CAP Aircraft Searching for Steve Fossett

[mawr]

Members having to pay for mission costs is absolutely not right!  We need national to issue credit cards for missions, and let them sort out the repayment in the end.

I've seen it done on a Wing level (for aircraft maintenance) and it was a disaster.  Pilots just could not understand when to use the credit card and when not.

[SoCalCAPOfficer]

I've seen it done on a Wing level (for aircraft maintenance) and it was a disaster.  Pilots just could not understand when to use the credit card and when not.

Lets see, us pilots are smart enough to aviate, navigate and communicate, but we cant figure out when to use a credit card?   I think we can be taught.

[RiverAux]

My wing has just the opposite experience.  A card in every airplane and it works just fine. 

[KyCAP]

Two things...

TFR's can be requested by any CAP IC through AFRCC.  I have made such a request last year on a mission.  Now with AFRCC in Tyndall there is an FAA Liasion in the OPS Center.  

Second, KY Wing has managed maintenance at the Wing level for well over a decade.  In our USAF compliance inspection I am pretty sure that we received an Outstanding in Operations and Aircraft Maintenance.  

Also, all of the planes have Multi-serve fuel cards and the pilots pay NO mission costs out of pocket.   In fact all fuel is purchased through multi-serve and the pilot reimburses the wing post flighteven for proficiency flights.  

[KyCAP]

Has anyone actually thought about this Mechanical Turk thing with Amazon / Google Earth as being "useful"?   Could there be some benefit from this??  Corporations like Satellite companies could write software interfaces for CAP and "donate" the images viewed and write this off their corporate taxes as "donations".   Then from a secure interface "trained" scanners could process this through an actual planning section to weed out the "leads"? 

Something like, Incident Commanders could define the containment area in WMIRS and the interfaces would get the data from their servers through XML/SOAP web services and populate back into Google Earth... (anyone at Google listening? hint)

Surely there are trained image analysts in the "know" lurking in the dark on this subject...

Personally, I have looked at the images and the resolution that I am seeing seems to be so low that I would really not see how this could be helpful for a crash site.   I suppose you could pick out a plane on a ramp in perfect shape, but anything else in any type of cover would be indistinguishable.  IMHO..  IC,  and MP

[bosshawk]

I happen to disagree with the opinion that the imagery on Google is not useful.  I am a trained imagery analyst of over 40 years and I looked at a lot of the imagery of the Fossett search area.  If you are trained to interpret imagery, the Google stuff is adequate for searching for a downed aircraft.  True, it is about 1 meter resolution and larger scale would be useful, but it is useable.  One of the nice things about digital imagery is the ability to rotate the image and tilt it so that you are viewing a more or less natural look.  You can also zoom in and out and that is sometimes useful.

All that said, crashes in the mountains usually involve going into the trees and no amount of high resolution is going to help much: the trees have a habit of shredding the aircraft and small pieces can hide under the trees and brush really well.  I could spend the next two weeks explaining how to search on imagery, but it would get boring really rapidly.  Of course, if the correct area is not imaged, it is all moot.

[KyCAP]

Good to know... Like I said, I was on a "fishing expedition" in my opinion as an "untrained" person not as a matter of fact....    I'm a software engineering by training.... If you see the imagery in there as useful, then what's your two cents on harnessing this type of tool for "real mission use" in an environement like I described?   

Think about it, your skills and otheres like you could be used to plug into the mission long-range.   Rather than hordes of untrained Googlers(?) real useful lead information could be headed into the mission base.

[a2capt]

I must have looked at over 2000 of those images, with holding an MP rating I have some better idea of what to look for vs. John Q. Public, and the very first thing I know is we are usually *NOT* looking for an airplane. What I was looking for was stuff out of the ordinary because I'm not going to see an airplane as plain as that sample image they pose next to the actual image you are looking at. I also feel I could have looked at images twice as large in about the same amount of time, I think the samplings should been a larger area.

[sardak]

While the imagery might be useful, the Mechanical Turk process leaves a lot to be desired.  The instructions read: Example of the size of object to look for. The white plane shown above (30 pixel wingspan by 21 pixels by length) is approximately the size of Steve's plane.  The image shows an intact aircraft sitting on a ramp.  It's highly unlikely the subject aircraft looks anything like that, structurally or size-wise, as trained searchers should know.

Many on-line "searchers" have figured that out.  To get an idea what people are reporting, there is a thread in the Google Earth "Community Forums" in which people are posting questions and coordinates and pictures of possible targets.  The thread is currently 103 pages long.  The link is:
Google Earth Fossett Search Thread

Another problem is the screening process.  Filling the IC's email inbox with pictures of possible targets (which has happened) is not useful. 

Assuming a reasonable initial screening process and delivery method can be worked out,  the next problem is sorting out these "clues."  Is there time and expertise to examine the pictures?   While there are some experienced imagery analysts, are there enough?  Are they available, etc?  How does one decide  which pictures to follow-up on?  This method produces far more clues than come in on a typical search.  What if one of the submitted pictures doesn't make the cut but turns out to be the right one?   

Aerial imagery is obviously a useful tool, but for use on a regular basis there needs to be a better system than the Google/Amazon one being currently being used.

FWIW, I spent today analyzing aerial imagery of forested mountains for work.  The photos range from 0.25 meter/pixel to 1 meter/pixel resolution, taken at different times of the year, at different times of day.  I was looking for objects on the ground one to four feet across, that I knew were in the images because the ground verification has already been done.  The viewing program can pan, zoom, and rotate in 3D.  The POD was very low.   As I looked at them, I thought about what it would be like to search them for airplane parts or evidence of a crash. 

Mike

[bosshawk]

I certainly concur with some of the foregoing comments.  We certainly are not looking for a whole airplane: mostly bits and pieces and fragments.  If you ever seen a crashed airplane, it resembles more nearly a pile of junk than a piece of operating machinery.

My whole point was that imagery MIGHT be useful, in the right hands.  From all of the posts, it looks like there are three of us who have any imagery analysis background.  That is probably typical of CAP: I doubt that there are more than a dozen people nationwide who have any imagery analysis background.

That says that using the stuff for a routine search is not a viable process.  It also takes a considerable amount of time to wade through the imagery of a search area.  That time could likely be better spent putting aircrews in the air.

You also have the issue that commercial imagery is not up to date: sometimes, the Google stuff is a year old and that isn't much help in a crash that happened last week or yesterday.  I understand that Google did manage to get some new coverage of the Fossett area, but how much I don't know. It would also matter where they imaged: you can't just willy nilly fly imagery strips and expect to get much in the way of results.  The classified systems are a different story, but we can only wish!!!!!!!!!!!!!!

The amateurs might come up with something, but the false positives will drive the mission managers crazy and they will tend to ignore that after awhile.

[RiverAux]

Unless the cost gets dropped significantly lower than the $100K reported in the media for the mechanical turk project it is not going to be feasible to do fthis for the regular folks we spend most of our time trying to find. 

Even without this high tech stuff CAP (and other agencies) have a pretty high success rate at finding missing airplanes.  In 2005 only 4/95 missing airplane targets searched for by AFRCC were not found. 

[eaglefly]

A question for the Ground Pounders involved in this search.  Are any of your teams saying out over night?  Any teams camping out in the mountains?  If so how are they doing?

[A.Member]

Unless the cost gets dropped significantly lower than the $100K reported in the media for the mechanical turk project it is not going to be feasible to do fthis for the regular folks we spend most of our time trying to find. 

Are you kidding?  We didn't blink an eye with ARCHER (an equally unfeasible product, IMO).

But I agree, the cost/benefit is not there right now.  Maybe...someday.

[PHall]

A question for the Ground Pounders involved in this search.  Are any of your teams saying out over night?  Any teams camping out in the mountains?  If so how are they doing?

The Ground Team, yes that's team, has not been out overnight. The mission they have been doing is interviewing people in the area of interest and going out to the old wrecks that have been found to mark and ID them.

The ground team game out here in California and Nevada is very, very different that what you guys have back east.

The closest that we come to teams being out overnight is that if a team will be in the same area the next day, and it's more then a 2 hour drive back to the search base, the may camp out overnight in the field. But it doesn't happen very often.

[wingnut]

I would like us to create an Intel unit that is well versed in Imagery, Geographical information systems, display of information, NTAP use and display, the correct mapping of all known wrecks so it can be displayed (we wasted a lot of time and money on that in the Fossett search)

Wait what am I saying? wait for what, I think I will get with Bosshawk and just do it. We teach High School kids how to look at Imagery and Hyperspectral change detection. Why not CAWG, and lets get a plotter donated, and a DLP projector or two for mission briefings.

I flew about 40 hours on that mission and I really admired the ICs for what they coordinated out of chaos and us primadonna aircrews.

[dougsnow]

I was reading somewhere, that CAP liaison personnel staffing the AFRCC were required to have a SECRET security clearance for working at the AFRCC. I was reading a Memorandum of Agreement about some Ops Control Center at Tyndall.

Overhead imagery is taken by the "national technical means", the CAP liaison personnel at AFRCC would be cleared for the take; take the product from the squints (thats what we called the imagery interpreters when I was a USAR 96B), and relay the interpretation reports to the IC, so he could effectively employ his assets in the areas that looked promising. OR, as part of the IC qualification, vet for security clearance and allow him to see the take from the national technical means, and judge for himself his employment of SAR assets.

Granted, it would be nice for the overhead take to be given straight to the mission pilot, planners, and ICs, but I can imagine that would be cost prohibitive just in security clearances - but hey, if it gets the job done, I'm all for it...

[eaglefly]

"The ground team game out here in California and Nevada is very, very different that what you guys have back east."

We don't do it either, my favorite 'pup tent' has Holiday Inn written on the door.   That is why I was asking, I figured that if CAP GT's were not staying out on this one they never would.   

Thanks

[NIN]

I was reading somewhere, that CAP liaison personnel staffing the AFRCC were required to have a SECRET security clearance for working at the AFRCC. I was reading a Memorandum of Agreement about some Ops Control Center at Tyndall.

Overhead imagery is taken by the "national technical means", the CAP liaison personnel at AFRCC would be cleared for the take; take the product from the squints (thats what we called the imagery interpreters when I was a USAR 96B), and relay the interpretation reports to the IC, so he could effectively employ his assets in the areas that looked promising. OR, as part of the IC qualification, vet for security clearance and allow him to see the take from the national technical means, and judge for himself his employment of SAR assets.

Granted, it would be nice for the overhead take to be given straight to the mission pilot, planners, and ICs, but I can imagine that would be cost prohibitive just in security clearances - but hey, if it gets the job done, I'm all for it...

FalconView

[dougsnow]

Actually, I found my reference again here...

http://level2.cap.gov/documents/CAP1AFMOUpdf.pdf

But yes, FalconView would be very cool

[ltcmark]

I just saw an interview with Maj Ryan on the Associated Press.

http://cosmos.bcst.yahoo.com/ver/242/popup/index.php?cl=4133989

Did she have her badges and name plate on backwards???