Tuesday, August 18, 2009

Thoughts on the Mid-Air Over the Hudson

A number of people have asked me for my reaction to the recent mid-air collision over the Hudson River in New York City. My inclination is to refrain from expressing any opinions while the NTSB conducts its investigation, but I do have a couple of comments that needn’t wait.
It is useful to read a 1991 report published by the Australian Transport Safety Bureau (ATSB) titled Limitations of the See and Avoid Principle. The report was prepared after a mid-air collision in Queensland involving a Cessna 172 and a Piper Tomahawk. It occurred in the traffic pattern in good visual meteorological conditions.
The report details many reasons – physiological, psychological and technological – for finding that unalerted see-and-avoid cannot be considered as a dependable anti-collision strategy. The recent collision in New York simply reinforces this point.
It was of great interest to me to read the ATSB’s findings on the difference between alerted and unalerted see-and-avoid. The report states:
“A traffic search in the absence of traffic information is less likely to be successful than a search where traffic information has been provided because knowing where to look greatly increases the chance of sighting the traffic. Field trials conducted by John Andrews found that in the absence of a traffic alert, the probability of a pilot sighting a threat aircraft is generally low until a short time before impact. Traffic alerts were found to increase search effectiveness by a factor of eight. (emphasis added) A traffic alert from ATS (note: Aussie for “ATC”) or from a radio listening watch is likely to be similarly effective.”
I can certainly attest to the increased likelihood of spotting traffic when it is “cued” by the Traffic Information System (TIS) feature ennabled by N631S’s Garmin 330 Mode-S transponder (and displayed on the GNS530W).

The picture shows TIS displayed on a GNS430.
The bearing and altitude information that this system provides is very helpful in spotting traffic and in my experience, is more useful than the controller’s verbal point-out (“Traffic at one o’clock, four miles, a Baron westbound at 5,000.”)
If we start with the premise that putting all flights under positive control at all times is both unachievable and undesirable (and I do think this is so), then we will need to look very hard at “see-and-avoid” to determine whether the strategy’s effectiveness can be meaningfully improved.
It seems hard to avoid the conclusion reached by the ATSB's Bureau of Air Safety Investigation (BASI):

"The see-and-avoid principle in the absence of traffic alerts is subject to serious limitations. It is likely that the historically small number of mid-air collisions has been in a large part due to low traffic density and chance as much as the successful operation of see-and-avoid.

Unalerted see-and-avoid has a limited place as a last resort means of traffic separation at low closing speeds but is not sufficiently reliable to warrant a greater role in the air traffic system."

The Hudson River VFR Corridor is intended to provide an alerted see-and-avoid environment. The means of alerting is a combination of rules-of-the-road ("Keep Right!") and self-announced position reports on the Common Traffic Advisory Frequency (CTAF) of 123.05 MHz. The Piper Saratoga that was involved in the mid-air was, regrettably, not communicating on the CTAF. Its pilot was talking to the Teterboro (KTEB) tower and may have just been switching (as directed and acknowledged) to the Newark (KEWR) tower frequency. It is unknowable whether the collision might have been averted if the Piper pilot had been on the CTAF and able to hear the helicopter's position reports.
Two things are disturbing here. Would the Piper pilot, who was a transient visiting from outside the New York City area, have been able to effectively process the position information from the reports of potentially conflicting aircraft even if he had been on CTAF? This is not in any way a disparagement of that pilot's skill, expertise or competence. But how many of us, hearing "Helicopter 405 climbing mid-river to 1100 abeam the Holland Tunnel Towers" will know just where to look...at a time when seconds count? And second, how useful are those sorts of position reports, in the last analysis, for raising the likelihood of timely target acquisition to an acceptable level?
I believe there are a couple of things that can be helpful. One is fairly easy and one is not so easy. First, consider as a model the on-line course that is required for pilots flying VFR within 60 miles of the Washington ADIZ. A similar course could be designed, and successful completion required (yes, I said required) for pilots wishing to transit the Hudson VFR Corridor. Such a course would cover altitudes, landmarks and proper position reporting.
Second (and harder, because now you'd be asking folks to spend money), a technological aid like the TIS system shown above, could be required for commercial operators and strongly urged for non-commercial users of the Hudson airspace.
These two steps (or others of similar effect and intent) would go far to making the congested Hudson VFR Corridor an effective alerted see-and-avoid environment.
Since I got my Private Pilot's License, I have flown the Hudson Corridor a total of three times. Each time it made me uncomfortable. Too many aircraft, not enough sky. My own workaround is to talk to New York Approach and request clearance into the Class B Airspace and the river tour at 1,500 feet. They are normally very accommodating (but might become less so, if everybody was asking). I hope that when the commotion resulting from the recent accident settles down, some well considered changes will be made that avoid over-reaction and result in a safer VFR environment over the river.


nec Timide said...

Very important points. You are fortunate (or smart, or both) to have that technology available.

One of the things I'm involved with (mostly on the periphery) is pilot traffic perception position processing done by the cognitive research department at a local university. Some of the preliminary results are interesting and match what you have reported.

Frank Van Haste said...


Re: lucky v. smart, I hope some of each. :-)

Thanks for commenting...when the traffic perception work has something ready for publication, I hope you'll come back with a link.



DitchCollege.com said...

I have been doing some research into the air traffic control profession for a website I am creating and this was a fascinating read.

Pretty intense stuff.