The question
Can separation by declared flight level retain its apparent benefit when craft are wrong about their own height?
The flight
The CITY campaign compared ordinary handling with altitude assignment at two report-validity settings, on a simulated 20 m crossing. The experiment varied own-height error and assumed communication tiers. These are controlled comparisons, not flights over New York.
What happened
No contacts occurred in the 3,456 exact-height runs. With 4 m own-height error, contacts appeared mostly with assignment and most with the longer report validity. On calm L2, the stock, 2-second and 8-second arms recorded 121, 258 and 573 contacts. In the 8-second arm, 1,146 of 1,150 lost craft were contact losses. Aggregate loss alone concealed the change in failure type.
Mid-air contacts · calm L2 · 4 m own-height error
Descriptive contact counts, not risk estimates. CITY cycle 34; proposed claim rewritten as revision c35 (amended 2026-10-10), awaiting the cycle-36 review.
What we learned later
Amendment (2026-10-10, after the CITY cycle-35 review). This observation has now been reviewed: both reviewers called the cycle-34 step flawed, its findings were fixed, and the counts above were re-screened and stand. Its proposed map sentence was disputed by all three reviewers and rewritten (revision c35) for the cycle-36 review, so the claim is still provisional rather than a recorded finding. Re-screened counts: the 8 s rule’s excess over ordinary handling was +182 contacts on calm L1, +452 on calm L2 and +36 in the descending L1 case, all beyond noise; at a 2 m error calm L2 already had +12. A contact is an overlap of the two craft’s padded safety spheres, not a demonstrated airframe impact. CITY 35 then isolated the cause on 48 new layouts: the excess replicated (+148 and +374), and nearly every contact followed a grant made while the pair was already truly inside its safety distance. Withholding exactly those grants with perfect knowledge (an oracle no craft has, so a diagnosis rather than a remedy) brought calm L1 back to ordinary handling’s count and most of calm L2, keeping a smaller loss reduction; that result is itself provisional until the cycle-36 review. A guard built from the craft’s own estimates was then tested (CITY 36, next note).
research/city/design/M-CITY-ASG-CONTACT-record.md
Amendment (2026-10-10, CITY cycle 36, provisional). The estimate-only guard this entry’s question led to has been measured. A guard that reads only the two craft’s own height estimates and an assumed declared error bound (4 m per craft, the error’s own standard deviation, not a worst case) was built into the simulator, off by default, and scored on 48 new layouts against ordinary handling, the unguarded 8 s rule and the cycle-35 oracle. On the calm crossing with the weakest link (L1) a 6 m allowance brought contacts from 188 to 10 (ordinary handling 6, oracle 5; the +4 is within noise) while still losing 107 fewer craft than ordinary handling, at the price of withholding 7.2 per cent of grant halves, of which about one in four was truly too close at that moment. On the better link (L2) and when a craft is pushed down, no allowance removed the contact excess while keeping the loss saving (L2 at 6 m: 185 contacts against 121, the oracle 146; 1,065 of the 8 s rule’s 1,549 fewer losses kept). The remaining contacts are pairs whose estimates lean apart while the craft are truly close, which no estimate-only rule can see. Nothing is adopted; the guard stays off. The proposed map sentence awaits the cycle-37 review. Alongside the guard, switching off speed-matching with the other height level (the alignment term LINK 29 identified, an existing switch) cut ordinary handling’s lost craft from 6,027 to 3,852 on L1 and from 2,761 to 270 on L2 (wall strikes 2,519 to 30) with contacts unchanged, and the 8 s rule’s from 5,820 to 3,626 on L1: LINK 29’s mechanism seen again, now with own-height error. On L2 the registered comparison (the 8 s rule with and without the switch) is empty by construction, because under that rule every cross-level pair is granted before the alignment term acts; the ordinary-handling comparison carries the result there. LINK 29’s caveats carry: one crossing geometry, band exits rose there, throughput was not measured, and nothing is adopted.
research/city/design/B-CITY-ASG-GUARD-record.md
Follow-up (2026-10-11, CITY cycle 37, provisional). The speed-matching result reported in the previous note was tested beyond its single crossing on 48 fresh layouts: with exact height, on wider lane spacing and in the demonstration fleet, switching the cross-level term off kept cutting lost craft (for example 2,471 to 16 on L2 with exact height; 91 to 4 in the fleet under the 4 m error), with no change in contacts detected. Its price is band-keeping: about 4 to 6 per cent more band exits per second of flying at the crossing, and under the 8 s assignment with exact height about a third more time outside the band. Nothing is adopted; see the CITY 37 entry.
research/city/design/M-CITY-ASG-ALIGN-SCOPE-record.md
What this does not establish
One crossing, assumed error process and communication settings. At the 2026-10-05 snapshot the proposed claim awaited cycle-35 review (see the later note). The experiment is consistent with granted pairs skipping avoidance, but does not isolate that cause or establish a remedy.
The next question
No operational validity policy or safe spacing follows from this result. The next question was which assignment decisions preceded the contacts; CITY 35 answered it in part (see “The pass that was already too close”).
Source references for reproduction
Optional repository references, pinned to this snapshot. Repository access may be required.