Insight for an Uncertain World
About the Aviation Safety Monitor
The Aviation Safety Monitor is a service provided by Robust Analytics to deliver timely information on terminal area safety in the National Airspace System (NAS). The safety monitoring and prediction technologies were developed by Robust Analytics over the past several years. Partial funding was provided by the NASA Small Business Innovation Research Program and the NASA System Wide Safety Project.
The Aviation Safety Monitor provides quantitative estimates of safety margins at 26 airports in 17 metropolitan regions in the United States. This information complements data on several safety-related events that are published elsewhere, with the FAA’s Runway Incursion Statistics website a good example. However, the available safety information can be misleading if it only reports the frequency of violations with no insight into how safety buffers may vary minute-to-minute and day-to-day. The Aviation Safety Monitor aims to provide this insight every week.
How Do We Measure Safety Margins?
The Aviation Safety Monitor summarizes output from Risk Tracker, the Robust Analytics in-time terminal airspace hazard and safety metrics monitoring system.
Aviation Safety Monitor Weekly Report for the Week Ending May 25, 2024
Safety margins at the 26 monitored airports increased slightly during the past week, reversing the negative trend from the previous two weeks. Safety margins in May remain higher than the average over the past two years.
Welcome back to the Weekly Aviation Safety Report. Every Monday Robust Analytics reports on safety margins at 26 United States airports. With this Aviation Safety Monitor Weekly Report, Robust Analytics offers the aviation community timely assessments of changing safety margins and safety-related events. Dates and times are tracked in UTC and the week ends at midnight every Saturday. This week’s report includes data through 2400 UTC on May 25, 2024.
For New Readers: Please read our article “Did Safety Degrade in the National Airspace System in the Winter of 2022-2023?” that applies our methods and data to examine whether safety margins decreased during the events of winter 2022-2023.
The Aviation Safety Monitor measures safety margins by estimating the frequency, duration, and severity of buffer encroachments. Our paper “How Do We Measure Safety Margins?” provides a detailed description of the methods and data. That can be found here https://www.robust-analytics.com/measure on the Robust Analytics website.
Weekly Safety Margin Update. Safety margins improved during the past week after decreasing for the previous two weeks. Total encroachment durations for the week ending May 25 were 8.3 percent lower than the previous week, and encroachment events dropped by 10.5 percent. Average weekly encroachment durations for the month of May are running 1.6 percent above the April average and event counts are 10.5 percent higher. These changes can be easily seen in Figure 1 that displays weekly summary metrics for the 26 monitored airports.
Figure 1. Weekly Trends in Encroachment Events and Durations
Let’s now take a closer look at safety margins for the seven days ending May 18. Figure 2 displays estimates of encroachment durations per aircraft for each hour over the previous seven days. This offers a detailed look at how safety margins vary over the operating day. Figure 2 also indicates the historical range of the data by showing the 25th, 75th, and 90th percentile values of the duration per aircraft metric. This week the percentiles estimates were updated using data from May 2022 through February 2024. The added eight months of data lowered each percentile when compared to the previous calculations that used data from May 2022 through June 2023.
Figure 2 displays the typical time of day pattern that we observe in the data, as encroachment events are highly correlated with traffic density. Even after aggregating the data over 17 metroplexes operating in three time zones the time of day pattern remains prominent. There were two high spikes in encroachment durations on Sunday May 19 and Saturday May 25. The other values through the week were within the normal range.
Figure 2. Hourly Encroachment Duration Per Aircraft for the Week Ending May 25, 2024
Figure 3 shows duration per aircraft for the four weeks from April 28 through May 25, 2024. With the updated percentile estimates we can observe more periods with encroachment durations above the 90th percentile, although the number remains a low fraction of all observations. Most observations are below the 75thpercentile (the yellow line in Figure 3).
Figure 3. Hourly Encroachment Duration Per Aircraft for April 28 to May 25, 2024
Trends are easier to detect visually with the moving average plats shown in Figure 4. The chart reinforces the observation about Figure 3 that safety margins over the past month are better than the historical average.
Figure 4. Moving Average Hourly Encroachment Duration Per Aircraft, April 28 to May 25, 2024
Figure 5 shows a longer perspective with three months of data back to February 25, 2024. This chart clearly shows the reduced safety margins for several days in early March that we reported on in our initial posting on April 8, 2024. That early March period had the highest durations and the longest extended period with high durations in the past two months, with metrics well above the 90th percentile for many hours.
Figure 5. Hourly Encroachment Duration Per Aircraft, February 25 to May 25, 2024
Figure 6 displays the moving average version for the past three months, which highlights the early March increase in encroachment durations that persisted for ten days. This is the only sustained period of high encroachments that lasted more than a couple of days for the past three months. April and May encroachment metrics are both much lower than March and below the long run average.
Figure 6. 24-Hour Moving Average Encroachment Duration Per Aircraft for the Three Months Ending May 25, 2024
We wrap-up this week’s report with updates to the two charts presented in the previous weekly reports. Figure 7 shows the total daily counts of the number of encroachments and their durations across the 26 airports in the 17 metropolitan areas that we monitor. The blue bars report estimates of total daily encroachment durations divided by the number of aircraft in the terminal airspace (approximately 50 miles of the airport center) for all 17 terminal airspaces. The red line reports the daily number of encroachment events per 100 aircraft.
Figure 7. Encroachment Duration Per Aircraft and Event Rates for 26 Airports
February 18, 2024 through May 25, 2024
As data are added to the chart each week, readers can begin to detect some mild day-of-week variation and, more importantly, periods of a few days and even a week or longer in which encroachment durations increase significantly. In the coming weeks, we will examine these fluctuations in more detail.
How severe are these encroachments? The FAA defines three separation conformance categories based on how far they are from the separation index. (See the description “How Do We Measure Safety Margins?” for details on the conformance categories and how we measure them.) In that classification system, Conformance Categories A and B are the most severe. Under our definition of a buffer encroachment, Category A and B encroachments are counted under all meteorological conditions.
Figure 8 displays information on the most severe separation conformance categories. The daily durations and event counts for the sum of Category A and B encroachments are shown in Figure 7. The pattern differs from the Figure 1, as there is no obvious trend during the time period. This suggests that Category PE and C encroachments are affected by different factors from Category A and B. We will take a deep dive into those differences in a future weekly report.
Figure 8. Encroachment Durations and Event Counts for Conformance Categories A and B February 18, 2024 through May 25, 2024
Overall, at the NAS level safety margins have been stable for the past several weeks. The industry shows evidence of recovering from the stress of September 2022 to February 2023. We will take a deeper dive into that time period and the subsequent partial recovery in future reports.