Quote:
I think Internet SAR is a great idea, but I have some suggestions to improve it.
Thanks BahiaMar and welcome to the GEC forums.
Quote:
(1) The GE forum thread is far too long and filled with unsorted posts with non-descriptive headings. No one has time and patience to read it all. It would be better to run a moderated or edited forum on the InternetSAR website with carefully constructed topics.
I agree that a single thread per search mission becomes unwieldy. Like I mentioned a couple posts up, I almost think we would be better served to have a separate forum for each search mission with people being able to start multiple threads on distinct topics. It is a factor that is being considered among many factors as we plan our long range forum strategy.
Quote:
Any valid information from the forum should be compiled into the briefing.
We will be keeping an official mission briefing on the InternetSAR.org website for each search mission. We currently have a mission briefing for the Boychuk search (note log in required), but not the Fossett search.
Quote:
(2) It is necessary for searchers to have ready access to reliable information so that they can be effective. Good information informs and guides a search, and it maintains the morale of the searchers. Some of the material is vague and conveys the impression that there are two levels of access: the 'real' search team, and the members of InternetSAR. I understand that is probably not the reality, but it is the impression conveyed.
Again, as I mentioned a couple of posts up, I am programming a system to manage information at this very moment. Accessing mission briefings do require individuals to have an account on InternetSAR.org. By the nature that some information is confidential in nature, not all information will be released (e.g. names of witnesses, contact information, etc.), but our goal is to provide a central clearinghouse for information and we want to provide as much information as we can. If we withhold information it will either be because we can not verify it or it is sensitive or confidential in nature.
Quote:
(3) Good information is mixed carelessly with bad. By bad, I mean any media report and all speculation. The media is horrendously, laughably poor in reporting any technical matter and must be almost completely disregarded. A separate thread should be created for any information derived from the media and labeled with a disclaimer like "CAUTION: Media Reports. For Entertainment Purposes Only." Information derived from media reports must be excluded from the briefing.
This only happens on these forums. The mission briefings on our website (like our Boychuk mission briefing) will be compiled by an information analysis team, which will be lead by our lead senior evaluator who is a professional imagery analyst and our Canadian SAR Liaison who has 20+ years of Canadian SAR experience and is search master trained.
Quote:
(4) The mission briefing should be written by someone who is familiar with aviation accidents (sorry, Ken).
No offense taken, you are correct a SAR professional needs to be responsible for the mission briefing and for leading our information analysis team. Again this is what is happening. You will not find me having much input in these mission briefings because this is not my specialty. Our Boychuk mission briefing was written by our Canadian SAR liaison, Greg Ursel, who again is search master qualified. This mission briefing is a living document and will be updated as more information is confirmed.
Quote:
This is where the most effort should be expended. Some of the material that is missing from the briefing is:
- What time did the pilot depart Calgary/Springbank?
- Were there any stops en route prior to Revelstoke?
- What was the arrival time in Revelstoke?
- What was the source of the Revelstoke departure time? How precise is it?
- What was the takeoff weight the aircraft?
- How much fuel did the pilot take on at Revelstoke?
- What were the aviation weather conditions and forecast that would have been available to the pilot at the time of departure?
- Did the pilot receive a weather briefing?
- With whom, and at what precise time was the flight plan filed?
- Were there any changes to the flight plan?
- Was the pilot instrument-rated?
- What was the actual aviation weather at reported at each of the local airports for the expected duration of the flight? (Vancouver weather and non-aviation weather reports are almost irrelevant.)
- Did the pilot make any position reports en route?
- If so, how and to whom and at what time precisely? (One of the forum posts says that he reported over Spence's Bridge. If that is true, the the time of that report is the most important piece of information we could have. If the report is spurious, the searchers must be emphatically told to disregard it.)
- What were the positions and times of the radar plot? We must have raw data, and all of it that is available. One poster said that the radar plot was made by a U.S. border patrol aircraft. If so, they will have a wealth of data. There can be no conceivable reason that it would be unavailable to us.
- What are the procedures for avoiding an incursion into the Vancouver TCA that would have been available to the pilot, given his skill, qualifications, terrain and the weather?
- What was the pilot's personal schedule for the days after his flight? Was there time pressure to arrive?
This is a good check list of questions, which Greg Ursel (gursel2) our Canadian SAR Liaison will surly work on answering. He has been doing a great deal of work interviewing witnesses and collecting data. The only information we make available on our mission briefing is that information that has been confirmed to his satisfaction. Thus he may already have answers to these questions, but has not made the information public pending confirmation.
Quote:
For an example of the quality and quantity of information required, check out this amateur site. There is an excellent radar plot with times, positions and altitudes.
The new mission information system I'm programming will be able to track all of this type of information and automatically plot it to Google Earth KML overlays. It may not have all of the features at first, simply so I can get something up and running, but in time it will be fully flushed out with all the features needed. Please understand that there is a lot of work involved with programming a site of this nature and it will be a work in progress for many months to come.
Quote:
Most importantly, the source of each piece of information must be clearly stated. If eyewitness information is given, the aviation and technical background of the witness must be given. An airport worker who thinks an engine sounds odd, for example, may have almost no experience (having left Starbucks last week for a better job at the airport), or may be a private pilot working on an ATP rating. It makes an enormous difference.
The new system will cover this concern and even has a rating system for information credibility.
Quote:
(5) The Images. We need to know the technical details of the images. When were they taken? What was the angle? What was the sun position? What was the ground cover? (Is it snow? It looks like snow, but who knows?) Is the border overlay correct? (If the border overlay is correct, then the images available are almost certainly not where the aircraft is located, based upon the available information. Since the information itself is poorly categorized and incomplete, there may be a good reason for searching that area.)
Noted it will be things I will work on having the ability to record into the databases as we upload imagery. I have an extensive punch list of programming tasks so I can't promise where in my programming queue it will fall, but I will get to it.
Quote:
In either case, the imagery does not cover the area that should be searched.
As the mission briefing states and has been stated here many times, we have areas of higher interest to search, but we can not get imagery for those areas until the snow melts. The imagery we have now was captured at some point in late November and was then brought to us for analysis in December. We had no input into the acquisition of the imagery, but I know it was based on the efforts of the Canadian SAR team and what the Boychuk family thought was of particular interest based on what was known at that time. Since it is the only imagery we had and we can not get new imagery until spring, it made sense for many different reasons to process the imagery to the best of our abilities. If for no other reason, it will bring peace of mind to the Boychuk family that the imagery was carefully reviewed and ruled out. It has also allowed us to greatly refine our systems.
Quote:
With better information, I could make a much better and more accurate wild-assed guess, and search the area predicted. So could everyone else, with the likely outcome being eventual success.
Absolutely, but part of improving the system is a matter of trial and error as well as figuring out what systems need to be built first. We really are in uncharted territory as far as what we are trying to do here and things won't go as smoothly as we would like in the beginning. In time we will have procedures and systems ironed out and we will have built a really effective tool. It is just going to take a lot of work.
A top priority for us is to write procedural and training manuals for what we are doing. Having standard operating practices (SOPs) and systematic training will greatly improve our effectiveness.
I would like to encourage anyone who has experience collecting and/or analyzing SAR information to PM Superquint or Gursel2 with your contact information (e.g. email address) and SAR background. Supersquint is working on putting together an information analysis team to help sift through everything posted here and collected via physical investigations (e.g. interviews).
--------------------
InternetSAR.org: Volunteers collaboratively analyzing aerial and satellite imagery to assist in search and rescue efforts.
DISCUSS ACTIVE SEARCHES: Ron Boychuk in British Columbia, Canada & Steve Fossett in Nevada, USA
My blog posts on my MTurk search experience:
Using the Internet to Revolutionize Search and Rescue
Internet search for Steve Fossett eight weeks later
|
Previous
Index
Next
Flat
Re: Humans Make Poor Robots
Edit
Reply
Quote
