AD 2026-17-02 adds repetitive ultrasonic and detailed inspections at certain 787 lower side-of-body splice plates after shim-gap findings. The directive is necessary. FLEX AI™ does not replace this AD, and it does not inspect a factory shim.
The FAA just put another inspection on the calendar.
This one is not a door cutout. It is a wing joint.
I am not arguing with the directive. I am arguing with the idea that a clock is the only memory the structure gets.
Last week's Federal Register added a new look at a known location on a known family. Operators will inspect. They will repeat. The paperwork will track the interval. That is how airworthiness still works when structure talks in a way the system can already name.
AD 2026-17-02 covers certain Boeing 787-8, 787-9, and 787-10 aircraft. A Boeing investigation found shim gaps at the lower side-of-body splice plates that may have exceeded engineering allowances. High pull-up forces at those fastener holes can start fatigue cracks. Operators are now directed to run repetitive ultrasonic and detailed inspections. The rule is effective October 1, 2026. The Federal Register notice is 2026-17553, published as 91 FR 55247 — https://www.federalregister.gov/d/2026-17553.
The FAA estimates 17 U.S.-registered airplanes and up to 286 work hours per inspection cycle. In plain language, that is a scheduled look at a wing/body splice location after a factory-related finding the system already has in hand. Inspect. Repeat. Repair on condition. Meet the October 1 clock.
This article does not claim that FLEX AI™ found these cracks. It did not. It does not claim that FLEX AI™ inspects a factory shim. It does not. It does not claim that FLEX AI™ is an approved substitute for this inspection. It is not. The directive exists because a specific structural location on a specific family required a specific look. That work has to be done.
A few days earlier, the industry was talking about a different calendar inspection — a 737 forward galley door bear-strap finding, another bulletin, another hangar look. That story was about a door cutout on a different family. This story is about a 787 wing/body splice and about how the inspection math itself already knows that flight length matters.
Mixing those two directives into one lesson would be lazy. They share a method — finding, bulletin, clock — but they do not share a location, an airplane family, or the same argument about what the structure already lived. Keep them apart.
One operator asked to simplify the compliance time to hours or cycles. The FAA said no. The splice plates are flight-length sensitive. How long the flight is, and how many times you take off, both matter. The inspection math already has to track the life the airplane actually lived.
That sentence is the whole industry argument in miniature. The airworthiness system will not let operators pretend that a short hop and a long over-water sector load the same joint the same way. The directive refuses a simplified clock because the physics of the splice already refuse it. Fatigue at those fastener holes is not a pure cycle counter. Duration matters.
And yet the trusted tool for acting on that truth is still a repetitive hangar look. Ultrasonic. Detailed visual. Interval. On-condition repair. The system knows flight length matters. It still treats the continuous memory of that life as something you reconstruct after the fact, when the next inspection card comes due.
That is the part aviation still treats as paperwork.
There is a version of commentary that treats every new airworthiness directive as proof that certification failed. That is not this argument. AD 2026-17-02 is doing the job an AD is designed to do. It takes a Boeing investigation finding, names the airplanes, names the lower side-of-body splice plates, and requires repetitive ultrasonic and detailed inspections with on-condition action. Operators who fly those 787-8, 787-9, and 787-10 aircraft now have an October 1 effective date. They should meet it.
The better question is whether flight-length sensitivity should only ever show up as a reason to refuse a simplified compliance clock — instead of as a reason to keep a continuous structural history while the airplane is still flying. If the only trusted memory remains the inspection card, then the industry will keep writing more cards. Each new look will be justified. Each new look will also be another block of hangar hours pointed at a joint the structure already lived through without a continuous record.
I would rather argue for a living structural history than for a fight with a necessary directive. The directive closes a known risk at a known splice. A living record is about the rest of the airframe, and about the flights before the next finding is long enough to see on a scheduled look.
A continuous structural history is not a photograph of a crack at a fastener hole. It is a record of what the airframe actually experienced — strain, load, the life it lived — collected while the aircraft is flying, so that engineering and maintenance are not starting from zero at the next ultrasonic pass.
That record is useful before a bulletin exists, not only after. It is useful across the whole airframe, not only at the lower side-of-body splice. It is useful for the life of the airplane, not only for the interval AD 2026-17-02 just added. The intended direction of that work is to eventually replace scheduled structural inspection as the primary way operators learn what an airframe has lived. Eventually is doing real work in that sentence. It is a goal for how structural knowledge is gathered. It is not a claim that this AD, or any AD, can be skipped.
FLEX AI™ is built to collect and trend that history while the aircraft is flying — not to replace this AD, and not to inspect a factory shim. That is the product Platinum Eagle Aerospace is building. The sentence is intentionally modest. Collect. Trend. History. While flying. It does not say the platform found the shim-gap fatigue risk. It does not say the platform replaces the ultrasonic and detailed inspections the FAA just required. It does not say the platform is already an approved change on these 787s.
The FAA STC is in process and will be finished before the end of 2026. That is the approved public language, and it is the only certification claim this article makes.
In process is not issued. Will be finished is not a guarantee written into anyone's maintenance program. An STC, when it exists, is the path to putting a living record on the airplane as an approved change. Until that path is finished, FLEX AI™ is not a substitute for anyone's inspection program, and it is not a substitute for AD 2026-17-02.
That does not replace this AD. Operators still inspect. They still meet the October 1 effective date. They still run the repetitive ultrasonic and detailed looks the directive requires. Nothing on this page is permission to do otherwise.
The useful industry question is not whether this directive should exist. It should. The useful question is how many more inspection clocks we write before we treat flight-length as something the structure already knows — not only as a reason the FAA refused to simplify a compliance table.
AD 2026-17-02 will do its job on the 787-8, 787-9, and 787-10 aircraft it covers. The larger job is still open: keep a continuous memory of what the structure was doing before the crack was long enough to see, across the whole airframe, for the whole life, with flight length treated as lived history rather than as a footnote in the compliance math. That is the work FLEX AI™ is built for. It does not replace this AD. It does not inspect a factory shim. It is the argument for not making another calendar look the only tool we trust the next time a splice starts talking.
Learn more about FLEX AI™ structural health monitoring.