Classify Safety Issues with Classifications
Organizing data and information is extremely important in SMS, as much of your data analysis and performance monitoring depend on such organization. One best practice to organize your data is assign classifications to issues.
A classification is an organization tool. A classification is a word or a short phrase that describes the safety issue in some way. In other words, classifications categorize your issues. They can be categorized in different ways, such as:
- It’s Type of Issue, such as “Lost time injury”;
- It’s central problem (hazard), such as "unmarked hazardous materials container";
- The central human component (Human Factor) that contributes to situation, such as "Overtime hours";
- The root cause(s), such as "Outdated safety checklist"; and
- Other factors, such as classifying the type of relevant procedure, policy, task, etc.
When you are managing safety issues, you assign classifications to the issue to organize it. Over time, you can easily research data regarding specific categories.
For example, you could find out how many runway incursions were caused by lack of communication by searching for issues with a “runway incursion” hazard classification and a “lack of communication” human factor classification.
Purpose of Classifying Issues in Aviation Safety Management
Classifications are perhaps the best way to organize your data. Its goals are to find a way to organize data:
- With consistent terms and verbiage;
- With multiple types of organization; and
- In ways that make it easy to retrieve.
Case in point. We recently worked with an operator who was attempting to adopt a classification system to organize data. We were reviewing their data and trying to establish how many “Go arounds” they had so far this year.
In their existing program, they were trying to search through their issue titles for the term “go around.” What we quickly found though, is that different managers had used 4 (and possibly more) different terms for this idea, such as “fly by”, “fly around”, and “go by.” And there may have actually been more with different terms, or spelling.
The result was that it was:
- Difficult to find even basic data; and
- Nearly impossible to trust the complete accuracy of the data.
The purpose of classifications is to solve these kinds of problems, which are significant and very real in the aviation industry.
Benefits of Organizing Data with Classifications
Some of the benefits of organizing safety data with classifications involves having classification trees that:
- Are easy to navigate;
- Comply with standards of oversight agencies; and
- Provide specific data organization (but not TOO specific) that allows for meaningful data analysis.
Good use of classifications provides the following benefits:
- Understand biggest safety needs;
- Understand safety performance;
- Data mine the safety program for lots of different data;
- Establish trends in safety issues;
- Search for data given multiple parameters; and
- Most importantly, do all of this very easily.
Classifications are fundamental to having the most reliable data possible. It is very difficult to have high quality data analysis and data mining without them.
1 - Classify Issues with Different Types of Classifications
Organizing your safety data with only one classification provides some minor benefit for acquiring data. However, data that can be organized in multiple different ways provides a powerful ability to understand data very specifically.
Using classifications, you grant yourself the ability to organize data in different ways by assigning multiple types of classifications. For example, we highly suggest that for each issue, you categorize issues based on, at the very least:
- Type of Issue;
- Root Causes; and
- Human Factors.
When you classify each issue with different classifications, you can search for issues based on any and all of your classifications. You might, for example, search for “Lost time injuries” that happened because of “lack of communication” and “faulty equipment”. This is a bit of a facetious example, but it highlights how you can really drill into your data.
2 - Have 3 Levels in Classification Trees
A very good practice for organizing your data is to create classification trees with three levels:
- Category - system related classifications, such as:
- Flight Ops,
- Ground Ops,
- On-Board Aircraft
- Rejected Landing,
- Aborted Landing,
- Unstable Approach
Classifications are organized this way not so much for the benefit of retrieving data and performing data analysis, but for the benefit of proving classifications to issues:
- With due considerations of all the relevant classifications; and
- In way that is manageable and easy to do.
3 - Ensure that Classifications are Specific (But Not TOO Specific)
One questions we often get asked is: how specific should classifications be? If you organize your data with classifications that are too specific, it will be very difficult for you to perform meaningful data analysis.
The point of organizing data is to clump data that is similar, not to describe your piece of data in intimate detail.
- Specific enough to be useful for making safety decisions and establishing trends;
- Not so specific that you need many, many classifications to capture the essential idea; but
- General enough to be useful for capturing the main idea, such as grouping highly similar situations.
Describing data in detail is the job of issue titles and descriptions.
Here are some examples of good specificity for hazards classifications:
- Flight Ops>>Wildlife>>Birdstrike;
- Flight Related>>Wildlife>>Wildlife on runway; and
- Flight Related>>Wildlife>>Rodent in aircraft.
These examples are specific enough to capture the essential piece of information (“wildlife” and “runway”) but not so specific that you will need to assign a couple of classifications just to capture one idea.
The point is: one type of problem, one type of classification.
4 - When Entering Reports, Give Title Meaningful Name
Often times, when employees submit hazard reports, the title they use on their report is not ideal. Title give you a meaningful opportunity to apply meaningful data keywords that will be useful for data mining.
For example, while:
- It’s useful to apply classification of “bird strike” for bird strike issues;
- It’s not useful to specify the type of bird in the bird strike classification (i.e., “goose strike”);
- You can indicate the type of bird strike in the title of the issue, such as “Bird Strike – Goose”
Hence, at the end of the year you can search through issues for “goose” to get a good estimate of how many goose strikes you had. This is a good practice to do for all issues.
In smaller organizations with one or two managers responsible for managing issues, this practice is easy to manage. In larger organization, we highly recommend you have guidelines for how to rename issues, including a glossary for common terms to use for various types of issues.
5 - Don’t Classify an Issue Twice With Same Classifications
An extremely important practice when you create classification trees and are classifying issues is DRY:
It is a very important term in many industries, as the goal is to ensure:
- Best practices for organization;
- Reduce useless clutter; and
- Reduce extra work.
In the context of classifications, it means don’t include the same classifications in different classification trees. This would cause you to have extra classifications and have to classify the same issue twice.
Furthermore, you avoid causing confusion when people are searching for specific data, and see the same classification(s) in different locations.
6 - Separate Classifications by Type
Different types of classifications should be organized into their own classification trees. Some common types of classifications we see used by aviation service providers are:
- Type of issue
- Human Factors
- Root Causes
- Risk Controls (i.e., create classification tree that includes all risk controls)
- Policies/Procedures (i.e., create classification tree that includes all policies/procedures)
- Job Duties
You can literally create any type of classification tree you want that best fits your organization. The point is that different types of classifications are separated. As noted, each classification tree should have its own unique classifications, and should not repeat classifications from other trees.
7 - Number of Classifications to Assign Per Issue
As discussed, it is very useful to assign different types of classifications to an issue to organize it. Here are some loose guidelines that we suggest for how many of each type of classification to apply when classifying issues.
- Type of Issue: one classification per issue is ideal, but sometimes two are necessary
- One Hazard: one classification per issue, as there is only one primary dangerous condition per safety issue;
- Root causes: multiple classifications per issue is good, as there are usually several root causes
- Human Factors: apply only the most important classifications per issue, usually 1-3; and
- Safety Policy/Procedure: usually none or one classification per issue, but up to two is okay
Remember, these are not hard rules, they are simply best practices – i.e. “soft rules”. Some issues will require you to be flexible.
Good classification practices will make a big difference in how easily and how well you can sort data.