TPM & Visual Factory
Key Takeaways
- Total Productive Maintenance (TPM) maximizes equipment effectiveness through shared ownership by operators and maintainers, aiming to cut losses from breakdowns, setup, speed, defects, and other wastes.
- TPM pillars commonly include autonomous maintenance, planned maintenance, focused improvement, training, early equipment management, quality maintenance, safety/environment, and office TPM; predictive maintenance uses condition data to service before failure.
- Visual factory makes process status, standards, and abnormalities obvious at a glance so problems are detected and corrected quickly.
- Andon provides visual (and often audible) alerts that stop or call attention to abnormalities; Jidoka (autonomation) builds in the ability to detect defects and stop rather than pass them on.
- TPM and visual controls sustain DMAIC improvements by keeping equipment reliable and process conditions visible so control plans and reaction plans can work.
TPM & Visual Factory (CSSGB BoK VI.C.1–2)
Quick Answer: Total Productive Maintenance (TPM) improves equipment effectiveness through operator–maintainer partnership, planned and autonomous care, and techniques such as predictive maintenance. A visual factory makes status and abnormalities obvious using tools like Andon and principles like Jidoka (detect and stop). Together they help control improved processes by reducing unplanned downtime and making out-of-standard conditions impossible to ignore.
Even a perfect control plan fails if machines are unreliable or problems stay invisible until scrap piles up. VI.C covers lean control tools that keep the physical and visual system capable of holding gains.
Total Productive Maintenance (TPM) — VI.C.1
TPM is a system for maximizing the productive effectiveness of equipment through company-wide participation—especially by involving operators in routine care traditionally left only to maintenance specialists. The classic aim is to minimize the major losses that steal available capacity and quality.
Why TPM Belongs in Control
DMAIC improvements often assume stable equipment. If a critical machine suffers frequent breakdowns, the process reintroduces variation, queues, expediting, and defects. TPM protects the capability and flow the project worked to create.
Equipment Losses TPM Targets
Organizations often group losses such as:
- Breakdowns — unplanned stoppages
- Setup and adjustment — long changeovers
- Idling and minor stops — small interruptions that add up
- Speed losses — running below design speed
- Defects and rework — quality losses tied to equipment condition
- Startup losses — yield loss after start or changeover
Reducing these losses raises Overall Equipment Effectiveness (OEE) conceptually (availability × performance × quality), even when the exam focuses more on TPM elements than on OEE arithmetic.
Core TPM Elements / Pillars (Conceptual Map)
Exact pillar lists vary by source; CSSGB expects the ideas, not a single corporate logo set:
| Element | Intent |
|---|---|
| Autonomous maintenance | Operators clean, inspect, lubricate, and detect abnormalities early (often after training and standards) |
| Planned maintenance | Scheduled preventive work based on time, usage, or risk—not only firefighting |
| Focused improvement (kobetsu kaizen) | Cross-functional attacks on chronic losses |
| Training & education | Skills for operators and maintainers to own equipment health |
| Early equipment management | Design maintainability and reliability into new equipment |
| Quality maintenance | Prevent defects by controlling equipment conditions that create quality problems |
| Safety, health & environment | Zero accidents and environmental incidents as part of equipment care |
| Office / administrative TPM | Apply loss reduction to support processes that affect equipment and flow |
Green Belts should recognize autonomous maintenance and planned maintenance as central, and know that TPM is proactive culture—not only a work-order system.
Predictive Maintenance
Predictive maintenance uses condition-monitoring data to service equipment before failure, based on actual health rather than fixed calendar intervals alone.
Examples of condition signals:
- Vibration analysis on rotating equipment
- Thermography (infrared) for hot connections or bearings
- Oil analysis for wear metals and contamination
- Ultrasonic leak detection
- Motor current or other sensor trends in modern plants
Contrast the maintenance modes:
| Mode | Trigger | Characteristic |
|---|---|---|
| Reactive (breakdown) | After failure | High downtime risk; emergency repairs |
| Preventive (time/use-based) | Calendar or cycle count | Reduces surprise failures; may replace parts early |
| Predictive (condition-based) | Measured degradation trend | Targets intervention when data show need |
Predictive methods often sit inside a broader TPM/planned-maintenance strategy. They do not replace operator autonomous care; dirty, uninspected machines still fail in ways sensors may miss.
TPM and the Control Plan
Link equipment CTQs to the control system:
- Critical process parameters that depend on machine condition (temperature uniformity, torque capability, seal integrity)
- Maintenance tasks and frequencies as process supports
- Reaction plans for equipment alarms and Andon stops
- Spare-part and setup standards that prevent “heroic” temporary fixes from becoming the new normal
Visual Factory — VI.C.2
A visual factory (visual workplace / visual management) designs the environment so that normal vs abnormal is obvious within seconds—without hunting through computer screens or asking a supervisor.
Goals include:
- Make standards visible at the point of use
- Show real-time process status (running, stopped, behind, quality hold)
- Expose waste, shortages, and safety risks immediately
- Enable faster reaction aligned with the control plan
Everyday Visual Tools
- Shadow boards and labeled locations for tools (support 5S)
- Floor markings for walkways, WIP locations, and min/max inventory
- Color coding for status, material type, or priority
- Performance boards with hourly production and quality results
- Standard work charts and photos of correct setups
- Limit samples showing acceptable vs reject product appearance
Visual management is not decoration. If a board is weeks out of date, it is noise—not control.
Andon
Andon is a visual (and often audible) signaling system that announces abnormalities and requests help. Classic forms include light stacks (green/yellow/red), boards showing line status, and digital dashboards.
Typical uses:
- Operator pulls a cord or presses a button when a problem appears
- Light turns red; team lead or support responds
- Line may stop or a station may stop until the issue is resolved
- Problem type may be coded (quality, material shortage, equipment)
Andon supports control by making problems visible now, enabling the reaction plan instead of hiding defects until final inspection. For the exam, associate Andon with immediate visual alert and response, not with annual performance reviews.
Jidoka (Autonomation)
Jidoka, often translated as autonomation or “automation with a human touch,” means building into the process the ability to detect abnormalities and stop rather than continue producing defects.
Key ideas:
- Separate human intelligence from pure machine motion: machines should not mindlessly make bad parts
- Detect the problem at the source
- Stop and fix the condition (or alert humans via Andon)
- Prevent passing defects downstream
- Enable root-cause correction so the same stop does not endlessly recur
Jidoka and Andon work together: Jidoka is the principle of detect-and-stop; Andon is a common signaling mechanism that supports that principle on the line.
Examples:
- A sensor rejects and stops the machine when a dimension is out of tolerance
- Software blocks the next screen if a required field fails a validation rule (transactional jidoka)
- A poka-yoke fixture that will not close on a missing component, triggering an alert
How Visual Factory Controls the Improved Process
After DMAIC Improve:
- Standards from the control plan are posted and obvious.
- SPC status or simple run boards show whether the process is on track.
- Andon escalates special causes immediately.
- Jidoka and poka-yoke prevent mass production of defects.
- TPM keeps equipment able to hold the improved capability.
- Abnormal conditions invite PDCA and documented reaction—not silent workarounds.
Without visual control, supervisors discover problems late; without TPM, even visible standards cannot be met because the machine cannot hold settings.
Integrating TPM, Visual Factory, and DMAIC Control
| Control need | TPM contribution | Visual factory contribution |
|---|---|---|
| Stable process capability | Reliable equipment, fewer breakdown-induced defects | Visible standards and status |
| Fast reaction | Alarms, condition monitoring, maintenance response | Andon, boards, color status |
| Prevention | Quality maintenance, predictive interventions | Jidoka, poka-yoke visibility |
| Sustainment culture | Operators own basic equipment care | Everyone can see normal vs abnormal |
Exam Focus
- Define TPM purpose: maximize equipment effectiveness via organization-wide participation and loss reduction.
- Recognize autonomous vs planned vs predictive maintenance ideas.
- Define visual factory purpose: abnormalities obvious at a glance.
- Associate Andon with visual/audible alert and help/stop response.
- Associate Jidoka with detect abnormality and stop (autonomation), preventing defect flow downstream.
- Connect both topics to sustaining improved processes in Control—not only to “lean vocabulary.”
Key takeaway: TPM keeps equipment capable; visual factory (Andon, Jidoka, and related signals) keeps problems visible and stoppable—together they control and sustain the process DMAIC improved.
Which practice best illustrates predictive maintenance within a TPM approach?
A sensor detects an out-of-spec fill volume, stops the filler, and lights a red Andon so the team lead responds immediately. Which pair of concepts is primarily demonstrated?
You've completed this section
Continue exploring other exams