Darner (@DarnerTech) header
Darner (@DarnerTech) profile picture

Building hybrid VTOL aircraft in Australia. Sharing what we test, what breaks and what we learn. Also into rockets, drones, EVs and AI.

Joined September 2026
Photo by Darner (@DarnerTech)Photo by Darner (@DarnerTech)Photo by Darner (@DarnerTech)Photo by Darner (@DarnerTech)Photo by Darner (@DarnerTech)Photo by Darner (@DarnerTech)Photo by Darner (@DarnerTech)Photo by Darner (@DarnerTech)Photo by Darner (@DarnerTech)Photo by Darner (@DarnerTech)
FPV won the expendable war. Nobody solved reusable speed at real payload. Hexes are slow. Fixed-wing needs a runway. One-ways die once. What’s missing: a VTOL that launches vertical, flies fast, recovers clean, and scales toward 100 kg — without becoming a flying truck. That’s the gap. We’re on it.
682
Dragon is back from its 35th resupply run to the space station. The cargo coming home matters as much as the cargo going up. How many hours from splashdown until the most time-sensitive experiments reach the scientists? teapawt.net/SpaceX/status/21242626…
Splashdown of Dragon confirmed, completing SpaceX’s 35th Commercial Resupply Services mission to the @Space_Station!
7
Satellites that beam 5G to phones could also spot drones. New paper: a ground receiver picks up faint 5G echoes off drones. It sends nothing, so it stays silent. Simulated result: first alert 36.5% sooner, 92.9% of drones found at 5-8 km. arxiv.org/abs/2611.20859
5
Shield AI's test order is smart: run the engine lying flat on the trailer, stand it up and run it vertical, then hang it from a crane in case the engine stalls. Only then fly free. When it lands tail first, how hot does the trailer get under the exhaust? teapawt.net/shieldaitech/status/21…
X-BAT made its Korean debut at Sacheon Airshow 2026. The world's first autonomous VTOL strike fighter drew Korea's defense and aerospace leaders to our booth, where we discussed how to keep pace as missions change, and what a fighter that needs no runway means for the future of Korean airpower. Continue the conversation with us at K-DEX, Booth 7-B21. Autonomy for the world. The greatest victory requires no war.
10
Someone can fake GPS for a whole drone swarm and keep its shape, so the formation check misses it. New paper: the fake transmitters sit a few km away, so each drone sees them at a slightly different angle. Satellites are too far for that. Simulation only. arxiv.org/abs/2611.26023
8
The real step here is the procedures, not the pad. Joby flew routes built with the FAA and NASA, and Orlando's air traffic controllers took part. How many air taxi landings per hour could that tower handle next to airline traffic? teapawt.net/jobyaviation/status/21…
One of this year’s highlights has been becoming the first eVTOL to utilize purpose-built electric air taxi infrastructure and procedures being developed around the world. Here’s a moment from our Florida campaign celebrating our landing at Orlando International Airport’s (@MCO) vertipad, the first of its kind at a US commercial airport. Thank you to Lance Lyttle and the Greater Orlando Aviation Authority team for their years of hard work bringing this vision to reality. Watch the whole story on YouTube ⬇️
31
Eight public numbers from aerospace and physical AI this week: 1. Dubai Marina Vertiport: certified by the UAE GCAA for Joby's planned air taxi route to the airport. Rated for up to 21,000 landings and 84,000 passengers a year. 2. Elroy Air: Barq Group deal for up to 20 Chaparral cargo drones, worth up to $70M. A $10.25M deposit covers 5 firm aircraft. 3. Air White Whale W5000: 30 minute test flight for an uncrewed cargo plane with a 5 ton payload and 2,600 km range. 835 letters of intent. 4. Baykar Kizilelma: first 3 uncrewed fighter jets delivered to Turkish forces after 600 test and training sorties. 5. Hanwha Systems: first US counter-drone radar deal, with Rebellion Industries. The short-range radar detects drones out to 10 km. 6. Rocket Lab: 100th Electron launch, the fastest commercially developed rocket to get there. It was the 11th Electron for iQPS. 7. Genesis AI: $500M Series A at a $3.5B valuation to roll out its Eno robot, first in factories, logistics and labs. 8. TARSBOT and Aptiv: 100 industrial robots expected at Aptiv by year end, across 33 use cases in 8 process types.
2
2
171
A drone loses one of its four rotors. Can it still steer the package hanging below it? New paper (Al Lawati, Sultan Qaboos Univ.; Akhtar, NJIT): a controller built for three-rotor flight keeps the swinging load on its path. Simulation only so far. arxiv.org/abs/2611.15787
18
The useful part is where the processing happens. The sensor sits on the loch floor, turns raw sonar into tracks itself, and sends only those tracks to Lattice. How long can one unit stay on the seabed before it needs new power? teapawt.net/anduriltech/status/212…
On the lookout for Loch Monsters. In this test footage, operators lower Seabed Sentry to the floor of a Scottish loch. The autonomous system activates an acoustic sensing payload, processes sonar data into undersea tracks, and sends them to Lattice. Real-time operating picture for the crew on deck. Monsters beware.
16
Should a hacked drone get to veto its own bad orders? A new paper says no. A separate safety board on each vehicle stopped every collision from forged or hijacked orders. Leave the veto to the drone's own software and it stopped none. Simulation only. arxiv.org/abs/2611.22773
12
X140 catch-up. A blended wing body, two jets, and rotors that fold away for cruise. Before it flies, we're building a hover rig to test the same flight computer and jets. The rig's frame is carbon tubes and printed nodes.
1
14
The plan is to fly the rig from a laptop dashboard over a telemetry radio, with no RC transmitter.
1
7
Exploring air entrainment: a lobed mixer in place of the stock exhaust cone. This one is printed for a fit check. The real one will be printed in Inconel to survive 650°C.
6
Catch-up: we've moved on from the XT7 to the X140. We reviewed the XT7 against CFD and simulated flights. It was complex to build, its fins added too much drag, and it flew 45 min in sim vs the 1.5 h we wanted. The X140 has a 50% better drag coefficient.
2
13
We kept the best of the XT7: the CFD pipeline, flight sim, jet bench testing, parts library and 3D-printed build process. The X140: two turbojets with steerable nozzles and folding lift rotors. Today it flew its first full simulated circuit on the wing, hover to cruise and back.
9
Eve's prototype just flew forward at about 44 knots (roughly 80 km/h) in stable flight. On a lift-plus-cruise design, the hard step is the speed where the wing takes the load off the lift rotors. At what airspeed does Eve expect that handover? teapawt.net/embraer/status/2122797…
Replying to @EveAirMobility
@EveAirMobility’s engineering prototype continues to expand its flight envelope, reaching a peak airspeed of approximately 44 knots during a stable forward-flight demonstration. Watch the video to see the prototype in action.
15
On a weak link, a drone can't send its operator video. New idea: send what happened, not pixels. The drone spots events on board and sends ~25-byte notes. Still worked at 0.3 kbps. Video needed ~100 kbps.* *New data + onboard detector: no gain yet. arxiv.org/abs/2611.19463
9
The US Air Force picked Anduril to build a shipping-container launcher for its low-cost counter-air missiles (CAMP). A box like that can travel by truck, rail or ship like normal cargo. How many missiles fit in one, and how fast does it reload?
The @usairforce has selected Anduril to develop the containerized launcher for the Counter-Air Missile Program. CAMP helps the service field affordable, ground-launched counter-air missiles at greater scale. Shown: Anduril’s internally funded Menace-based hardware, software, and testing that will ultimately build the CAMP launcher.
1
1
42
Counter-drone radar has a blind spot: drone models it never trained on. Rice University paper: passive radar on 5G cell signals, trained on data built from rotor physics. Unseen drones get flagged and sorted by type, no retraining. Simulated so far. arxiv.org/abs/2611.15612
15