alt Hacker NewsIn 2026, with how much money their is in aviation, it seems wild to not have digitized this ages ago. The runway should be essentially 'locked' when in use, if they don't want screens in every ground vehicle that may cross a runway, at least display it at runway entrances.
That ATC still takes place over radio just seems insane at this point. And there's pretty much no way to make ATC's job not stressful, its inherently stressful. Taking out how much of their job is held in the current operators mind versus being 'committed' seems like low hanging fruit 30 years ago.
The whole system's just begging for human error to occur. There's 1700+ runway incursions a year in the US alone, each one should be investigated as if an accident occurred and fixes proposed. Like when an accident occurs.
Replies
"each one should be investigated as if an accident occurred and fixes proposed"
I feel the same way about close calls on the road, especially ones involving a vehicle and a vulnerable road user like a pedestrian or cyclist. Way too many lives being saved by a person jumping out of the way at the last minute who shouldn't have had to do so, and then cops and bureaucrats shrugging with "well what do you want us to do, the numbers don't show enough fatalities here for it to be worth fixing" and later when someone actually does die it becomes "this is a horrible tragedy that no one could have seen coming, let's focus on thoughts and prayers rather than accountability that could lead to structural change."
Air traffic (and ground traffic) control are not simple problems. La Guardia has 350k aircraft operations (takeoffs and landings) every year. 1000/day. Peak traffic is almost certainly more than 1 plane every minute. Runways are always in use and the idea that some simple software will solve all the safety problems is not grounded in reality.
You can't just throw software at this. It's a complex system that involves way more than just an airplane and someone in a tower. Systems engineering, human factors, and safety management systems are the relevant disciplines if you'd like to start reading up. In addition there are decades of research on the dynamics between human operators and automation, and the answer is never as simple as "just add more automation." Increased reliance on automation can paradoxically decrease safety.
CPDLC is already being deployed domestically. It's currently available to all operators in en route segments.
All runway incursions at towered airports are reported, classified according to risk, and investigated.
> The runway should be essentially 'locked' when in use, if they don't want screens in every ground vehicle that may cross a runway, at least display it at runway entrances.
It does, the Runway Status Lights System uses radar to identify when the runway is in use and shows a solid bright red bar at every entrance to the runway. I'm curious what the NTSB has to say about it for this incident. From the charts LGA does have RWSLs. I didn't check NOTAM to see if they were out of service though.
How would you exactly "digitize"? While that sounds like a nice idea in theory it's the same as "digitizing" road traffic.
In the end the air traffic system is a highly complex but also a highly reliable system, especially when you compare accident rates.
I am sure the working conditions of ATC staff might be improved - but being both a pilot and a programmer, I know that there is no easy digitalization magic wand for aviation.
Automating ATC is similar to automating flying in general. Even if it's possible to automate 99% of 99% of flights, including even takeoff and landing, commercial flights still have two pilots because if things start to go wrong there's just so many edge cases that you can't easily write automation to handle all of them. Same thing for ATC, except even worse. They still have control towers because controller eyeballs still work even if nothing else does, if ground radar fails, or if a vehicle doesn't have an ADS-B transponder, or if a crash eliminates the radios, etc. There's just so many edge cases that making automation be able to handle everything is extremely difficult
Ha. My first job in '89 was working for an FFRDC reviewing IBM's Jovial code that was going to "revolutionize ATC" by modernizing everything.
I'm gonna guess that code never went into production. The problem seems easy until you start looking under the hood.
There are systems for it, just not really integrated into emergencies and ground vehicles. Mistakes also happen even if all info required to avoid is present
The BBB allocated $12B for ATC modernization. https://www.faa.gov/new-atcs
Money isn't the only reason it's so old. The coordination problems are huge. https://www.theregister.com/2024/09/24/us_air_traffic_contro...
> That ATC still takes place over radio just seems insane at this point.
There is digital comms with ATC without voice:
* https://en.wikipedia.org/wiki/Controller–pilot_data_link_com...
* https://www.faa.gov/air_traffic/technology/DataComm
But in the highly dynamic environment of final approach, landing, and taxiing, I doubt it would be practical. Unless we want to try autonomous 'driving' on taxiways and runways?
> There's 1700+ runway incursions a year in the US alone, each one should be investigated as if an accident occurred and fixes proposed. Like when an accident occurs.
How many runways crossings are there in a year? How much is "1700+" a percentage of that total?
You seem to be giving too much credit to the singleton design pattern. We know exactly how well that works on a modern, multi-tasking, preemptible operating system (hint: not well at all).
> That ATC still takes place over radio just seems insane at this point.
Voice communication is insane? I suspect you are ignorant of what it is like to actually fly a large aircraft into a busy airport. Fault-tolerant and highly available hardware must facilitate low-latency, single-threaded communication with high semantic density in order to achieve multi-dimensional consensus in a safety-critical, heterogeneous, adversarial environment.
There is some interesting research that captures this sentiment and shows how complex a solution might need to be (replace "faulty agent" with "human error"): https://www.sciencedirect.com/science/article/abs/pii/S00051...
I would not trust my life to a government software project (See Phoenix Payroll for a typical case)
While modernizing ATC in the US may be overdue, the real issue here is that ATC in the US has been understaffed, underpaid, and overworked for a while now.
My father works ATC and his schedule has him working overtime, 6 shifts a week, including overnight shifts, meaning that there is literally not a day of the week where he doesn't spend at least some time in the tower.
If that's the reality for even half of the controllers, it's no surprise that we've been seeing more and more traffic accidents lately.