Insights

What is Failure Mode and Effects Analysis (FMEA)?

A FMEA is pretty much what the letters spell out. A FMEA is a Failure Mode and their associated Effects Analysis.

FMEA is a step-by-step approach for identifying possible failures modes and can be applied to a wide variety of situations including:

• Identification of Reasonably Probable Failure Modes
• Elaboration of Implausible Failure Modes that may have an undesired Effect
• Manufacturing and Service industry process Failure Modes, and of course,
• Human Error Causes

These failures are studied in design/development, maintenance, manufacturing and operational processes, maintenance/service, as well as software phases and studies their consequences, to evaluate their effect – typically categorized by the immediate, next level and system level impacts. In a pure FMEA, Failures are not prioritized based on their severity and consequences of the failure modes.

When twisted a bit a “functionally focused” FMEA is a cornerstone step within any RCM analysis as well and has been traditionally used to improve system concept and design. The functional FMEA of RCM differs from a hierarchical FMEA sub-system division in that ift is based upon the functions, sub functions and functional failures.

When the FMEA concept is “extended” to evaluate criticality it then identifies highest priority tasks for mitigation based on criticality – hence the “C” in (FMECA). FMEA/FMECA approaches are not limited to production areas and can be utilized in administrative, maintenance and service departments as well to identify and eliminate loss and waste.

What is Failure Mode Effects and Criticality Analysis (FMECA)?

As noted above – the FMECA is an extension of FMEA that adds a criticality analysis to the basic FMEA to further prioritize failure modes for remediation. Its purpose is to help prioritize potential failures and allow organizations to proactively focus resources on critical issues before they impact reliability of operations, safety, or other critical aspects. In a qualitative construct, it typically charts the probability of failure modes against the severity of their failure consequences.  In a quantitative construct – like those often associated with Mil-1629A approaches, the FMECA adopts failure rates, and failure mode apportioning concepts.

Failure Modes and Criticality Analysis (FMECA) to many is a process that operates off of 1 to 5 or 1 to 10 scales. It catalogues and helps to manage potential failure modes in systems, products, or processes. It also builds upon the principles of Failure Modes and Effects Analysis (FMEA). Not only failure modes, causes, and effects but also their criticality and consequences and at times – the failure rate and apportionment of failure modes to the overall failure rate. Irrespective of the qualitative or quantitative approach chosen, the FMECA analysis aims to focus on the most critical failure modes. It means those that the criticality could have severe impacts on safety, performance, or other aspects such as environmental harm. Besides, it evaluates factors such as likelihood of a failure mode, its severity, and detectability it is well positioned as an impressive methodology. By doing so, FMECA helps organizations assign more resources to address high-risk issues.

Why Launch a FMEA/FMECA Study?

FMEA/FMECAs are launched to proactively identify, assess, and mitigate potential failures in a system, process, concept, or product to help improve reliability, safety, and efficiency. The outcomes of deploying an FMEA/FMECA ultimately lead to better asset reliability, quality and a reduction in costs.

The Purpose of FMEA/FMECA

The FMEA/FMECA results define the system, identifying potential failure modes, analyzing their effects, assesses their criticality, and then developing and implementing corrective actions to proactively identify, evaluate, and mitigate potential failures in designs and processes before they occur.

In many ways – the FMECA is a one-dimensional risk assessment. When we choose to utilize 1 to 5 or 1 to 10 scales, we catalog the risk, and the artifact so we can calculate a Risk Priority Number (RPN).

The RPN Helps to Prioritize Risks (Based on criticality and consequences of failure.)

Evaluate the severity of the effects, the likelihood of occurrence, and the ability to detect the failure (and its causes and effects), and then prioritize the failure modes based on risk (typically using a Risk Priority Number or Action Priority).

Develop and Implement Corrective and Preventive Actions

The natural extension of the FMECA RPN is to design and implement strategies to eliminate or reduce the likelihood or impact of the most critical failures.

Document Findings and Support Continuous Improvement

FMEA/FMECA’s of course serve a second purpose to record the analysis, recommended actions, and their effectiveness to inform future projects and drive ongoing improvement. 

In essence, FMEA/FMECA aims to:

Improve Product Quality and Asset Reliability

By identifying and addressing potential failures early, manufacturers can create more reliable and higher-quality products.

Reduce Costs

Preventing failures early avoids costly rework, scrap, warranty claims, and downtime.

Increase Efficiency

By optimizing processes and reducing failures, manufacturing efficiency is enhanced.

Improve Customer Satisfaction

Delivering high-quality, reliable products leads to increased customer satisfaction.

Support Regulatory Compliance

FMEA/FMECA helps meet quality and safety standards required in various industries.

Mantua’s FMEA/FMECA at a Glance: The 4 Phases

Phase 1: Cause and Effect Analysis

Step 1: Build a Team

Just like in an RCM engagement, we assemble a multidisciplinary/cross-functional team of personnel with diverse knowledge about the systems, processes, product, or service, and the customer needs/expectations. The team can consist of a cross-section of representatives from design, manufacturing, quality, testing, reliability, maintenance, purchasing (and vendor/suppliers), sales, marketing (and customers), and customer service.

Step 2: Define the Scope

Is it for concept, system, design, process, or service? What are the boundaries? How detailed should the scope be? Use flowcharts to identify the scope and make sure every team member understands it in detail.

Step 3: Identify Information

Fill in the identifying information regarding systems, processes, or products, etc. within the scope of work. 

Step 4: Identify the Scope Functions

Identify the failure modes of your scope: In what ways can the system fail? What is the purpose of this concept, system, design, process, or service? What do our customers expect it to do? Usually, you can break the scope into separate subsystems, items, parts, assemblies, or process steps and identify the function of each.

Step 5: Identify the Failure

For each system, sub-system or component, identify the ways failures that could happen. This is a brainstorming session. These are failure modes.

Step 6: Identify Criticality/Consequence

For a FMECA – for each failure mode, identify the criticality and consequences on the system, related concepts, systems, process, secondary/tertiary processes, product, service, customer, or regulations. Assess the severity, occurrence probability, and detectability of each failure mode to prioritize them based on risk priority numbers (RPNs). Identifying critical failure modes that require immediate attention and mitigation. This is the most important activity in FMEA. These failure effects ranked by criticality (if performing a criticality analysis with FMECA).

Phase 2: Criticality Ranking

Step 7: Rate Severity

Determine how serious each effect is. This is the severity (S) rating. Severity usually is rated on a scale from 1 to 10: One being insignificant and 10 signaling catastrophic. If a failure mode has more than one effect, add only the highest severity rating for that failure mode.

Step 8: Determine Root Causes

For each failure mode, determine all the potential root causes. Use information from the Root Cause Analysis to inform the failure modes if you have one handy!

In summary, FMEA/FMECA while often confused as to which is which, are both essential in manufacturing and operations today and can assist organizations in achieving long-term success and improvements in quality, efficiency, and competitiveness by ensuring predictions of failure, criticality, consequence, and mitigation upon which decisions can be informed. Are you ready to start?

Phase 3: Risk Management

It does little to chalk up a win in the benefit column if you only assess the problem and do not recommend mitigation strategies.

Step 9: Mitigation Strategies

Developing risk mitigation strategies and controls to reduce the likelihood or impact of identified failure modes is essential. This may involve design improvements, process changes, redundancy measures, or enhancing monitoring and detection methods.

Step 10: Integration with Design and Process Improvement

Integrating FMEA/FMECA findings into design reviews, process improvement initiatives, and continuous improvement programs. Ensuring that lessons learned from FMEA/FMECA are applied to enhance reliability, safety, and performance.

Step 11: Verification and Validation

Verifying the effectiveness of mitigation actions through testing, validation, or simulation. Ensuring that implemented controls adequately reduce risk and meet desired performance criteria.

Step 12: Documentation and Reporting

Documenting the FMEA/FMECA process, results, and action plans in a structured format. Providing clear and concise reports to stakeholders, management, and regulatory bodies as required.

Phase 4: Continuous Improvement

Step 13: Follow-up and Review

Conducting periodic reviews and updates of FMEA/FMECA to reflect changes in processes, systems, or operating conditions. Continuously monitoring and reassessing risks to maintain proactive risk management practices.

Coupled with RCA, FMEA/FMECA are essential tools for identifying and addressing potential risks, failures, and inefficiencies in systems, processes, and products. They help organizations improve reliability, safety, and performance while supporting continuous improvement efforts.

Why Mantua Group?

• Independent Expertise: We apply structured, vendor-agnostic FMEA/FMECA methodology.
• Industry-Proven Tools: Digital scribing, logic trees, consequence ranking models.
• Cross-Sector Know-How: Clients in energy, mining, defense, and heavy industry.
• Execution Focus: We ensure FMEA/FMECA outputs are deployable in your CMMS with clear work instructions, frequency logic, and resource planning.

One-Page Visual Workflow

Download our FMEA/FMECA Project Infographic or connect with our team to learn how to apply Mantua’s pragmatic approach to your next critical asset.

Visit www.mantua.group to explore how we support reliability transformations that deliver sustainable results.

What is Weibull Analysis

Weibull analysis is a statistical method used in reliability engineering to analyze failure data and from the parameters estimated you can also predict future failures. It utilizes the “Weibull Distribution”, which is a flexible probability distribution that can model various failure patterns shown above, making it a valuable tool for understanding and improving product or system reliability. By fitting the Weibull distribution to failure data through a method called statistical regression, reliability engineers can predict failure rates, and use this to optimize maintenance schedules, predict the need for future CAPEX replacements, and improve product design. 

The challenge with Weibull analysis revolves around the data and the algorithm used. Complicating the data problem is that many firms miss the need to deal with incomplete data at all and as a result get the wrong answers. At the Mantua Group we know how to counter this challenge correctly.

Why Launch a Weibull Analysis

The “WHY” is simple. We want to estimate the past performance from the data we have, and predict the future performance to answer a question.

The Weibull model is a statistical distribution used in reliability analysis to model the time to failure (TTF) observed data (sometimes called technically longitudinal data) and then estimate the asset performance. The typical values of interest are the mean time to failure (MTTF), mean time between failures (MTBF), and mean time to repair (MTTR). The Weibull distribution is quite flexible, and allows for both increasing and decreasing failure rates, and can describe various types of failures, such as infant mortality failures and wear out failures. The Weibull distribution is also used to often model asset or component service life data analysis, which is the time until device failure of many different physical systems, such as a bearing or motor’s mechanical wear.

Weibull analysis can work with functional failure data, catastrophic failure data, and non-longitudinal time data like revolutions, or counts.

In other words, it can assess asset reliability and model failure times from many different types of data.

The Purpose of Weibull Analysis

Core Objectives of Weibull Analysis

Weibull analysis is essential in many fields including asset management and product manufacturing but for different reasons. In manufacturing of a product it is often employed to understand the products reliability, and forecast the warrantee return impacts from initial data. In asset management and reliability, it is used to optimize maintenance strategies, spares forecasting, and justification of capital renewals.

This approach leads to several key benefits:

Modeling Probability of Failure (PoF)

The fundamental goal is to use the Weibull distribution to statistically analyze “life data” (typically, time to failure) to model how likely an asset, product or component is to fail over time. One of the more useful pieces of information recovered from the data is the PoF and the hazard rate, (given an item has survived up until now, what is the likelihood it will survive the next period of time?

Understanding Failure Characteristics

By analyzing the Weibull distribution’s shape parameter (β), manufacturers and maintainers gain insights into the nature of failures occurring. Alignment of the type of maintenance deployed is different for each of the three cases below:

• β < 1 indicates early life failures (infant mortality)
• β = 1 suggests a constant failure rate (random failures)
• β > 1 points to wear-out failures

Failure Prediction and Maintenance Planning

Weibull analysis helps predict when failures are likely to occur, enabling asset maintainers to proactively schedule preventative maintenance and optimize resource allocation. Ideally you maximize the asset’s useful life, and retire or repair it just before it would have failed in service.

Product Design Improvement

The insights gained from Weibull analysis inform design decisions to enhance product reliability, addressing root causes of early failures or implementing material upgrades or higher quality for wear-out failures. The topics of Reliability Growth (over time) and the use of Accelerated Life Testing (ALT) are intertwined here.

Cost Reduction

When we bring cost implications into play, we associate the cost of failure of an asset with the probability of the asset failure. By identifying and addressing failure patterns, manufacturers can reduce costs associated with warranty claims, repairs, and unplanned downtime.

Risk Management

It helps assess and manage the risk of failure, particularly in safety-critical industries like mining, transportation, utilities, aerospace, healthcare, and energy. In manufacturing, Weibull Analysis aims to improve operational efficiency and product quality by providing data-driven insights into systems, equipment, component, and/or product reliability and failure patterns. Most enterprise risk models now utilize the Weibull distribution to predict the failure probability.

The Three Basic Applications for Weibull Analysis

1. Modeling Time-to-Failure Data
2. Situations Where Failure Rates May Vary
3. Applications in Other Fields

At Mantua, we guide clients through each application with discipline and clarity, using a blend of expert facilitation, cross-functional collaboration, and modern tools.

Application 1: Modeling Time to Failure (TTF) Data in Assets

The world’s infrastructure assets are as diverse as the planet’s fish—each uniquely adapted to its environment, with distinct lifespans, vulnerabilities, and roles in the ecosystem it supports. The deep-ocean leviathans of this system include power grids: vast, foundational networks of transmission lines, bulk substations, and central transformers that span regions. In contrast, local distribution systems resemble schooling fish: smaller, numerous assets like poles, fuses, and smart meters that operate within communities. Together, these layers determine the stability of the infrastructure ecosystem.

Like marine biologists who must study a single habitat rather than the entire ocean, asset managers must focus their analysis on particular sectors to derive actionable insights. This research casts its net into the electrical utility sector, where data challenges frequently cloud decision-making. Incomplete or inconsistently recorded asset data acts like sediment in a stream, obscuring the view beneath and distorting reliability assessments.

Left-truncation, described by Emura and Michimae (2022), refers to assets that failed before data collection began—like fish that lived and died before the rod was ever cast. Right-censoring, as examined by Abernathy et al. (1984) and Wu and Carroll (1987), describes assets still in service when data collection ends; they are like healthy fish still swimming downstream, their outcomes unseen. Informative censoring, explained by Lagakos (1979), occurs when weaker assets are selectively removed from service early, akin to fishing out only the weakest fish, thereby distorting our understanding of survival patterns (Piovanni, 2021; Yanling, 2022).

Traditional reliability models, which assume full and unbiased data visibility, often fail to account for these phenomena. They count only what is caught in the net, ignoring what evaded observation—leading to underestimation of risk and misinformed lifecycle planning.

While biostatistics and survival analysis offer robust methods to manage such data irregularities, their application to infrastructure asset management remains limited. Most asset-intensive industries continue to rely on models that were not designed to manage left-truncated or informatively censored data.

At the Mantua Group we understand these shortcomings that often generate the wrong answers and guide you towards the truth.

Reliability Analysis

The Weibull distribution is widely used to model failure times, allowing engineers to predict failure rates, optimize maintenance schedules, and improve product design. 

Equipment Lifespan Analysis

It can help determine the expected lifespan of a product or component, facilitating informed decisions about maintenance intervals and replacement strategies. 

Warranty Claim Analysis

Analyzing warranty claims using Weibull can provide insights into product reliability and potential design flaws. 

General Failure Analysis

It can be applied to various industries, including manufacturing, electronics, aerospace, and healthcare, where understanding failure patterns is crucial. 

Application 2: Situations Where Failure Rates May Vary

Increasing Failure Rates

When the likelihood of failure increases over time (e.g., due to wear and tear), the Weibull distribution with a shape parameter (beta) greater than 1 can model this behavior. 

Decreasing Failure Rates

When the failure rate is higher at the beginning of equipment life (e.g., infant mortality) and then decreases, the Weibull distribution with a beta value less than 1 can model this. 

Constant Failure Rate Situations

When the failure rate is roughly constant over time (similar to an exponential distribution), the Weibull distribution with a beta value of 1 can be used. 

Infant Mortality

The Weibull distribution can also be used to analyze and address infant mortality, which is defined as a higher probability of failure at the start of equipment life and historically the terminology has root sin Human mortality statistics where children under the age of 5 were 100 years ago much more likely to perish than those who had survived to be 6 years or older.

Application 3: Applications in Other Fields

Biology

The Weibull distribution can be used to model survival times in biological studies. 

Other Engineering Disciplines

Beyond reliability, it can be used in material science, quality control, and other engineering areas where analyzing time-dependent phenomena is important. 

Weibull at a Glance: 5 Key Steps

Weibull analysis, also known as life data analysis, is a statistical technique used to analyze the reliability of products or systems. 

Step 1: Data Collection

Gather Time to Failure (TTF) Data

Collect data on the time it takes for products or systems to fail. This includes failure times, time intervals, or other relevant data. This can be particularly difficult in industries where assets are removed preemptively before they fail or have been removed because they failed a condition-based inspection.

Identify Censored and Truncated Data

Note any censored data points, where a product was still in operation during the analysis period. 

Identify left truncated and informatively censored data.

Step 2: Data Preparation

Clean and Validate Data

Ensure the data is accurate and free from errors. This is the most difficult step as most organizations do not possess data systems to curate data in a clean format. Our experience using AI tools is that you “Can” cleanse almost any dataset, with persistence, blood, sweat, and tears.

Consider Data Type

Determine if you are working with complete, censored, informatively censored, truncated or interval data. 

Step 3: Data Analysis

Choose a Distribution Model

Select a Weibull distribution (2-parameter or 3-parameter) that best fits the data is a good place to start. For prediction and optimization models we recommend the model selection be the best mathematical fit and therefore include the normal, log-normal, exponential, log-logistic, and Cox proportional hazards models (not an exhaustive list of available distributions) if they fit your data better.

Estimate Parameters

Estimate the shape parameter (β) and scale parameter (η) of the Weibull distribution using various methods like maximum likelihood estimation (MLE) or Ordinary Least Squares (OLS) with Rank Regression methods. Which methodology depends on your data as OLS can only handle a portion of the complexity, and MLE methods must be employed for complex data situations

Generate Plots

Create Weibull probability plots to visually assess the fit of the distribution. The plots also provide insight, and allow key statistical metrics to be calculated.

Step 4: Interpretation and Application

Analyze Results

Analyze the estimated parameters and Weibull plots to understand the reliability characteristics of the product or system. Here we look at the characteristic life, the shape parameters, but also the percentile values of 1% and 99%, the confidence limits (typically 95%), and various Goodness of Fit, AIC, BIC and possibly the Anderson-Darling indicator to assess the correctness of your analysis.

Determine Reliability and Failure Rate

Use the Weibull parameters to calculate reliability (probability of surviving) at specific times and failure rate over time. The basic parameterized equations are useful for calculating the reliability, unreliability (CDF), Failure rate, and probability density functions (PDF).

Most people however are more interested in the conditional probabilities given their asset has survived up until now, what is the likelihood it will survive without failure over the next scheduled uptime period (perhaps 1 month)?

Make Predictions

Use the Weibull distribution to predict the future reliability of the product or system. Again, with the basic Weibull parameters, we can calculate precisely a variety of scenarios and answer great many questions of interest with analytical calculations, but also noting the parameters obtained from regression are “estimates”.

Identify Failure Mechanisms

Analyze the shape parameter (β) to understand the failure mechanisms (e.g., infant mortality, random failures, wear-out failures).  The specific relation to components may take some interpretation.

Determine Mean Time to Failure (MTTF)

Calculate MTTF to understand the average lifespan of the product, asset or system. Once the parameter estimates are known mathematically using the Beta function, we can then calculate the MTTF directly from te Weibull parameter estimates. If we boldly estimate the MTTR, then we can also calculate the MTBF = MTTF + MTTR (at steady state) and estimate the availability.

Step 5: Implement Strategies

Implement Strategies

Use the Weibull analysis findings to develop strategies for improving reliability, optimizing maintenance, and reducing costs. Typically, this involves optimizing the frequency of preventative maintenance (PM) and on-condition inspections or Condition Based Maintenance (CBM) used to detect progressing potential failures.

Review and Update

As time goes on, so does more failure data get generated. Periodically review and update the Weibull analysis with new data to reflect changes in the product or system’s reliability. 

In Summary

Weibull Analysis is essential in manufacturing, reliability, asset management, health care, survival analysis, medical trials, and operations today and can assist organizations in achieving long-term success and improvements in quality, efficiency, and competitiveness by ensuring predictions of failure are based in data on which decisions can be informed. Are you ready to start?

Why Mantua Group?

• Independent Expertise: We apply structured, vendor-agnostic Weibull methodology.
• Industry-Proven Tools: Statistical software, Prediction models, optimization models.
• Cross-Sector Know-How: Clients in energy, mining, defense, and heavy industry.
• Execution Focus: We ensure Weibull Analysis optimization outputs are deployable in your CMMS with clear work instructions, frequency logic, and resource planning and predictions are deployable in your investment planning software.

Weibull with Real Business Impact

Mantua’s Weibull Analysis programs reduce unplanned downtime, improve asset reliability and availability.

One-Page Visual Workflow

Download our Weibull Analysis Project Infographic or connect with our team to learn how to apply Mantua’s pragmatic approach to your next critical asset.

Visit our Failure Data / Survival / Weibull Analysis page to explore how we support reliability transformations that deliver sustainable results.