PINAX Ai lets airline operations teams follow on-time performance live on one wallboard, see which department leads each controllable delay cause, find the block times behind late arrivals, and start each morning with an emailed operations brief. The wallboard, daily brief, weekly delay report and up to a year of punctuality and delay-code analysis read the same flight records with the same definitions. D15 and A15 sit beside the flight counts behind them.
Email me the brochure, PINAX Ai
PINAX Ai is FlightAtom's airline on-time performance software for operations control, performance teams and airline management. It runs a live operating-day dashboard designed for wall screens, analyzes up to a year of punctuality, delay codes and block times, and emails a daily operations brief and a weekly delay report from the same flight records. Every punctuality figure is shown with the flight count behind it.
Each tile is one figure PINAX Ai computes and the rule or statistic that produces it.
PINAX Ai screens captured in October 2026 from an isolated demo with synthetic records and neutral geography. Open a screenshot for a closer look.
The full-screen wallboard rates the day on the weaker of D15 and A15, here A15 at 75.0%. It puts flight progress, D0 and A0, delay per sector, rule-based actions with owners and the day's delay codes on one screen. Synthetic demonstration data.
Across 365 selected days, every delay code is ranked by coded minutes and assigned once to Engineering, Ground Operations, Flight Operations or external and knock-on causes. The footnote states the controllable share, here 61.9%. Synthetic demonstration data.
PINAX Ai compares each route's actual gate-to-gate time with its published block. YYZ-JFK and JFK-MCO run a median 9 minutes over their published block and get a suggested block that would contain 90% of sectors, while BWI-JFK runs a median 11 minutes under its block and is marked over-blocked. Synthetic demonstration data.
Deterministic k-means clustering groups operating days by punctuality, cancellations, block pressure and delay-cause mix and names each group from what its days share; here 27 days disrupted by departure-station weather carry 22.7% of the delay on departures from New York. The scope can be narrowed by date, airport or city pair, downloaded as CSV or printed as a PDF report. Synthetic demonstration data.
The readiness view counts aircraft on the board, open deferred defects, crew duties within two hours of their duty-period limit and training due. It rates each aircraft's tightest turn, all without naming anyone. Synthetic demonstration data.
For each completed operating day, PINAX Ai produces a PDF operations brief. Page 1 sets the day's technical dispatch reliability beside its rolling 30-day figure, plots each departure's delay against its scheduled time, totals delay by code and station, and lists the sectors that need attention. Synthetic demonstration data.
The wallboard, the analysis, the briefs and the reports read the same flight records with the same definitions, so their figures agree.
D15 and A15 sit beside the flight counts behind them, and the day is rated on the weaker of the two, so strong departures cannot hide late arrivals. Completion, delay per sector, station departure punctuality and zero-tolerance D0 and A0 for the same sectors share the screen, and authorized users can step the board back through the past year.
The wallboard states how much of the day's delay was controllable and names the lead department for each controllable cause: Engineering (40-49), Ground Operations (10-39 and 50-59) or Flight Operations (60-69). External, knock-on and uncoded delay stay visible on their own, and a Friday report groups the week's delay log by department, reconciles coded minutes against observed delay and shows what share of delayed sectors carry a code.
Each completed operating day becomes a PDF brief, emailed at a set hour to each account that opts in, so nobody assembles the morning report by hand. It covers completion, D15 and A15 with D0 and A0, daily and rolling 30-day technical dispatch reliability, every sector with its delay reasons, station punctuality and open deferred defects, then looks ahead to the coming day's sectors with passenger counts, crew cover and a TAF check at each expected arrival.
PINAX Ai compares each route's and flight number's actual gate-to-gate time with its published block, classes it as under-blocked, over-blocked or holding, and counts sectors that left on time but still arrived late. It splits taxi-out, airborne and taxi-in time, and suggests a block that covers 90% of sectors once five or more measured sectors show a route is under-blocked.
Filter up to the past year by date range, departure airport, arrival airport or directional city pair across punctuality, delay, block-time, day-archetype and trend views. Any scope exports as an A4 PDF report of one view or all five, with flight, delay, route and schedule CSV files.
Rule-based checks raise up to three actions at a time, each with an owner and a due time: cancellations, missing movement times, overdue departures, punctuality misses and airport weather from current METAR observations. The board refreshes every minute, keeps the last good picture if an update fails, and keeps the screen awake in full-screen mode.
PINAX Ai can email the pilots rostered on a delayed flight a signed survey link that needs no account and does not show the recorded delay codes. An optional crew delay log, granted one account at a time, lists late departures by roster code, never by name, and ranks a code only after it has flown ten sectors in the window.
The readiness view shows each aircraft's rotation and tightest turn, active and next-seven-day maintenance, open deferred defects, crew duties within two hours of their recorded flight duty period limit, training expiring within 30 and 90 days, and overdue operations tasks, all as counts without crew names.
Administrator, Operations manager and Wallboard viewer presets, or a custom set, decide which views each account can open, and every account change is recorded in an audit trail. Accounts set up as wallboard screens stay signed in while the screen keeps refreshing, and personal accounts confirm sign-in with a six-digit code sent by email. With Keycloak single sign-on, the roles assigned there set PINAX Ai permissions and are rechecked every few minutes.
PINAX Ai uses deterministic calculations and classical statistics. Each measure is defined with its formula in the operator manual, which must be updated in the same change as the code, and the wallboard labels its data as live, limited, stale, unavailable or demo.
Every punctuality figure carries the flight count behind it. Cancellations stay in the delivery denominator, range D15 and A15 are summed from flights rather than averaged from daily percentages, and A15 is shown as unavailable rather than calculated on a smaller base when arrival times are missing.
OTP trends fit flight-weighted lines over 30 days, three months, six months and a year, and label each fit from 'Clear trend' to 'No clear trend'. PINAX Ai withholds the direction when the fit is too weak and states when the fitted line would reach the target only if that date falls within 90 days.
PINAX Ai groups 30 or more operating days by outcome and delay-cause mix using k-means clustering with fixed seeding, so the same dates always give the same groups. Each group is named by its severity within the selected scope and its dominant cause, with how often it occurs and its delay per sector.
Agree a representative sample and a baseline from your current process. These are proposed evaluation measures, not published performance results.
An operating period with a known source and a specific punctuality question.
Trace the denominator, inspect flights, and prepare a management brief.
Report preparation time; source coverage; agreement with independently calculated metrics.
D15 and A15 are the shares of flights that depart or arrive within 15 minutes of schedule. PINAX Ai shows each beside the flight count behind it, adds zero-tolerance D0 and A0 for the same sectors, and rates the day's punctuality on the weaker of the two. When arrival times are missing, A15 is reported as unavailable rather than calculated on a smaller base.
PINAX Ai compares each route's and flight number's actual gate-to-gate time with its published block. A route whose median runs 5 or more minutes over its block is under-blocked, and one 10 or more minutes under is over-blocked. When at least five measured sectors show a route is under-blocked and one published block applies across the range, PINAX Ai suggests a block that would have covered 90% of them, rounded up to the next 5 minutes.
PINAX Ai assigns each delay code once to an accountable group: Engineering for codes 40-49, Ground Operations for 10-39 and 50-59, Flight Operations for 60-69, and external or knock-on causes for other numeric codes, including 70 and above. Records without a code are reported as uncoded. The wallboard and the weekly report use the same mapping to separate controllable from external delay minutes.
The daily brief is a PDF for each completed operating day. It covers completion, D15 and A15 with D0 and A0, delay per sector, daily and rolling 30-day technical dispatch reliability, every sector with its delay reasons, station punctuality, open deferred defects by aircraft, and the next day's sectors with passenger counts, crew cover and a TAF check at each expected arrival. PINAX Ai can email it at a set local hour, 07:00 by default, and Monday's edition covers Friday to Sunday.
PINAX Ai counts a technical incident when a flight leaves the gate 15 or more minutes late with a delay code beginning with 4. Technical dispatch reliability is the share of departed sectors without such an incident, reported for the day and for a rolling 30 days. Cancelled sectors are measured separately through completion.
Yes. PINAX Ai can email the pilots rostered on a delayed flight a private, signed survey link that is valid for 30 days and needs no account. The survey asks what they saw without showing the recorded delay codes, and each response is kept with a snapshot of the flight. Requests go out manually or automatically from a crew directory the operator maintains.
PINAX Ai reads schedules, aircraft, delay codes, OUT, OFF, ON and IN times, cancellations, crew assignments, and readiness records such as maintenance, deferred defects, crew duties and training from myairops. Current METARs and TAFs come from the Aviation Weather Center. Flightradar24 positions can refine flight progress without changing movement times or KPIs, ForeFlight can add passenger counts to the day-ahead brief, and Open-Meteo can add a seven-day destination outlook. PINAX Ai is a standalone product and does not require AURORA Ai.
No. PINAX Ai's measures, briefs and reports are deterministic calculations. Its OTP trends use flight-weighted linear regression, and its day archetypes use k-means clustering with fixed seeding to group past operating days by outcome and delay cause. Both describe recorded performance; PINAX Ai does not forecast punctuality.
As your fleet, routes, rules or teams change, we adapt PINAX Ai to match, and you approve each change before it goes live.
A focused walkthrough, with the people and decisions that matter to your team.