kristian1109

4 ms·

1.
by kristian1109 2y ago
> Do you think your model's word error rate would decrease if it had access to raw feeds from towers or was trained on data from even more airports? Yes, especially for local things like navaids, cities, etc. Our expectation is that
2.
by kristian1109 2y ago
There's lots to discuss about the most recent Potomac River crash, but I'll let that rest for now. Your general point is 100% correct. It's not enough to only do the command parsing; you have to also compare that to radar + t
3.
by kristian1109 2y ago
Sure, feel free to email founders @ ... One of the best parts of this job is how awesome the labeling community is. So many super qualified folks like yourself. Thank you!
4.
by kristian1109 2y ago
But only after 1am when you're not fighting a 15 knot headwind with an A320 cleared number two.
5.
by kristian1109 2y ago
A lot of drones these days will have ADS-B. The ones that don't probably have geo-fencing to keep them away from airports. There's also all kinds of drone detection systems based on RF emittance. The bird problem is a whole other
6.
by kristian1109 2y ago
At an ASDE-X equipped airport like DCA, there is surface radar that the tower manager, for example, might be keeping an eye on. This is just some screen with a high-res view of everything on the ground. Then you "notice" the mista
7.
by kristian1109 2y ago
Not to spoil the reveal here, but ... yes. We're working on compiling more of these cases and will probably publish a website or paper when it's solid.
8.
by kristian1109 2y ago
A lot of glass panels these days will do this in the PFD. You get your clearance via datalink from ATC, it loads everything right up, and you just keep the plane in the box or turn the autopilot on once you're in the air. Of course, th
9.
by kristian1109 2y ago
Speaking as a pilot for a moment, I think your instincts are correct in theory but hard to actually implement. In a critical function like control, you don't want to split a pilot's attention. You wouldn't want them to someti
10.
by kristian1109 2y ago
Good callout — we're being a little vague to the public while we build out these first use cases (HN doesn't count as the "public"). Expect more detail soon!
11.
by kristian1109 2y ago
Good callout, thanks
12.
by kristian1109 2y ago
Exactly, this is one of our early beachheads. If we can train better pilots, the airspace system gets safer immediately.
13.
by kristian1109 2y ago
No stress, appreciate you chiming in. > We've accelerated past our capabilities and need to slow down. This is a super interesting meditation. As much work as there is to be done now, the demand for air traffic is growing and power
14.
by kristian1109 2y ago
Interesting paper, thanks for sharing here. Encrypting ADS-B is a whole discussion... in my opinion, it's a good thing to have largely public location data for folks to consume and study. What were your main takeaways from this paper?
15.
by kristian1109 2y ago
Short answer: both. Pitching air traffic management is going to be a years long process. Getting certified for on-board avionics is similarly challenging. In the meantime, we'll get better and better at monitoring the airspace system a
16.
by kristian1109 2y ago
Yes, this is an excellent prompt and we're working on it. One problem is a lot of these visual sources require permission, integration, and regulation. That's going to move slower than something we can proceed directly with (VHF a
17.
by kristian1109 2y ago
The problem with datalink systems is they are poor substitutes for immediate control & confirmation. My co-founder Eric wrote a short piece about this: https://www.ericbutton.co/p/speech . This is why they are mainl
18.
by kristian1109 2y ago
The systems I've flown have definitely been altitude aware. They won't alert if the plane is sufficiently deconflicted above or below. One problem for sure is that when you're close to the ground you have to be careful about
19.
by kristian1109 2y ago
Yes, fair points. In talking to controllers, this has already come up. There are a few systems that do advisory alerting and controllers have expressed some frustration because each alert triggers a bunch of paperwork and they are not 100%
20.
by kristian1109 2y ago
1. Our first product is post-op review at airports. We're selling that to airport managers who use our system for training and incident review. Today, when a ground ops vehicle (for example) makes a mistake, the airport manager has to
21.
by kristian1109 2y ago
Yes, this is a challenging issue in ATC. Also, some safety systems are turned off or diminished at critical phases of flight (near the ground) because of the noise problem (TCAS, for example). ADS-B helps with this as it's self-reporti
22.
by kristian1109 2y ago
Thanks! Appreciate your inputs here.
23.
by kristian1109 2y ago
It's been interesting to see that a product as simple as combining data from multiple frequencies at once has been really compelling to folks. Can't tell you the number of times we've heard "wait, can you compile ground,
24.
by kristian1109 2y ago
I have a bit of an ax to grind with the state of pilot training tools. There's some interesting new work being done here, but agreed lots to do in this realm. What does a compelling "automatic CFI" look like to you?
25.
by kristian1109 2y ago
Who gave you our event schema!? ;) Kidding aside, yes, you're exactly right. We're already doing this to a large degree and getting better. Lots of our own data labeling and model training to make this good.
26.
by kristian1109 2y ago
Language is critical input data. Appreciate it!
27.
by kristian1109 2y ago
AGI is a moving target, but agreed, lot's more research to be done.
28.
by kristian1109 2y ago
Appreciate you!
29.
by kristian1109 2y ago
Good and fair questions. In the very short term, we're deploying this tech more in a post-operation/training role. Imagine being a student pilot, getting in from your solo cross country, and pulling up the debrief will all your co
30.
by kristian1109 2y ago
Post product market fit :D

More ›