutting campaign but a way of managing work, learning, and improvement.

Key Terms

Takt time: Available productive time divided by customer demand; the rhythm required to meet demand.

Cycle time: Time required to complete one unit at a particular process step.

Lead time: Elapsed time from customer request to delivery, including processing and waiting.

Throughput: Rate at which completed units leave a process.

Work-in-process: Units or cases started but not yet completed.

Lean management: A management system focused on customer value, flow, pull, waste reduction, and continuous improvement.

Value stream: The complete set of activities and information flows required to deliver an outcome.

Kaizen: Continuous, participative improvement through repeated problem solving.

Pull system: A system in which downstream consumption authorises upstream replenishment or production.

Kanban: A visual or electronic signal controlling movement or replenishment.

5S: Workplace-organisation method involving sort, set in order, shine, standardise, and sustain.

Standard work: The current best-known safe and effective method for performing a task.

Visual management: Display of information so normal and abnormal conditions are readily understood.

SMED: Method for reducing equipment or process changeover time.

Cellular design: Arrangement of resources in a sequence or unit serving a product or service family.

Poka-yoke: Mistake-proofing that prevents or immediately detects errors.

Gemba: The actual place where work and value creation occur.

Waste: Activity that consumes resources without creating required customer value.

Little’s Law: Relationship stating that WIP equals throughput multiplied by lead time.

Review Questions

Short Answer Questions

1. Distinguish takt time from cycle time.

2. Identify the eight wastes recognised in Lean management.

3. What is the purpose of a Kanban signal?

4. Explain why standard work is considered a baseline for improvement.

5. What is meant by a gemba walk?

Long Answer Questions

1. Explain the five principles of Lean and show how they are interdependent.

2. Discuss how 5S, visual management, and standard work contribute to process stability.

3. Explain SMED and cellular design, including their effects on batch size and flow.

4. Discuss the role of respect for people and leadership in sustaining Lean management.

5. Analyse the benefits and limitations of applying Lean in service organisations.

Analytical / Application Questions

1. A claims department has 480 productive minutes per day and receives 80 cases. Calculate takt time and explain what additional information is needed before staffing decisions are made.

2. A process has throughput of 12 cases per hour and average WIP of 60 cases. Use Little’s Law to estimate lead time and recommend two questions for further diagnosis.

3. Design an initial Lean improvement plan for a university admissions office experiencing duplicate data entry, long queues, incomplete applications, and frequent status enquiries. Specify the wastes, measures, and tools you would use.

22. Multiple Choice Questions with Answers

Question 1

A service centre has 420 productive minutes available each day and demand for 70 cases. What is its takt time?

A. 3 minutes per case

B. 6 minutes per case

C. 70 minutes per case

D. 490 minutes per case

Correct Answer: B. 6 minutes per case

Question 2

Which situation is the clearest example of overproduction?

A. A worker walks repeatedly to obtain tools

B. A report is produced before it is required and is never used

C. A customer waits for an approval

D. A product is assembled incorrectly

Correct Answer: B. A report is produced before it is required and is never used

Question 3

In a pull system, what normally authorises upstream replenishment?

A. Maximum theoretical machine capacity

B. A supervisor’s informal preference

C. Downstream consumption communicated through a signal

D. Production of the largest possible batch

Correct Answer: C. Downstream consumption communicated through a signal

Question 4

Which statement best describes standard work?

A. A permanent method that may never be changed

B. The current best-known safe and effective method used as a baseline for improvement

C. A document used only for regulatory audits

D. A method for calculating inventory value

Correct Answer: B. The current best-known safe and effective method used as a baseline for improvement

Question 5

A barcode system prevents a medication from being matched to the wrong patient. This is primarily an example of:

A. Poka-yoke

B. SMED

C. Overproduction

D. Cellular design

Correct Answer: A. Poka-yoke

Question 6

A process has throughput of 10 cases per hour and average WIP of 40 cases. Under stable conditions, what is the approximate lead time?

A. 0.25 hours

B. 4 hours

C. 30 hours

D. 400 hours

Correct Answer: B. 4 hours

Question 7

Which SMED action converts an internal changeover activity into an external one?

A. Performing all adjustments after the machine stops

B. Preparing and presetting tools while the equipment is still running

C. Increasing the production batch size

D. Adding a final inspection after changeover

Correct Answer: B. Preparing and presetting tools while the equipment is still running

Question 8

Why can excessive inventory conceal quality problems?

A. It eliminates all variation from the process

B. It allows defects to accumulate without immediate detection

C. It guarantees one-piece flow

D. It reduces the need for material handling

Correct Answer: B. It allows defects to accumulate without immediate detection

Question 9

What is the primary purpose of a gemba walk?

A. To rank employees by individual speed

B. To observe actual work and understand conditions directly

C. To replace all process data with managerial opinion

D. To approve larger production batches

Correct Answer: B. To observe actual work and understand conditions directly

Question 10

A hospital digitises an inefficient paper approval sequence without removing duplicate approvals. Which Lean lesson does this illustrate?

A. Automation always creates flow

B. Digitising a process can preserve or automate existing waste

C. Pull systems require no process stability

D. Visual management is unnecessary in services

Correct Answer: B. Digitising a process can preserve or automate existing waste