Catching Your Airplane Using Lean and Six Sigma

We have all had to plan a trip to the airport. Sometimes it goes well and sometimes not so well. One of the problems we have is dealing with the variation in the trip time to the airport. Since I travel to the airport a lot and like analyzing data, I keep track of my travel time to the airport. And living in Boston has always been an adventure when going to the airport.

Our question today (or our problem statement) is simple: “How much time should I allow for my trip to the airport so I don’t miss my plane?”

Here in figure 1 is a run chart (data over a time period) that shows my trips to the airport over time. The numbers on the left show my total trip time in minutes—from leaving the house until completing the security check in the airport. Note that I have made 58 trips to the airport and that my average time to get there is 93 minutes. 

Figure 1: Trip Times to Airport (minutes on left, number of trips at bottom)

Based on the data in figure 1, what happens if I always plan to take 93 minutes (the average) to the airport? You’re right; I miss a lot of flights. But how much time should I plan so that I can make my flight and not waste a lot of time sitting at the airport? Let’s apply lean and Six Sigma to help answer the problem.

Don’t be surprised at our answers—you know more about lean and Six Sigma than you may suspect.

Applying lean

Lean is a structured methodology to document a process with a high level of detail and then drive improvement in time and quality. Lean is also used to understand the workplace layout. There is a rich set of tools and methods (e.g., 5S, kaizen, value stream, kanban) to manage time, waste, yield, and queue time. By first documenting the process and then leaning it out, you can achieve major process improvements. So let's lean out our airport trip.

Figure 2 shows some details about my trips to the airport:

 
Figure 2: Trip details

Note that the above times are average times. Remember that the average time is calculated from 58 separate measurements.

Lean helps us document the process and then identify ways to eliminate the step or to improve the time that particular step takes. Based on this concept, some immediate ideas come to mind to improve our process:

• How about purchasing a toll road fast pass?

• Parking always takes a long time as well. How about having someone drop us at the airport so that we can save the 15 minutes for parking?

These small improvements definitely affect the process and can produce measurable results. The way we improve our airport trip can also be used in improving our business processes. The tools can and are more complicated, but the concept is exactly the same; document your process and then identify and implement small improvements over time. Measure your success and continue to improve the process over time.

But is there a better way to improve the process? Yes. Small incremental steps can drive incremental change, but sometimes a whole new approach is necessary.

Lean results

I did purchase a fast pass and saved my toll time. I also purchased a “gold passport” card to get preferred parking that is close to the airport terminal. However, those changes only reduced by travel time an average of 10 minutes.

So how did I cut an additional 20 minutes off my travel time and significantly reduce my trip time variation? By drastically changing the process.

As you know I live outside of Boston. Going to the airport always involved driving in downtown Boston (20 minutes in the average process time along with a lot of variation). Downtown Boston is known for its small one way streets, notoriously poor roadway signs, traffic, and the Big Dig. Part of the Big Dig project was to create a two-mile tunnel that connects the toll road to the airport, which allows you to bypass downtown Boston. When the Ted Williams Tunnel (figure 3) opened up in January 2003, it immediately eliminated the need to drive through downtown traffic. It reduced my trip time by 20 minutes (on average). The much bigger benefit was that this process change significantly reduced the trip time variation.

Figure 3: Ted Williams Tunnel

The moral of the story: Analyze your existing process and make minor improvements but don’t be afraid to investigate radical process changes to make major improvements. Don’t be afraid to challenge existing norms and rely on the strength of the team to drive success.

Six Sigma

A Six Sigma level quality process is 99.99966-percent perfect. It’s also typically called a five nines process because of the first five nines in the percentage. The major goal of Six Sigma is to reduce variation in the process. How good is a Six Sigma process? A Six Sigma process has 3.4 defects per 1 million opportunities. Figure 4 shows a few examples of day to day processes we live with and their quality levels.

Figure 4: The difference between 3.8 sigma and 6 sigma in everyday processes

Applying Six Sigma

The airline industry example in figure 5 below exposes us to the fact that not all processes have to be 99.999-percent perfect. I am sure the quality of baggage handling doesn’t meet your expectations since we have all lost a bag during our travels. But the airline industry does boast a 99.99998-percent (six nines) quality level safety record based on the number of fatal events per million.

Figure 5: Airline safety record
(June 2009 Airline statistics. www.consumerwarningnetwork.com [1] )

So not only do you need to consider where to apply Six Sigma, you also need to determine what quality level is required. When a human life is involved, what quality level and what risk level is acceptable? Is the acceptable quality level different when considering our trip to the airport? What’s the impact of missing a flight or having to rush? A good rule of thumb: pick the right quality level that is necessary to delight the customer.

Figure 6 below is a histogram of my airport trip information. It’s the same 58 trips, now sorted by the length of time of the trip.
 
Figure 6: Histogram of trips sorted by length of time

What does the chart tell us? The good news is that the data is starting to tell us more about the distribution of data around our average trip time. In fact, it shows us that our trip is never less than 50 minutes and those 36 trips were actually less than our average time! We can also see a few very bad trip times of more than 170 minutes (almost three hours). But the problem still remains. How much time should I plan for the trip? Well the good news is that we know more about the data. We all know that Saturday morning will have less rush hour traffic. We also know that during the week, we can work around the rush hour by leaving early. So let’s create one more chart before answering the question on how to reduce our trip time and reduce wasted time at the airport.

Figure 7: Trip times when leaving house between 4 a.m.and 6 a.m.

The chart above shows my trip times when I leave the house between 4 a.m. and 6 a.m. We all know that the time of day is a major source of variation in our data and this chart highlights that fact. The chart above shows that our average time to the airport, if we travel between 4 a.m. and 6 a.m. (before rush hour traffic!) is 65 minutes which is significantly less than our overall average time of 93 minutes.

Six Sigma results

I constantly collect data on my trips to the airport. And I constantly adjust my trips based on the data. Seasonality is a good example of a macro trend that would change my average trip times. A snow storm is a local issue that can also impact specific trip times and must be considered.

For my specific experience, the opening of the Ted Williams Tunnel had the biggest impact on my airport experience. It reduced the overall trip time by almost 18 minutes with one process change and reduced a significant amount of variation since I no longer had to leave the toll road and drive through the city.

That process change represents the best of both worlds—a reduction in time and variation.

Overall results

Using basic lean and Six Sigma tools, we can significantly improve our airport experience:

• The time you leave for the airport is a significant factor on overall trip time.

• The time you leave also turns out to be a significant factor on variation.

• A significant process change (the Ted Williams Tunnel) has the biggest effect on my trip time and variation.

However, even with these great results, our improvement project never really answered our question: “How much time should I allow for my trip to the airport so I don’t miss my plane?”

To answer this question we need to add another component to our analysis; risk and confidence factors. Understanding and managing risk is a necessary part of any analysis. If we never want to miss a flight (100% quality level) we could simply plan three hours for every trip. But what if you only allow two and a half hours? How many flights could you miss and what are the consequences of a missed flight? Did you miss an important business meeting? Or maybe it makes very little difference since the Boston to New York shuttle flies every 30 minutes.

Based on these considerations, our real question is: “How much time should I allow for my trip if I want to have a 90-percent confidence factor that I will make my flight 95 percent of the time?”

In another installment, we can answer that question.

Recommendations

Combining lean with Six Sigma provides a powerful methodology you can implement to drive operational change and improvement. Applying lean Six Sigma helped me improve my airport planning by making it more visible and predictable.

These simple concepts can be applied to your business processes. Focusing on improving or eliminating as many steps in the process as possible, and then by reducing the variation in the remaining processes will help you improve the quality of delivery of your product and services. When you apply this methodology to customer processes you can also influence customer loyalty.

Sustainable, fundamental improvements can be achieved when lean and Six Sigma are part of an overall quality management system across the enterprise.