July 03, 2009

An Operational Critique of NextGen




Let’s review Robert Poole and Kathryn Wylde’s Cause. Perhaps later we’ll look at who profits from their rhetoric.

They champion a new air navigation system called NextGen. Using these technologies, pilots will be free to navigate dynamically. If you buy the hype, freed of a route structure that looks like the 1950's interstate highway system, pilots will be able to reroute themselves, and they'll separate themselves from other aircraft.



The problem with our archaic World War Two system, Poole pontificates, is that it makes airplanes line up for runways, airports, cities, and routes. If we’d use GPS (and other variations on the theme) to let pilots navigate direct, we’d save millions, and avoid the queuing that causes such onerous delays.

I generally write in a self-deprecating tone, but I know that in order to persuade I should present credentials and attempt to establish credibility (ethos). I’ve been a pilot and a controller. I understand and have used these technologies.

GPS and NextGen do present remarkable navigation capabilities. They allow individual planes to navigate direct, to develop creative routes around weather, and they would free airplanes from having to participate in a rigid system.

There is very little new under the sun. Philosophically, the problem is Immanuel Kant's Kant’s categorical imperative. In response to the question, Is it OK for me to do (yada yada)?, Kant would ask, Would it be OK for everybody to do it?

If there’s only one driver in town, can he drive wherever he wants? Sure. Drive one-way streets the wrong way, ignore stop signs and lights? Yes.


And in fact, if there’s only ten cars in the city, you can still mostly get away with that sort of chaos, although – eventually – someone’s going to get killed, and then you’ve probably only got eight cars.

It’s tremendous for one person. It's amazing in a limited situation. Kant would ask, Is it okay if everybody does it? And the answer is: No. It would be a disaster with 500 drivers and cars.


             Überlingen Midair        Re-enactment Video

When you scale the city up to a thousand drivers, all of a sudden we don’t permit that sort of libertarian chaos. We have one-ways, two-ways, turn lanes, flyovers, HOV lanes, limited access highways, right on red – we get pretty gritty about the details. Overall it’s a mostly efficient, mostly safe system that permits unfettered access without pre-arrangement – you can drive to any address you want, whenever you want – and does it without direct user fees or reservations. Society manages the cost of the system, paid for by the common wealth through a variety of public taxes.

Ahhhh. I'm glad that's settled.

So I’m wrapping up this car analogy, I’m leaning back smugly, and Robert Poole might say: but wait, this is the informal fallacy of false analogycars and roads are only two-dimensional, and airplanes fly in a three-dimensional construct. There’s no reason to force them to go even-streets north, odd-streets-south; put them at different altitudes. Let them fly where they want! Embrace the flexibility inherent in the system!

And Robert Poole would be right. Aviation is three-D; in fact, if you’re good at it, it’s four-D, but that’s another topic. But there’s two times during each flight when airplanes are just like cars, when they’re two-dimensional just like cars, and that’s when they’re using runways (roads) to takeoff and land.

A funny thing about runways: they’re long, and they’re straight, and they’re not very adjustable. Planes have to “line up” on runways. You can’t move a runway (although aircraft carrier runways do move around pretty well). When planes line up with runways to land and to take off, they tend to fly over the same houses all the time, and NIMBY neighbors don’t like that.

You can run all the 3-D matrix airspace configurations you want, but the bottom line is that airplanes have to line up for a runway. With airliners, generally speaking we only want one airplane using a runway at a time – so each runway has a maximum capacity of x airplanes per hour. When the weather goes from CAVU to cloudy to nasty, a good runway's occupancy rate goes from 40 planes/hour, to 25 planes/hour, to 20 planes/hour.



Runways are the choke points that cause air traffic congestion in New York, Chicago, and LosAngeles. Planes have to line up on runways, they’re a fixed property with a limited capacity (that decreases in bad weather). You want a new technology that reduces delays? Build a better mousetrap – figure out a way to do without runways.



Runways and runway capacity are the keys to avoiding air traffic delays. NextGen does nothing to address runway acceptance rates.

Some Wonderful Things About NextGen

I don't want to give the impression that I'm anti-NextGen, that I'm a Luddite who misses the days of four-course radio ranges - which, by the way, was a navigation system based on Venn diagrams. So let me mention some wonderful things about NextGen.


NextGen is going to open up new remarkable capabilities for use in limited situations. It's going to be wonderful for the oil platforms in the Gulf of Mexico, which is completely (and almost negligently) underserved by the present system.

NextGen is wonderful for "Midnight Rules". In the wee hours, when traffic is down to about 5% of peak times, you can run a lot of direct routes and approve things you'd never do on a dayshift. In fact, UPS and FedEx use a lot of cool techniques that are available right now, today on the midnight shifts. The problem is: you could never use these schemes on the day shift.

Here's a slightly dramatic example of NextGen saving lives:
EastBumfuStan is located 150 miles from the Metropolis mega-trauma center. It's snowing heavy. Out on the interstate, one car spins into another, there's a chain reaction in the low visibility, there's a twenty-car pileup. There's three severely injured people. Local volunteer firefighters have made it to the scene. Ambulences are on their way from 40 miles away; they're travelling in heavy snow, and the local hospital is not up to treating those people and saving their lives.

Back in 2009, those people would die. But it's 2015. Over at Stat MedEvac, a dispatcher double-clicks on the area map, and sees the local high school football field, the tall AM radio antenna a mile away, and the power lines running east-west. He designs an instrument approach and departure procedure so a helicopter can land at the high school, even in these snowy conditions. He thinks he's done, so he submits it to a validation program that checks his work. The software points out that he failed to consider a hill in the local terrain. He adjusts the approach course, the validator confirms the procedure, and he datalinks the details to the helo crew that's already in the air.

The crew lands, picks up two patients, and takes off for Metropolis. Another helicopter lands five minutes later and picks up the other critical patient. They all live. They'd have died in 2009.
NextGen can do that. For instrument flights at low altitude, in areas without ground-based navigation aids, NextGen is going to be remarkable.

Let me point out something about my two examples: they're both about helicopters. Helicopter aviation, specifically the Medevac and Gulf communities, is completely underserved in the current system. You should care about this, because when there's a terrible accident, it's a helicopter that means life or death.

The other side of the coin is: in both of these examples, the helicopters benefit because they don't need runways. Neither of these examples apply for fixed-wing (ie, normal) (sorry) aircraft that need runways.




NextGen would be a remarkable supplemental system. If you overlaid NextGen on top of today's system, there'd be an awful lot of small benefits realized.

Here's the primary issue: NextGen is so expensive, the industry and the salesmen can only cost-justify it by presenting it as a complete replacement for the current system. In other words, if you shut off all the radars and all the navigation beacons, and save all the money from maintaining them, only then might NextGen be justifiable. Trouble is: NextGen is not an acceptable stand-alone system.



In a June 14, 2009 story in the Dallas Morning News, aviation consultant Michael Boyd expressed his doubts.
Not only is the program management lacking, he said, but the technology of managing the air space may not be up to task. "I have no confidence this is going to work," Boyd said. "The public is simply being bamboozled... ...about how this is working."


There’s a few other details about NextGen.
  • NextGen offers a remarkable capability for navigation. That new capability doesn't scale well; you can let 5% of today's volume use it, and you can use it tactically when situations permit, but you can't use it as the sole basis for 5000 airplanes at a time.
  • NextGen cheerleaders say that more accurate navigation (that's true) will permit aircraft to fly closer together on final approach (that's false). Separation on final is driven by wake turbulence, not navigation ambiguity.
  • Way back in the days of Najeeb Halaby, there were conflicts between civilian and military expectations. The system is obligated to provide "primary radar" coverage. NextGen advocates want to shut that off in order to cost-justify NextGen.
  • There is no complete, published technical specification for NextGen systems. No defined deliverables. It's vaporware. It's a vendor's fantasy.
  • NextGen would not have tracked any of the 9-11 hijacked aircraft.
  • If an aircraft doesn't want to be tracked, NextGen can't track it.
  • If an aircraft has a total electrical failure (like say, AirFrance447, the Airbus off Brazil), NextGen can't track it.
  • NextGen depends on a military system. The military has the right to turn it off.
  • It depends on a satellite system that’s subject to solar interference.
  • We’re behind on maintaining the satellite array.
  • By the way, the Chinese have figured out how to disable satellites. It doesn't seem prudent to put all our eggs in that basket.


The vision of NextGen does some really impressive things. It would be a great supplemental system, but it's not sufficient as a stand-alone system. It does not provide sufficient stand-alone tracking capability. It can't support the volume of traffic at intermediate altitudes. It's tremendously expensive. Although the vendors want to sell it, it's not an acceptable "equal or better" replacement for the existing aviation system.


See Also:
ADS-B promises a lot
Why NextGen will not triple capacity


2 comments:

Anonymous said...

the kessler effect, basically space becomes un-useable because space junk keeps hitting the new spacecraft/ads-b and turnning it into more space junk, until it is impossible to have a space based anything, except space junkyards

Anonymous said...

I'm curious what your definition on "NextGen" is? Are you referring to ADS-B, or do you consider performance-based nav, data communication, and other communication and automation changes?

Post a Comment

Comments and Feedback? Love that stuff. Please leave your thoughts in the box below--