AirNav's AI NOTAM Tool and How LLMs Draft Flight Notices

AirNav Indonesia and Telkom University presented an AI NOTAM drafting tool at an aviation safety forum in Arlington, Virginia, that runs through November 6, 2026. The application uses a large language model (LLM) to turn an officer’s plain-language report into a Notice to Airmen (NOTAM) in the standard ICAO format.

A NOTAM is an official notice that tells pilots, airlines, and other aviation users about changes to facilities, services, or procedures, and about hazards that could affect flight safety. Crews and dispatchers make operational decisions based on these notices, so the way they are written matters. That makes AirNav’s use of AI worth following, especially for other public bodies that produce rigidly formatted documents every day.

The tool on show in the United States

According to a Telkom press release dated November 5, 2026, AirNav Indonesia introduced the application during the Aviation Safety and Security Reverse Trade Mission organized by the United States Trade and Development Agency (USTDA). The mission ran from October 26 to November 6, 2026, in Arlington. It brought together representatives from Indonesia and the Philippines to share information and explore cooperation on aviation safety and security technology.

The program included a business briefing on October 27, 2026. In its invitation, USTDA asked US companies to meet senior decision-makers from Indonesia and the Philippines and to present American aviation technology. The forum was therefore a showcase for Indonesian work and also a chance for both countries to look at what the market offers.

Telkom University rector Suyanto described the partnership as an example of university research applied to an industry need. He said he hopes the technology will support efficiency, accuracy, and safety in air navigation services. IndoTelko reported the same details on November 6, 2026.

The collaboration itself is older. Aviation outlet IndoAviation reported that the two institutions began working together in September 2024. On July 28, 2025, they held a soft launch of the NOTAM-AI application and signed a cooperation agreement at AirNav’s headquarters in Tangerang, Banten.

How the AI NOTAM tool works

Writing a NOTAM takes care because the text must follow international aviation formats and abbreviations. With the new application, an officer describes the situation in everyday language. The LLM, a type of AI model trained to understand and generate text, then converts that input into a concise NOTAM that follows the rules of the International Civil Aviation Organization (ICAO).

AirNav gave a concrete example at the soft launch. An officer enters a note that runway 07L/25R at Soekarno-Hatta International Airport is closed for overlay work. The application produces “RUNWAY 07L/25R CLSD DUE TO OVERLAY WIP”. CLSD is the standard abbreviation for closed, and WIP stands for work in progress.

The workload is significant. AirNav Indonesia’s technical director told IndoAviation that the organization issues 50 to 100 NOTAMs a day. He said drafting a NOTAM used to take several minutes and that with AI it can take less than one minute.

In this setup, the AI writes the draft. According to AirNav, a NOTAM still goes through several stages before publication: the event is recorded, the text is composed in international aviation language, the draft is validated, and the notice is distributed to the parties concerned. The validation step is what keeps a model error from reaching the cockpit.

Why NOTAM accuracy matters

NOTAM systems have made global headlines before. On January 11, 2023, the US Federal Aviation Administration (FAA) halted domestic departures from 7:30 a.m. to 9:00 a.m. Eastern time after its NOTAM system failed. According to FedScoop, a preliminary FAA review found that contract personnel had unintentionally deleted files while correcting synchronization between the live primary database and a backup. The FAA said it found no evidence of a cyberattack.

That outage had nothing to do with AI. It does show that NOTAMs are a critical service and that problems in producing or distributing them can stop air traffic. Any new tool added to the NOTAM workflow, an LLM included, deserves the same level of testing as other airport systems.

Details still undisclosed

Several important details are still missing. AirNav and Telkom University have not said which LLM the application uses, where the model runs, or what data was used to train or fine-tune it. No accuracy figures from testing have been published.

Reports on deployment also differ. IndoAviation wrote that version 1 of the NOTAM-AI application was ready for AirNav personnel at headquarters and branch offices. ANTARA, reporting on the same day, wrote that version 1 would be ready after design, testing, and evaluation. The November 2026 press release says the tool was presented, without stating whether it already drafts NOTAMs that are actually published.

AirNav has also said the application still needs refinement and that it continues to coordinate with the regulator on the use of AI. Until more information is released, it is too early to judge how far AI is involved in daily operations.

What this means for government agencies

AirNav’s approach can apply to many public bodies. A transport office issues road closure notices, a regional hospital writes referrals with standard codes, and a licensing office produces decrees in a fixed format. In each case the input arrives as free text while the output must follow strict rules.

1. Define the format and abbreviation list first. An LLM is more consistent when it works from a template and an official glossary, such as ICAO abbreviations in aviation or region codes in local government. Without that reference, a model can produce text that looks plausible but is not standard.
2. Check drafts automatically before a person reads them. Simple rules can reject drafts with empty fields, impossible dates, or abbreviations outside the glossary. These checks catch mistakes that a busy officer can miss.
3. Require sign-off by an authorized officer. An AI draft should be published only after review by someone who understands the context, as in AirNav’s NOTAM validation step.
4. Keep a record of every draft. Log the original input, the model output, the officer’s edits, and who approved the result. That record supports audits and lets the agency measure model accuracy over time.

Questions about where data and models are hosted should be settled early, especially for sensitive operational data. We looked at this issue in our piece on Telkom’s AI ModelHub and agency data control. Agencies that want to start can pick one high-volume document type, then measure drafting time and correction rates before expanding. An AI development team can help design a workflow like this, including the automated checks and the audit trail.

Frequently asked questions

What is a NOTAM?

A NOTAM, or Notice to Airmen, is an official notice about the condition of or changes to facilities, services, and procedures, and about hazards that may affect flight safety. Pilots and airlines rely on NOTAMs before and during flights.

Are NOTAMs now written entirely by AI?

No such claim has been made. The AI drafts text from an officer’s input, and the NOTAM is still validated before distribution. AirNav also says it is coordinating with the regulator on the use of AI.

Will the tool be used in other countries?

That has not been announced. The application was presented at a USTDA forum that brought together Indonesia and the Philippines, but no deployment outside Indonesia has been announced.

Closing thoughts

The AI NOTAM tool from AirNav Indonesia and Telkom University shows how an LLM can support work with strict formatting rules, with AI drafting and people validating. Test results and deployment status are still to come, but the pattern is relevant to many public services. If your agency is considering a similar step, Merakit can help through digital government solutions designed together with your team.

Sources

• Telkom Indonesia: Dukung Akurasi Informasi Penerbangan, Telkom University dan AirNav Kenalkan NOTAM Berbasis AI di AS (November 5, 2026)
• IndoTelko: Telkom University dan AirNav kembangkan AI untuk penyusunan NOTAM (November 6, 2026)
• USTDA: Indonesia and Philippines Aviation Safety and Security Business Briefing (event on October 27, 2026)
• IndoAviation: Kolaborasi AirNav Indonesia-Universitas Telkom, Notam Berbasis AI untuk Layanan Lebih Cepat dan Tepat (July 28, 2025)
• ANTARA Banten: AirNav Indonesia kembangkan aplikasi Notam berbasis AI (July 28, 2025)
• FedScoop: Federal Aviation Administration says NOTAMs outage was caused by contractor deleting files (January 20, 2023)