The Flights Caught Up to the Pilot Program

Last Friday this digest reported that Beta's announced Sugar Land to Galveston route had not appeared in the observed record. On September 2 it did. The same CX300 flew Brenham to Sugar Land to Galveston and back, four legs, all three announced Texas test sites in one afternoon. That afternoon a Joby aircraft logged its first Texas flight, a local sortie at Alliance Airport in the Dallas area. The FAA confirmed the first remotely piloted test under the pilot program, at Houma, Louisiana. Archer announced a multistate flight tour. All 25 market scores held. This is the week announced activity became observable, and the week the platform updated its own prior reading because the evidence changed.

What Changed This Week

Four developments, ranked by what the platform can independently verify. Observed means the aircraft broadcast and the ledger holds the track. Confirmed means a federal source stated it and the platform has no track. Announced means a company said it will happen.

The Route We Reported as Unobserved

Issue 22 followed N401NZ 2,765 miles from the Vancouver area to Houma and noted that the Texas announcement and the transponder record disagreed. Texas DOT and the FAA had said the CX300 would connect Galveston, Sugar Land, and Brenham. As of last Friday morning the ledger showed two local sorties at Galveston and no leg to either of the other two sites. That reading was accurate on the day, and it is no longer the reading.

The aircraft left Houma on August 27 for New Iberia, crossed to De Quincy and then Cleveland, Texas on August 28, and landed at Brenham on August 29, three more hops in the 80 to 100 mile shape the whole transit has kept. It then flew seven local sorties out of Brenham over four days. On September 2 it flew the route: Brenham to Sugar Land, 39 minutes; Sugar Land to Galveston, 27 minutes; Galveston back to Sugar Land, 33 minutes; Sugar Land back to Brenham, 41 minutes. Four legs, 228 miles, two hours and twenty minutes in the air, and every one of the three announced sites touched in a single afternoon. It is the first day in the ledger on which the Texas program looks like the program that was announced. As this issue closed, the September 3 sweep added one more leg: Brenham to Houston Executive Airport, west of the city, 23 minutes.

The platform records what this does and does not change. Brenham to Sugar Land and Galveston to Sugar Land now exist as observed route pairs, two legs each, one aircraft, one day. Neither carries a corridor identifier, because identifiers are minted only when a pair is flown on at least two separate days. Eighty four of the 375 pairs the ledger has ever recorded have crossed that line. A second day on either Texas pair would add to them, and that is the first thing to watch next week. The larger point is the one the lede makes: the prior reading was not wrong, it was dated, and a record that cannot be revised by new evidence is not a record. This issue is the revision.

The Engine Caught the Sugar Land Landing

Last week the inference layer opened situations at hospital pads near Galveston and Houma after N401NZ landed there, without anyone tagging the aircraft. On September 3 it did the same thing around Sugar Land. It opened twelve situations across the Houston metro, eight of them at hospital pads that cannot take the aircraft now flying past them: CHI St. Luke's in Sugar Land, Oak Bend in Richmond at 20 feet, Memorial City General at 40 feet, Memorial Hermann Southwest, Memorial Hermann Pearland, the Women's Hospital of Texas, Texas Children's West Campus, and the Houston VA. Each cites the same evidence, four observed flights at Sugar Land Regional inside fifteen nautical miles, against a pad on record under the 50 foot eVTOL landing floor. The other four are records the layer suspects are mislocated, now with traffic nearby. The rules fired on the ledger and the federal facility record alone.

Houston's score is 53 and did not move. The classifier refuses credit for a test flight unless the source names a designated program, and the observed ledger does not feed the score directly. So the market whose hospital pads the engine flagged twice in two weeks is the market the scoreboard has not repriced. That gap is deliberate, and it is where the next Texas score change will come from.

Observed, Confirmed, Announced

Joby, observed. Joby said on August 7 it would base its Texas pilot program flights at Alliance Airport in the Dallas area and fly piloted routes across the metro for a week in September, routes unannounced. On September 2 the ledger recorded N545JX, a JAS4-1 registered to Joby, flying a ten minute local sortie at Alliance. The airframe's 67 previous logged flights, going back to June 2025, were all at Marina, Monterey, or Salinas. It is the first Joby track in Texas the platform has ever recorded, and it is a local flight, not a route. The five other Joby flights this week were N542JX at Marina.

Elroy, confirmed. The FAA said on August 31 that Elroy Air and the Louisiana Department of Transportation and Development flew a hybrid electric cargo demonstration at Houma, and called it the first remotely piloted aircraft test flight under the program. Elroy's own release on September 2 called the late August series autonomous and uncrewed, flown with LIFTOFF Louisiana and Bristow, carrying parcels, medical supplies, spare parts, food, and water. The platform holds no track for any of it. Elroy's Chaparral is not on the FAA eVTOL airframe roster the ledger sweeps, so this is a federal confirmation, not an observation, and this digest keeps that distinction visible rather than blending the two.

Archer, announced. Archer's September 3 release says the No Roads tour follows more than 70 test flights in August and will begin with a flight to Hollister, then four more Northern California stops ending in San Francisco, before extending to the Los Angeles area, Texas, and Florida on no stated schedule. The ledger shows Archer's five flights this week were all local sorties at Salinas on August 28 and 29 and nothing since. The news layer read the release and mapped two overrides to San Diego at medium confidence, a city the tour does not name. Both are at the audit gate, unpublished. That is the gate doing what it is for.

The Weekly Operating Board

Observed flights cover August 28 through September 2 as the ledger stood this morning. Counts backfill for about a day; last week's figure now reads higher than it did when published.

The standing caveat applies. Observed flight is a floor, never a census. Experimental aircraft may legally fly without broadcasting, and a missing track is recorded as unobserved, never as proof an aircraft did not fly.

Markets to Watch This Week

Top three by predictive significance, from the AirIndex Forward Signals pipeline as of September 4.

See full ranked digest →

What to Watch Next Week

• Joby moving from a local sortie at Alliance to city to city legs across the Dallas area. Joby said a week of piloted routes in September. The ledger will count them as they broadcast.
• Archer's first observed tour leg, Salinas or Marina to Hollister, and then whether the Bay Area sequence produces distinct city pairs or repeats one.
• A second day on Brenham to Sugar Land or Galveston to Sugar Land. That is the threshold for a corridor identifier, and it would be the first minted from a pilot program route.
• The six overrides at the gate. Two are Archer tour items mapped to San Diego, two are Joby's Texas facility and its Vandalia, Ohio acquisition at high confidence. Whether any publish is next week's first score question.

UAM Market Pulse is a weekly intelligence digest from AirIndex. Observed flights come from the AirIndex eVTOL movement ledger, built on FAA registry data and public ADS-B broadcasts. Forward signals are derived from the AirIndex Forward Signals pipeline aggregating classifier outputs, MarketWatch trajectory, and predevelopment facility milestones.

AirIndex · UAM Market Readiness Intelligence · airindex.io