Problems Happen. It’s How They’re Solved That Matters.

Best practices for conducting effective root cause analysis in manufacturing

Published: Thursday, November 4, 2021 - 12:02

Within every organization, problems or incidents arise that can affect the quality of your operations. Take for example, food recalls due to improper food labeling that not only could cause sickness in humans, but also result in a hit to a company’s reputation. Or, automotive product recalls due to defective parts.

Today, given supply-chain disruptions that require broader sourcing of materials, along with a shortage of skilled workers, the chances of problems rising are increasing exponentially.

Whether causing a minor blip in productivity or tragic results, any problem should be evaluated to minimize the chances that it could occur again. And, while we read about manufacturing issues almost daily, we don’t always learn what the root cause of the problem was and how it was corrected.

When failures occur, it’s not always because a worker forgot to read a process manual or because a battery was sourced from a new provider. Instead, the root of the problem might be systemic. The manual might not have been easily accessible, or new suppliers may not have been properly audited. Root cause analysis means going past the apparent source of the problem and uncovering deeper issues to increase quality and safety at the organizational level.

Barriers to effective root cause analysis

Although many organizations may think they have a master plan for root cause analysis, there are many barriers to its successful implementation. Time constraints can get in the way of thorough analysis, and often the urgency to explain the reason for the incident to a corporate board, other company stakeholders, or the public can cause companies to speed up the process, with no thoughtful path for avoiding it in the future.

In addition, many incidents, especially those involving safety, are too quickly assigned to human operator error, but that can be a cop-out. Often, human error can be a contributing factor, but more likely inefficient processes or workflow can be the real reason for the incident. Another thing to consider is that many companies that assign human error as the root cause choose more training as the corrective action. But if the training didn’t work the first time, why would it work the next time around? Maybe there needs to be a new approach to the task.

Even when a root cause analysis is done correctly, it’s not the end of the story. Many organizations declare success when they identify the issue and take corrective action to solve it. Yet it’s important to regularly monitor the action and verify its effectiveness because needs can change at any time and may require a new approach.

Six steps to better resolution

Many problems can originate in equipment, materials, processes, or technology, yet it’s usually one primary factor that is the cause, even though contributing factors can impact the outcome. The ultimate goal of root cause analysis is to trace the problem to its origin, using key steps along the way. Consider the following six steps:

1. Define the problem. Start with the basic question: What happened? The root cause analysis (RCA) team should first document and contextualize the incident. They should dig deep to look for early warning signs that could have indicated a problem and then document the symptoms.
2. Measure the impact. Determine if it’s an isolated incident or a recurring one. Understand the organizational impact it has caused, such as unplanned plant downtime, a negative brand image, or compliance violations. You also should determine the duration of the problem and its impact on organizational processes.
3. Uncover contributing factors. You might determine that the root cause of recalled hamburgers, for instance, may be contaminated beef, but go beyond the obvious. Contributing factors that caused the beef to be contaminated could be inadequate worker training or a failure in the automated system being used. These factors also must be considered and addressed.
4. Gather the facts. Take a data-driven approach to root cause analysis by leveraging standard methodologies and tools that guide you to the culprit (see examples below).
5. Provide resolution. Depending on the results of the analysis using standard methodologies, create and clearly define the solution that mitigates or eliminates the original problem. This should include timelines and specific steps that will be taken.
6. Secure C-level buy-in. For root cause analysis to be effective it requires commitment from the top. By educating company leaders on the role of RCA done right and its benefits, it’s easier to get enterprisewide buy-in, as well as openness to new ways of working.

Below are some of the primary tools and methodologies being employed to guide a data-driven, structured root cause analysis.

Fishbone diagram. Also known as the cause and effect diagram, it was first developed in the 1960s by Kaoru Ishikawa to determine potential root causes of complex problems. The “fish’s head” is considered the problem, and the six “fishbones” are the areas of investigation to uncover potential reasons for it: measurements, materials, personnel, environment, methods, and machines. The method is conducted in four steps, during which you identify the problem; ask questions (e.g., Was it an environmental problem? Human error?); then you drill down further to identify potential reasons for it; and finally, you analyze it to determine the root cause.

5 Whys process. Created by the founder of Toyota for its production system, the 5 Whys process is used to identify and correct moderate-complexity problems. It starts by creating a problem statement and asking “why” five times to get to the root cause. Each “why” drills a bit deeper into the cause. You go through five steps: assemble a team familiar with the problem; define the problem with a clear statement; ask why and get factual, data-driven responses; develop a corrective action plan; and finally, monitor results to measure effectiveness.

Pareto chart. The Pareto principle follows the 80/20 rule: “For many events, roughly 80 percent of the effects come from 20 percent of the causes.” Comprised of bar graphs, it helps you prioritize the most important issue and broad causes. This methodology follows four steps: Deciding which categories and measurements are appropriate (e.g., frequency, quantity, and cost); determining the time frame; collecting and assembling the data and total measurements for each category; and calculating and plotting cumulative results.

Problems happen in manufacturing, but the key is in identifying the root cause to learn from it and prevent it from replicating and spiraling out of control. Effective root cause analysis avoids the blame game, which can lessen employee and partner morale, and uses an approach that is guided by data, driven by insights, and seeks a common goal of continuous improvement in processes, people, and performance.

About The Author

David Isaacson

David Isaacson is the vice president of product marketing at ETQ. Isaacson focuses on developing market strategies and product positioning for cloud-based solutions.