Enhancing Operational Efficiency Through TPM and Lean Six Sigma

The 8 Deadly Wastes Black Belt Training Module #2 Session #3 Analyze December 16, 2024

Agenda Total Productive Maintenance (TPM) Lean Layout Design Design of Experiments Hypothesis Testing

TPM Total Productive Maintenance Total Productive Maintenance (TPM) is a proactive approach to equipment maintenance that seeks to maximize the operational efficiency of equipment by involving all employees in maintaining the equipment Goal: To improve equipment reliability and uptime, reduce downtime and costs, and increase overall efficiency through a holistic maintenance strategy TPM aligns with Lean's focus on eliminating waste by ensuring that equipment is running efficiently, reducing breakdowns, and maximizing production capacity.

TPM & Lean Six Sigma Eliminating Equipment Downtime: TPM helps identify and resolve underlying issues that cause breakdowns, directly reducing downtime and waste Improving OEE: TPM contributes to higher Overall Equipment Effectiveness (OEE) by improving availability, performance, and quality Supporting Continuous Improvement: TPM emphasizes teamwork, collaboration, and continuous improvement to achieve better equipment reliability and efficiency Reducing Costs: By preventing equipment failures and improving efficiency, TPM leads to lower maintenance costs, less scrap, and reduced operational disruptions

8 Pillars of TPM Safety, Health & Environment Autonomous Maintenance Education and Training Planned Maintenance Quality Maintenance Focused Approach Early Management Office TPM

8 Pillars of TPM Early Management Applying knowledge from other pillars (Autonomous Maintenance, Focused Improvement, etc.) to minimize downtime Autonomous Maintenance Enables Operators to fulfill duties normally performed by Maintenance (cleaning, inspection, lubrication, vibration monitoring, etc.) Intention is to detect faults and issues early so minor issues can be corrected before they turn into large problems Quality Maintenance Aims to reduce quality defects and eliminate the need for re-work to correct mistakes and defects Focused Improvement Empowering teams with Kaizen Events to quickly address problems related to machinery and equipment Safety, Health & Environment Create a workplace free of hazards and safety risks, a healthy work environment, includes health and safety metrics tracking Planned Maintenance Scheduling preventative maintenance activities for when a machine is not scheduled for production Office TPM Includes other aspects of the business outside of the shop or plant floor like supply chain, HR, finance, etc. Education and Training Providing continuing education for operations, maintenance and management

Autonomous Maintenance Autonomous Maintenance involves operators in the upkeep of equipment, allowing them to perform basic maintenance tasks like cleaning, lubrication, and small adjustments. Activities: Cleaning and inspecting machines Performing lubrication and minor adjustments Identifying wear and tear before it becomes a failure Benefits: Increased operator engagement, reduced downtime, and early identification of potential problems

Focused Improvement Focused improvement targets specific issues causing equipment inefficiencies and addresses them through problem-solving and continuous improvement efforts. Activities: Root cause analysis of frequent breakdowns or performance losses Implementing countermeasures to address recurring problems Benefits: Targeted solutions, improved equipment performance, and a culture of continuous improvement

Planned Maintenance Planned maintenance focuses on scheduling regular maintenance activities to prevent unplanned breakdowns and extend equipment lifespan. Activities: Setting up a preventive maintenance schedule based on manufacturer recommendations or historical data Using condition-based monitoring (vibration analysis, temperature checks, etc.) Benefits: Improved equipment reliability, reduced breakdowns, and predictable maintenance costs

Education & Training Providing ongoing training to all employees to ensure they are equipped with the knowledge and skills to perform effective maintenance and support TPM initiatives. Activities: Offering training on maintenance best practices, problem-solving, and equipment operation Cross-training operators to perform basic maintenance tasks Benefits: Empowered workforce, reduced dependency on specialized maintenance staff, and faster issue resolution

Early Management Early equipment management involves designing and selecting equipment with a focus on long-term reliability and ease of maintenance. Activities: Involving maintenance teams in equipment design and procurement Ensuring new equipment is easy to operate, maintain, and service Benefits: Increased equipment reliability, reduced downtime, and better overall equipment performance

Quality Maintenance Quality Maintenance ensures that machines are kept in optimal condition to support high-quality production processes and minimize defects caused by faulty equipment. Activities: Regular checks for precision and calibration Maintenance to prevent equipment malfunction that could affect product quality Benefits: Higher product quality, reduced defects, and less rework

Safety, Health & Environment TPM incorporates safety, health, and environmental considerations into maintenance practices to ensure equipment is safe to operate and minimizes environmental impact. Activities: Conducting regular safety audits of equipment Identifying and addressing health and safety hazards related to equipment operation Ensuring compliance with environmental standards during equipment operation Benefits: Safer work environment, reduced accidents, and improved compliance with safety regulations

Office TPM TPM principles are applied not just to physical equipment but also to administrative and support functions to improve overall efficiency and effectiveness. Activities: Streamlining administrative processes related to maintenance and equipment management Improving communication and collaboration between maintenance teams and other departments Benefits: More efficient administration, reduced delays in maintenance scheduling, and better alignment between production and maintenance teams

Lean Layout Design Flow Efficiency: Proper layout design reduces the time spent moving materials and personnel, leading to faster cycle times Waste Reduction: A Lean Layout eliminates waste related to transportation, motion, inventory handling, and waiting Improved Communication: Effective layouts facilitate communication between workstations and improve team collaboration Employee Engagement and Safety: A well-designed workspace improves ergonomics, reduces clutter, and promotes a safe work environment Flexibility: Lean Layouts support flexible production, allowing quick changes to adapt to different products or production volumes

Lean Layout Design

Lean Layout Design Principles Minimize Transportation: Reduce unnecessary movement of materials and people; this can be achieved through proximity and optimal arrangement of workstations Optimize Workflow: Design the layout to ensure a smooth and continuous flow of materials and information, minimizing interruptions or delays Simplify Processes: Remove unnecessary steps and activities that don t add value to the product or service Minimize Inventory: Use Just-In-Time (JIT) to reduce the need for large inventories, which saves space and reduces waste Design for Flexibility: Ensure that the layout can adapt to changes in product types, volumes, or processes Standardization: Implement standard work practices and station setups to reduce variation and improve efficiency 1. 2. 3. 4. 5. 6.

Lean Layout Tools Value Stream Mapping (VSM): Identifies Value Added and Non-Value Added activities in the flow of materials and information 5S (Sort, Set in Order, Shine, Standardize, Sustain): A workplace organization method that supports Lean Layout by reducing clutter, improving organization, and promoting efficiency Spaghetti Diagram: A visual tool to trace the path of workers or materials in the workspace to identify excessive movement and areas for improvement Kanban System: A scheduling system that controls the flow of materials based on demand, ensuring the layout supports just-in-time production and minimizes inventory Facility Design Software: Tools such as AutoCAD, MS Visio, and others help simulate, design, and optimize layout before physical changes are made. 1. 2. 3. 4. 5.

Lean Layout Designs Product Layout (Flow Shop Layout) 1. Equipment and workstations are arranged in the sequence of the product s production process Best For: High-volume, standardized products with little variation Advantages: Efficient material flow, minimal transportation, and quick changeovers Example: Automotive assembly lines

Lean Layout Designs Process Layout (Job Shop Layout) 2. Equipment and workstations are grouped by function (e.g., all lathes in one area, all grinders in another) Best For: Low-volume, custom production where products vary significantly Advantages: Flexible, can handle a variety of products with different requirements Example: Machine shops or custom fabrication shops

Lean Layout Designs Cellular Layout 3. Workstations are arranged into "cells" that are dedicated to producing a specific set of similar products Best For: Mid-volume products with similar processing requirements Advantages: Reduces transportation time, improves flow, and enhances employee teamwork Example: Electronic component assembly

Lean Layout Designs Fixed-Position Layout 4. The product stays in one place, and workers and equipment come to it Best For: Best For: Large, complex, or custom products that are not easily moved Advantages: Suitable for large or bulky products (e.g., ships, airplanes) Example: Shipbuilding, construction projects

Lean Layout Designs Hybrid Layout 5. A combination of two or more types of layouts, designed to meet the needs of a specific production process Best For: Low-volume, custom production where products vary significantly Advantages: Flexible, can handle a variety of products with different requirements Example: Machine shops or custom fabrication shops

The Founder

Designing Lean Layouts Define Objectives and Constraints: Identify the goals of the layout redesign (e.g., reduce cycle time, minimize transportation, improve safety) and consider any physical constraints (e.g., space, equipment) Map Current State: Use tools like value stream mapping or spaghetti diagrams to assess the current state and identify inefficiencies and waste Identify Improvement Opportunities: Based on the current state analysis, identify areas where layout changes could improve flow and reduce waste Develop Layout Alternatives: Create several layout alternatives, considering factors like space utilization, flow efficiency, and flexibility Evaluate and Select the Best Layout: Compare alternatives using criteria like cost, flexibility, and ease of implementation Implement the New Layout: Make the changes and ensure that employees are trained on the new layout and processes Monitor and Adjust: Continuously monitor the performance of the new layout, using metrics like cycle time and throughput, and make adjustments as necessary 1. 2. 3. 4. 5. 6. 7.

Design of Experiments (DOE) Statistically determine ways to improve an existing process Define the optimal operating conditions for a process Identify savings opportunities to optimize material and energy use Recognize opportunities to reduce process variability

DOE Benefits Objective: Identify the key drivers of process performance and quantify their impact Focus on Variability: DOE allows teams to identify sources of variation and reduce them, improving process consistency and reducing defects Optimization: DOE helps find optimal process settings to maximize output while reducing waste and inefficiencies Data-Driven Decisions: Provides a rigorous, statistical foundation for decisions, ensuring improvements are based on data, not assumptions

Design of Experiments (DOE) Input Factors Controllable and Uncontrollable Inputs that affect the outcome of a process Levels Defined levels of variation of the Input Variables, typically set at low, medium and high levels for the experiment Response The measurable output of the experiment Input Variables Process Results

Design of Experiments (DOE) Replicates Controllable and Uncontrollable Inputs that affect the outcome of a process Interactions The effect that two or more factors combined have on the response, which may differ from their individual effects

Steps to Conduct DOE Define the Problem and Objective: Determine what you want to optimize or understand (e.g., quality, yield, throughput) Select Factors, Levels, and Responses: Identify the key factors to study, their levels, and the corresponding responses Design the Experiment: Choose the appropriate experimental design (full factorial, fractional factorial, etc.) Conduct the Experiment: Run the experiment, ensuring randomization to reduce bias and error Analyze the Data: Use statistical tools to analyze the results and identify significant effects Interpret Results and Make Decisions: Evaluate the results to determine the optimal settings or identify areas for improvement Confirm Results and Implement Changes: Validate the findings with additional experiments, if necessary, and implement the changes to the process 1. 2. 3. 4. 5. 6. 7.

DOE Results Terminology Main Effects The individual impact of each factor on the response Interactions How the combination of factors affects the response, often in a non-additive manner Confounding A situation where the effect of one factor is mixed with the effect of another, making it difficult to isolate their individual impacts

DOE Results Terminology Randomization The process of randomly assigning experimental runs to minimize bias and ensure reliable results Blocking Grouping similar experimental units together to account for variability and reduce experimental error

DOE Analysis Techniques 1. ANOVA (Analysis of Variance): A statistical technique used to determine if the means of different groups (factors/levels) are significantly different 2. Regression Analysis: A method for modeling the relationship between the factors and the response, often used in response surface methodology (RSM) 3. Effect Plotting: Visualizing the effects of factors and their interactions on the response to identify trends and patterns 4. P-Values: Used to assess the statistical significance of the factors and interactions in the experiment 5. Confidence Intervals: Provide a range of values for the estimated effects, giving an indication of their reliability

DOE Challenges Confounding Effects: Ensuring factors are not confounded with each other is critical for reliable results Complexity in Setup: Designing experiments for multiple factors with several levels can become complex and resource-intensive Interpreting Interactions: Interactions between factors can complicate the interpretation of results and require advanced statistical knowledge Data Collection: Collecting consistent and accurate data can be challenging, especially for high-variability processes

Design of Experiments (DOE) Video

Full Factorial Experimental Design Full Factorial Design is a type of experimental design where all possible combinations of factors and their levels are tested. Purpose: To understand the individual and interactive effects of factors on a response variable Scope: Suitable for experiments where all factors are of interest and the number of factors and levels is manageable Advantages: Provides detailed insight into factor interactions Allows for the estimation of all main effects and interactions Disadvantages: As the number of factors and levels increases, the number of experiments grows exponentially, which can be resource-intensive

Full Factorial Design Key Concepts 1. Factors: Independent variables (inputs) that are varied in the experiment (e.g., temperature, pressure) 2. Levels: The different settings or values each factor can take (e.g., low, medium, high) 3. Main Effects: The impact of each factor on the response 4. Interactions: The combined effect of two or more factors on the response 5. Runs: The number of different combinations of factors and levels to be tested; in a full factorial design, all possible combinations are tested

Full Factorial Design Example A company wants to determine the impact of two factors, temperature and pressure, on the yield of a chemical process. Factors and Levels: Temperature: Low (300 F), High (500 F) Pressure: Low (50 psi), High (150 psi) Full Factorial Design: All combinations of temperature and pressure will be tested: Temperature (Low), Pressure (Low) Temperature (Low), Pressure (High) Temperature (High), Pressure (Low) Temperature (High), Pressure (High) Number of Runs: 4 runs (2 factors 2 levels = 4 combinations)

Full Factorial Design Benefits & Limitations Benefits: Comprehensive understanding of the effects of each factor Precise determination of interactions between factors Provides a clear picture of how all factors contribute to variation in the response Limitations: Can be resource-intensive if the number of factors or levels is large Requires significant time and resources to execute experiments for large-scale designs

Fractional Factorial Experimental Design Fractional Factorial Design is a reduced version of Full Factorial Design where only a fraction of the total combinations of factors and levels is tested.. Purpose: To reduce the number of experiments required while still capturing the main effects and important interactions Scope: Suitable when there are too many factors or levels, and performing a full factorial design would be too costly or time-consuming Advantages: Fewer experiments compared to full factorial design More cost-effective for large experiments with many factors Disadvantages: Does not capture all possible interactions, especially higher-order interactions Risk of confounding effects between factors.

Fractional Factorial Design Key Concepts 1. Fractional Replicates: Instead of testing all combinations, a subset (fraction) of the combinations is selected 2. Resolution: A measure of the design s ability to distinguish between main effects and interactions; higher resolution designs can separate main effects from interactions 3. Confounding: Occurs when the effect of two or more factors is indistinguishable from one another because not all factor combinations are tested

Fractional Factorial Design Example A company wants to determine the impact of three factors, temperature, pressure, and time on the yield of a chemical process. Factors and Levels: Temperature: Low (300 F), High (500 F) Pressure: Low (50 psi), High (150 psi) Time: Short (30 min), Long (60 min) Full Factorial Design: Number of Runs: 2 2 2 = 8 runs (All combinations of factors) Fractional Factorial Design: Number of Runs: 4 runs (A fraction of all combinations) Confounding effects may occur, but key main effects and some interactions can still be captured

Fractional Factorial Design Benefits & Limitations Benefits: Significantly fewer experiments than a full factorial design More efficient and cost-effective, especially with a large number of factors Identifies the most important factors and interactions while minimizing experimental effort Limitations: Higher-order interactions may be confounded, making it difficult to interpret results fully Some important interactions may be missed if not carefully considered during the design phase

Hypothesis Testing How to tell if it made a Difference? There are three main ways to improve a process: 1. Shift the Mean up or down 2. Reduce Variation 3. Reduce Defects Hypothesis Testing tells us whether two sets of data (before and after improvements) are statistically different Continuous Data detect the difference between Means and Variance Discrete Data detect the difference in the number of Defects

Hypothesis Testing Stating the Hypothesis Null Hypothesis, Ho There is no statistically significant difference between the two datasets Alternate Hypothesis, Ha There is a statistically significant difference between the two datasets

Hypothesis Testing Performing Hypothesis Tests 1. Determine the appropriate Hypothesis Test based on the types of data 2. State the Null Hypothesis Ho and the Alternate Hypothesis Ha 3. Calculate Test Statistics / Check P value in Test Statistics Table 4. Interpret Results Accept or Reject the Null Hypothesis

Normal Distributions A dataset is considered normal if it follows the curve of the normal distribution

Types of Hypothesis Testing Output Variable / Dependent Variable / Y Variable Continuous Discrete Continuous Simple Linear Regression Logistic Regression Normal Data Input Variable / Independent Variable / X Variable T Test (1, 2 Sample & Paired) ANOVA, HOV Non-Normal Data Moods Median, HOV Chi-Square Test of Independence Discrete Discrete Data (Attribute Data) Categories Good/Bad Machine 1, Machine 2, Machine 3 Shift number Counted things Continuous Data (Variable Data) Distance Time Pressure Temperature Conveyor Speed

Types of Hypothesis Testing Data Types Situation Example Statistical Test To see how output variable Y changes as the input variable X changes To compare the mean of a dataset against a given standard value To compare the means of two different populations To compare the means of more than two populations To compare the variance of more than two populations Comparing the medians of two or more populations To compare the variance of two or more populations To see how the number of units output by two or more populations differ To see if there is a correlation between the accuracy of staff members performing a task with how many years they have worked for the company Simple Linear Regression Continuous Y / Continuous X To determine if the mean turnaround time for an order is less than the standard value of 15 minutes 1 Sample t-Test Does production vary between Day Shift and Night Shift? Is the mean order processing time the same across Day Shift, Afternoon Shift and Night Shift? Is the order processing time variance the same across Day Shift, Afternoon Shift and Night Shift? Is the median order processing time the same across Day Shift, Afternoon Shift and Night Shift? Is the order processing time variance the same across Day Shift, Afternoon Shift and Night Shift? 2 Sample t-Test Normal Continuous Y / Discrete X ANOVA Homogeneity of Variance (HOV) Mood s Median Test Non-Normal Continuous Y / Discrete X Homogeneity of Variance (HOV) Are the number of defects produced by Day Shift and Night Shift statistically different? Chi-Square Test of Independence Discrete Y / Discrete X

Stating the Null and Alternate Hypotheses Statistical Test Normality Run Tests 1 Sample t-Test 2 Sample t-Test (2-tailed) 2 Sample t-Test (1-tailed) Chi-Square HOV (3 Sample) ANOVA (3 Sample) Mood s Median (3 Sample) Null Hypothesis, Ho Data is Normal No Special Causes = constant or T 1 = 2 1 2 2 = 0 21 = 22 = 23 = 2n 1 = 2 = 3 = n Median1 = Median2 = Mediann Alternate Hypothesis, Ha Data is not Normal Special Causes Exist constant or T 1 2 1 > 2 2 0 21 22 23 2n At least one is different Median1 Median2 Mediann