David

@puca.bsky.social

Wolkenjägerblicker

@puca.bsky.social

This period remains, perhaps, the most interesting of the 2026 test season - so far. Vessel layout wise, basically a repeat of activity from the only confirmed test last year. The presence of Rossita in Severny at same time is interesting. There were other vessels that didn't transmit on AIS.

@puca.bsky.social

Novaya Zemlya, showing the Pankovo and Severny test sites. I don't think it renders as well in Copernicus Browser anymore; looks a bit screwy with the shadows - whatever algorithm they're using now.

@lemonodor.bsky.socialReposted by @puca.bsky.social

More explorations using my custom AI tool: People on social media were talking about how many E-6Bs were airborne at one time, so I built a heuristic classifier to identify E-6Bs and looked at the trend over time.

Two charts of how many E-6B aircraft were airborne at the same time each day. Top: a calendar heatmap with one row of weeks for 2025 and one for 2026. Each cell is one UTC day, shaded from dark brown to bright orange by that day's peak count of E-6Bs airborne at once. 2025 is complete except for a blank stretch in late May and early June, where days are missing from the archive. 2026 runs from January to late August. The brightest cells are in late November 2025, February to March 2026, and mid-August 2026. Bottom: a stacked bar chart titled "Daily peak simultaneous airborne E-6B aircraft (UTC days)", covering August 2025 to August 2026, on a scale of 0 to 5. The blue part of each bar is aircraft confirmed by known fleet hex codes and labels. The orange part is additional probable aircraft flagged by the heuristic rule. Most days peak at 1 to 3 aircraft. Before a dashed line in early April 2026 the bars are mostly blue. After it they are mostly orange, with only 0 or 1 confirmed. A few days reach 5: one in September 2025, a cluster in early December 2025, and several in August 2026, when the peak stays at 4 or 5 for a run of days, the highest sustained level on the chart. Below the bars, a gray line shows archive coverage. It rises and falls on a regular cycle, with a dip around late December 2025.
@lemonodor.bsky.socialReposted by @puca.bsky.social

I've been working on a AI tool focused on aviation investigation and analysis. I think it's incredibly sick, the coolest thing I've done since GPSJAM, and I use it every day. I'm going to post some examples of how I've used it lately in this thread. This is the collision at LAX yesterday.

Satellite view of an LAX gate area with aircraft outlines overlaid. A red silhouette shows Delta flight DAL7, an A350-900 (N512DN), at the moment of first contact. It's taxiing left to right along taxiway C, and its left wingtip touches the tail of a blue silhouette, Air Canada C-FIVK, parked nose-in at gate 69A. A yellow dot marks first contact at 17:29:18.4Z. Small red dots trace DAL7's path at 29:02Z, 29:09Z, 29:16Z and 29:19Z, and a dashed red outline shows where it came to rest. Below the map are key findings: C-FIVK stopped 15.9 m short of where the same aircraft parked on 9 September, while Air Canada 787s and Hawaiian A330s normally stop within 1 to 2 m of that spot. DAL7 held the taxiway centreline (+0.1 m, against a median of +0.4 m across 198 A350 passes) and moved 3.5 m right in the last 15 m before stopping. First contact is estimated at 17:29:18.4Z, with a 95% interval of 17:29:17.1Z to 17:29:19.8Z.
@dpa-intl.comReposted by @puca.bsky.social

#BREAKING August Hanning, the former head of Germany's BND foreign intelligence service, has been arrested on suspicion of espionage, sources tell dpa

Older posts