1 Problem description - Delaying methods on approach
In order to detect potential anomalies in airports, we would focus on understanding the arrival delay introduced in the vecinity of the airports, via any type of holding patterns. In airports where those anomalies introduced via those flight patters need to be automattically flagged, looking at the main parameters that identify those anomalies. What matters is how arrivals are delayed on approach until a runway is available, since runways are typically the system’s main bottleneck.
Large airports manage this delay in several ways:
Vectoring. At low traffic levels controllers simply shortcut routes. At some busy hubs, even at high density, radar vectoring remains the standard method (60%)
Point-Merge. An increasingly common technique: aircraft fly a sequencing arc and leave it at the assigned time—simple for controllers to operate (24%)
Linear path-stretch. Published STARs can incorporate extended downwinds or “trombones” to add distance and absorb delay without classic holding (17%)
Conventional holding stacks. Fixes are still published in the AIP but now serve mainly as contingency options rather than the primary tool.
The focus is therefore on identifying which of these methods each airport uses when operating near capacity and the delays they can cause.
2 Our goal: Computing additional arrival sequencing and metering area (ASMA) time
The most widely recognised key performance indicator for quantifying arrival delay inside terminal airspace is Additional ASMA Time. This metric was developed within the EUROCONTROL Performance Review Unit (PRU) and is incorporated into the Single European Sky (SES) Performance Scheme as well as the ICAO Global Air Navigation Plan performance framework.
“Additional ASMA Time is the average time that arriving aircraft spend flying in the Arrival Sequencing and Metering Area (ASMA)—a 40 NM radius cylinder around the airport—in excess of a statistically derived unimpeded reference time, grouped by runway, arrival flow and aircraft category. It is expressed in minutes per arrival.”
The ASMA is defined as a cylinder of 40 nautical miles radius centred on the Airport Reference Point (ARP). For each arrival, the ASMA transit time is calculated as the period between first entry into this 40-NM cylinder and touchdown. A reference “unimpeded” time—derived from operations under low-traffic conditions and stratified by runway configuration, entry flow, and aircraft type—is then subtracted from the actual transit time. The resulting value, in minutes per arrival, represents the additional delay incurred in the terminal manoeuvring area.
This KPI directly reflects controller-induced sequencing actions such as vectoring, extended downwinds (“linear path-stretch” or “trombone” procedures), or point-merge operations. Unlike en-route efficiency indicators (e.g. KEP/KEA) or ATFM ground-delay measures, Additional ASMA Time isolates delay that occurs strictly in the approach phase and is therefore well suited for projects assessing runway-capacity constraints or approach-management techniques.
Official methodology and current European performance data are maintained by EUROCONTROL at https://www.eurocontrol.int/prudata/dashboard/metadata/additional-asma-time/ and summarised annually in the Performance Review Report. For instance, in the latest report: “Additional ASMA times at T₃₀ in 2024 showed a slight increase compared to 2023 but remained below the levels recorded in 2019. The highest additional ASMA times were observed at the London airports with Heathrow averaging 7.48 min/arr and Gatwick at 6.02 min/arr. Lisbon followed showing a significant increase and reaching 4.87 min/arr (+1.41 min more than in 2023). This deterioration of the ASMA times was mostly observed since the implementation of the Point Merge procedure in mid-May 2024, alongside a reduction of the arrival ATFM delays at this airport.”
In order to generalize, Eurocontrol uses RWYs in use and specific entry sector for each airport for the ASMA calculation: “Entry sectors are defined for each airport according to the main arrival flows and may be revised when traffic patterns change. Aircraft are assigned to the sector corresponding to the point where it first crosses the ASMA boundary, regardless of the filed procedure.”
Sector definition and reporting specifications
3 Variables to consider
Input Data
Airport list and ARP coordinates – one ARP (latitude/longitude) per airport.
Runway geometry – for each runway end: threshold coordinates and published approach course.
(No information on runway‐in‐use is assumed.)
From ADS-B
Callsign, Position, Altitude, Cleared altitude, Heading, Ground speed, Type (L, M, H, HEL)
Arrival Sectors
Divide the 40 NM Arrival Sequencing and Metering Area (ASMA) around each airport into fixed angular sectors referenced to the ARP.
Use a uniform azimuth grid (e.g. 12 sectors of 30° each) measured from true north.
Sector boundaries remain constant over time and are independent of runway configuration to ensure stable sampling even when procedures or STARs change.
Runway Inference (problem already solved in DataBeacon for RWY detection in LEMD)
For each arrival, estimate the likely landing runway or runway group using the last 20 NM of track:
Compare the aircraft’s final course and lateral proximity to each published runway centerline.
Select the runway (or parallel group) with the smallest heading error and cross-track distance.
Store a confidence score (High/Medium/Low) based on heading agreement, cross-track distance, and threshold proximity.
Alignment Classification (Reporting Layer)
After runway inference, compute the bearing from the ASMA entry point to the ARP (θ_in).
Let θ_rwy be the inferred runway approach course.
Calculate the absolute angular difference Δ = min(|θ_in – θ_rwy|, 360 – |θ_in – θ_rwy|).
Assign an Alignment Class:
Aligned: Δ ≤ 60°
Cross: 60° < Δ ≤ 120°
Opposed: Δ > 120°
This alignment is used only for reporting and does not influence sector boundaries or the ASMA time calculation.
KPI Calculation
Detect ASMA entry time (first crossing of the 40 NM circle) and touchdown time from ADS-B surveillance. Interpolate between 5-min points on a geodesic.
Compute ASMA transit time = touchdown time – entry time.
Calculate an unimpeded reference time. See Annex.
The final metric is Additional ASMA Time = actual transit – reference, reported in minutes per arrival.
4 MVP0
To develop the Minimum Viable Product (MVP0) for mesuring ASMA
Select Dataset: Use data from Madrid/Barajas airport, high quality ADS-B data from DataBeacon’s network.
Choose Timeframe: Select a specific day and time range for the measuring and detecting additional ASMA time.
Visualize in Bravo5: Display the information in Bravo5 (replay), highlighting the moments where additional ASMA time were found. Direct links from the anommanly detection to the replay tool.present without labels, focusing solely on flights.
Full development:
Define the Scope: Look into 200 top airports of the world
Select Time Slot: Continuous monitoring
Access Dataset: Data streamed directly from ADS-B Exchange.
Reporting: continuous reporting through a dynamic table.
Visualize Results: Generate links to our replay tool with ADS-B exchange data (probably interpolated)
Produce Report: Compile a daily, weekly report summarizing the simulation results.
Annex: Additional considerations
1 Unimpeded Reference Time Calculation
While Eurocontrol derives the unimpeded reference time from statistical samples of low-traffic flights (typically night-time operations), a more precise calculation is possible when the exact ASMA entry point and the subsequent interception of the ILS—or alignment with the runway centreline—are known.
The unimpeded reference time is derived by estimating the minimum feasible transit from the ASMA entry point to the identified runway threshold.
The lateral segment follows the shortest track that allows a standard-rate turn and proper alignment with the runway centreline.
For the vertical component, the altitude to lose between the entry point and the final approach altitude is divided by an assumed continuous descent rate (e.g. 2 000 ft per minute).
If the horizontal distance of the lateral path is insufficient to achieve this descent—meaning that the required vertical distance divided by the assumed vertical rate would demand more time than the lateral leg provides—the path length is extended until the available horizontal distance supports the necessary descent.
The total unimpeded time is the sum of the lateral travel time, based on entry and final approach ground speeds, and any additional time required to meet the vertical descent constraint.
Why this is better for our global use-case
Mechanical and airport-agnostic: no STARs, no custom sectors, no per-airport tuning.
Per-flight: respects where and how each aircraft enters the ASMA (heading, speed, altitude).
Physically flyable: bounded by turn mechanics and descent geometry, so it won’t invent impossible “direct-to ILS” times.
Comparable: same rules everywhere; runway knowledge improves only the orientation, not the logic.
Handling the 5 min data constraint
Entry time: linearly interpolate the boundary crossing; typical error is ≤2–3 min at 5-min cadence—acceptable at KPI scale.
Touchdown time: infer with a speed/position heuristic; if last in-air point is >5 min before touchdown proxy, interpolate on the descent segment.
Flag arrivals with >3 min interpolation uncertainty as low-confidence (keep for aggregates if sample size is limited, or exclude for “strict” reporting).
2 Time absorbed en-route
Arrival Management (AMAN) systems influence the measurement of Additional ASMA Time (ASMA) because they shift part of the sequencing delay from the terminal area into the en-route phase.
When an AMAN is active, aircraft are given Controlled Times of Arrival (CTAs) at defined metering points. To meet these CTAs, crews typically reduce speed during the descent or in the upper portions of the cruise profile. As a result:
Less delay is absorbed inside the 40-NM terminal cylinder, which is the area used to calculate ASMA.
The aircraft reaches the ASMA boundary closer to its scheduled landing time and often requires shorter vectoring or path-stretching before intercepting the ILS.
The overall arrival delay may be unchanged, but a larger share occurs earlier in the flight, outside the ASMA measurement window.
For a global project, this means that ASMA values cannot be interpreted in isolation.
An airport with a well-tuned AMAN may show low Additional ASMA Time even when total arrival delay is significant, because a portion of that delay has been strategically absorbed en route through speed control rather than in the final 40 NM.