*** Accurate POD's & Efficient SAR Team Deployments *** To: Search & Rescue Info Date: Fri, 19 Apr 96 17:20:28 -0800 From: sar-l-d-request@islandnet.com Subject: sar-l-d Digest V96 #87 Date: Thu, 18 Apr 1996 12:59:45 +0900 From: SARINFO@mindlink.bc.ca (Search & Rescue Info) (by way of Mike Doyle ) To: sar-l@islandnet.com Subject: Re: cost of air and ground searches My sympathies to the comments about the cost of performing air searches. The world is a strange place. I have spent a fair amount of time collecting POD data, refining the theories of how we search and have developed the concept of search efficiency into a process called Search Priority. By analysing more efficient (ie less manpower to a given POD) we can now select more efficient search techniques that get the job done faster, ie with less time and manpower, than before. This is totally mathematically based and seems to have been accepted by my colleagues in the ground SAR community. (It is now included in ERI's 'Search is an Emergency' manual). With the exception of a few scientific types the air search people have not shown any interest in examining this work from the point of view of reducing the cost or air searches, which as you have pointed out, are far more expensive than volunteer ground searches. Is this because of tradition, the military environment, which I have had great difficulty penetrating, or what I do not know, but it seems odd that those who stand to gain the greatest economical benefit have been the least interested in examining the concept. Just an observation. Regards, Martin Colwell. ------------------------------ Date: 18 Apr 96 19:54:15 EDT From: Steve McConaughy <73764.3070@CompuServe.COM> To: "INTERNET:sar-l@islandnet.com" Subject: Re: cost of air and ground searches To Martin Colwell. Re Your comment : <<>> One of the troubling aspects of air POD to me has been defining what it is that we are determining a POD for? Probablility of detecting and airplane that went smashing into a hillside is quite different from observing clues left behind by a mobile lost subject on foot. We also quite often (almost always when air is concerned) mix resources in the same areas. Since you have been dealing with collecting these POD figures; Id be very interested in hearing how you would think that a search manager might balance 3 debrief reports for the same search area that might read as follows: Task 1 - SAR Aircraft trained to find downed planes reports 35% certainty (POD) that there are no planes on the ground in the area. Reported after about 15 minutes Search time in area. Task 2 - SAR ground team reports 35% certainty that there are no clues (footprints, broken branches etc.) left by missing subject in the area. 5 hours SAR effort. Task 3 - Air scent SAR dog team reports 35% certainty that subjects body scent is not in the area. 2 hours SAR effort. The mathematical Cumulative POD for this area would be: +/- 72.5% 1-((1-.35)x(1-.35)x(1-.35)) Is this a realistic measure? Or are we comparing apples, oranges and grapefruit? If these are equal it seems that based on this wild guess of search hours, an airplane is about 8 times as effective as dogs. (It used 1/8th the time and reported the same POD.) What were your conclusions off efficiency based on your Search Priority study? Does your study go as far as providing a guideline or framework for prioritizing resource utilization? Just starting to realize what I don't know!! Steve -------------------------------------------------------------------------- Steve McConaughy for Search Equipment Company Providing quality equipment for Search Teams everywhere P. O. Box 70626 Richmond Virginia, 23255 Phone: (804)967-9264 FAX: (804)967-9265 E-mail: 73764.3070@compuserve.com ------------------------------ Greetings Steve, Interesting questions you ask. But, what is the search area? ie. mountains, flats, night, day, experience of the teams, etc. Also, how large is the search area? Task 1 - What kind of search area can a team cover in fifteen minutes? Does this whole task need to be reasigned? Task 2 - Again, what kind of search team was this? It would be interesting to have a team give that kind of POD that includes broken branches and tracks. Is this team trained to find tracks? Task 3 - Who is your dog team? Why does it take them 2 hours to cover what a ground team did in Task 1 in 15 minutes and Task 2 took 5 hours? What kind of cover was/is in the search area? Sometimes I have a lot a questions and few answers. What I am getting at is that, what I have seen, there isn't anyway to give a good POD unless you can answer the above, and many other, questions. Is the LKP of a missing aircraft that much different from the LKP of a missing subject? Distance could and will be a factor. Missing subject - generally - up to several miles. Aircraft - many miles. Should't that be our most important pieces of information? Need answers to balance.. No search is the same. Searchers are not the same. An aircraft, at times, could get a better POD than a ground search team. But looking at Virginia, how many missing aircraft searches, that ground search teams were called out on, did the find happen because of an aircraft spotting the missing plane. Not many. Lots of the missing aircraft went straight down through the canopy. Look at ground teams. Even though they may be "certified" at a certain standard, that doesn't mean that they can preform at, or have kept up their skills, that stardard or above on a search. One search that I was a division commander on, one of the dog teams requested that the police meet them at a certain area. I passed that informatin onto Command and was questioned at length as to why this dog team needed the police, etc. My answer was " if "----" requested the police, then you had better get them". This dog and handler had been the field for many years, had many finds and shouldn't have been questioned by unknowing or untrained base Staff, not even considering the time lag. No trust! Didn't know their teams! Getting a realistic POD on a search not only comes from Book training and stats but from knowing your teams. This comes from experience... and I don't know if you can put that into stats. I think that this is what you are looking for but I wouldn't have a "clue" as to how to put all of the variables into a usable stat. As for the POD being different from aircraft to lost subjects, I feel that they shouldn't be. A POD is a POD. But, are the searchers (or observers in the search aircraft/s) trained in realistic POD's for the type of task, search, terrain, night/day conditions, weather, etc. that they are doing or are you leaving it up to Base Staff to try to decifer and redue their POD's without even being out in the search area? ie. is there no trust in your searchers, haven't the searchers been trained correctly, are they incompentent, or base folks falling into a trap? Is to much time in the training classes and field work being spent upon lessons that may be used once in a lifetime or is the time spent on the real search world. ie. learning realistic POD's, clue awareness, signcutting, tracking, orienteering, etc. It seems that a lot of us get caught up in management stuff and forget that if the search teams don't do their job, the mission can fail, or that the very least, take longer. Can't talk about aircraft, don't like to fly. Have talked to many flyers though. I'll keep coments to myself for lack of experience and not being a doer/flyer. Yep, I answered the bait. I just don't like to seen tunnel vision. Your questions are very real happenings, but, again, do we not need to address these issues at the basic level? I really believe that we, as search and rescue folks, should pay more attention to the basics which is, of course, the ground pounders. Had to put my 2c's in. Now that I have taken you question onto other questions, I feel better. Later, Greg ---------------------- Greg Fuller - sar1track@esva.net Director - Search and Rescue Tracking Institute Mantracking Course Developer and Instructor - Department of Emergency Services Member - Applalachian Search and Rescue Conference Member - Virginia Search and Rescue Council, etc. ------------------------------ Date: Fri, 19 Apr 1996 07:53:39 -0300 From: Phil Jennex To: sar-l@islandnet.com Subject: POD - AIRCRAFT Message-Id: Content-Type: text/plain; charset="us-ascii" Watching/reading the threads on POD I have come to realize that there is a misunderstanding on how an aircraft searches. TRAINED air sar people look for signs, signals and clues. We are not just trained to look for downed aircraft. When we are looking for that missing aircraft, we are looking for missing people. In a lot of cases people try to walk out. (Big mistake but..) We use a variety of search patterns, the selection of which depends on the type of subject, terrain, time of day etc. For localized searches such as missing persons we use a pattern called a "Sector Search" It is a bit hard to describe without a diagram but I will try. We center on the point of highest probability. This could be the LKP (we use NASAR search master methodology to determine this position). We fly a track the prescribed distance (varies according to subject) usually 3 miles, and turn right 120 degrees another equal distance and we turn right 120 degrees again. This time we go through the center point and on out the other side the same distance and do it again. We continue this until we have done what looks like three equilateral triangles centering on the mid point. If we have not spotted the target we offset 30 degrees and do it again. Based on 3 mile legs we will have covered over 50 miles of airspace within 3 miles of the search point. We will have crossed the midpoint 6 times from 6 different directions. We will have had two highly trained spotters looking at that small piece of land for at least an hour. These spotters are looking for signs and or signals. Aircraft are best used when the missing subject wants to be found. A sign or clue, for an aircraft, is anything that looks out of place. Signals are man made and meant to attract attention. Phil Jennex MCTS Training Dept. CCGC Sydney, NS ------------------------------ Date: Fri, 19 Apr 1996 08:15:30 -0300 From: Phil Jennex To: sar-l@islandnet.com Subject: Re: cost of air and ground searches At 12:59 PM 4/18/96 +0900, you wrote: >My sympathies to the comments about the cost of performing air searches. You may have missed my point. Because there is a cost to the tasking agency, air search teams are sometimes not used in cases where they are very appropriate. The sympathies are to the missing persons who are not getting the full scope of services available. Your points are well taken but... For us, POD is a search masters tool. It tells him weather or not to send a second crew to the same area, using a different search pattern, with hopes of better coverage. We do not and can not change the search patterns. We can add new ones if anyone can give us ones that give better coverage etc. If you took the NASAR search masters course, you are taught how to limit area, how to determine the manpower you need for a given POD etc. This is all good stuff and can be applied to searches for missing persons but cannot be applied to a missing aircraft or boat. For a missing boat you can limit the area only by disregarding the land. The boat can drift so you must search every square inch within the drift model for that period. (Another mathematical problem brought into our domain.) For aircraft we must search every square inch of the area within the max distance that that aircraft could have flown, with the given fuel load. We can limit this area only if we know the aircraft had less than full fuel at takeoff. Where the ground search master can use POD in planning a search the air search master (one person can be doing both jobs) can only use POD in recording the progress of the search. Our planning for missing aircraft and boats is based on areas of highest probability and we search those first and then expand from there. When we search for missing persons we take our direction from the ground search master. (SRI to be so long winded) Phil Jennex MCTS Training Dept. CCGC Sydney, NS ------------------------------ From: "Robert J. Koester", INTERNET:rjk5a@avery.med.virginia.edu TO: Steve McConaughy, 73764,3070 DATE: 4/18/96 10:42 PM RE: Cost of air and ground Sender: rjk5a@avery.med.virginia.edu Dear Steve: I'm not into posting on the mailing group yet, but I think this is how Martin Colwell might answer and this would be my answer. Are we comparing apples, oranges, and grapefruit: YES! Because what we really want is a fruit salad and not just a raw number that the PODcum gives the ill informed or the inexperienced. If an aircraft search could actually give you a POD of 35% go for it. It does not conflict with any other resource and if 1/3 searches could be solved by an aircraft in a single 4 hours worth of flight over the entire area, one would be a fool not to request one right away. Dogs are usually given priority over a grid team becuase of the effectivness issue. A greater POD per unit time per unit of people. Yet we don't keep sending dog team after dog team after dog team because we all realize that sometimes grid teams make finds that dogs can't (poor winds, no winds, vertical cliffs, heavy brush, etc). So like most things, search effectivenss is a tool that MAY be used to prioritize resource utilization all things being equal. If you could work two resources at the same time in the same area, GREAT. If you can't you might have to think a little bit more. Second pass effectivness is another factor few talk about. The aircraft does not truly raise its POD on the second search as much as a Grid team. A dog's team real increase in PODcum on its second pass will also depend highly on changes in wind or search technique (grid patern) used within the sector. So you can compare apples and oranges if you are looking at weight or amount of vitiamin C. But in the end you really want a fruit salad. -- Robert J. Koester Virginia Search and Rescue Council - President Appalachian Search & Rescue Conference - Incident Commander Web Site: http://galen.med.virginia.edu/~rjk5a/rjk5a.htm ------------------------------ Date: Fri, 19 Apr 1996 20:28:43 -0400 From: Peter Boucher To: sar-l@islandnet.com Subject: Re: cost of air and ground searches At 12:59 PM 4/18/96 +0900, you wrote: >My sympathies to the comments about the cost of performing air searches. > Hello, it seems just the opposite is occuring here in the western mountain region of maine, where the 112 Army Air National Guard is called out frequently to use their ships visually or with their flir unit attached. I think the wardens have seen it used successfully and figure a larger area can be searched quicker (although not as carefully) and there seems to be much less logistical and support hassles with organizing groud parties, etc. I can speak more at length about this as the discussion develops. Sincerely....P. Boucher NREMT-P -------------------------------- From: Martin Colwell sar_admin@mindlink.bc.ca 19th April 1996 Dear Friends, I was quite surprised by the number of responses to my comments about POD's for air and ground searching, with particular relation to the cost of searching. I will try and reply to all of the comments within this one reply. 1. Fundamental POD Parameters: To define and determine POD we have to be quite specific about what we mean by POD. To do this the following five parameters have to be defined: The POD 'Target' - is it a human, (standing, sitting, lying down, shouting etc) or is it a crashed plane, wreckage debris, forest burns from a crash, small objects such as wallets, handguns, cigarette packs, or tracking sign such as scuff marks, broken vegetation etc, the colour of the targets must also be included. Is the target stationary, moving continuously or moving in response to stimuli (whistle blasts, helicopter noise etc). All of the above are valid targets but they must be accurately specified, including, for POD calibrations, how many are within the search area. The Resource - Is it a human gridsearcher, a dog & handler, trackers, a helicopter, airplane, hovercraft, FLIR etcetera. The Detection Sense - is it visual, scent, sound or heat. The Search Effort - How many search hours by how many resources, what speed, what spacing, how many sweeps, for aircraft at what elevation and, most importantly, for how large a search area. The Terrain - Is it dense coniferous forest, subalpine forest, grasslands, desert scrub, jackpine forest etcetera, is it summer or winter (snowcovered), daytime or night. By defining the above factors we can both measure the POD and determine which search patterns and techniques can most efficiently find the missing person or other target. When we send searchers into a search area, using defined search patterns, we are essentially requesting a predetermined search effort be applied. Using predetermined search patterns (spacing, number of sweeps) we can predict the POD of the resource, based on the earlier field POD calibrations. These can quite easily be measured for different types of target within the same area, so our searchers actual POD, for a fixed spacing and number of sweeps, may be, for example, simultaneously: 40% POD for a sound responsive subject, 25% for a sitting subject, 8%POD for an unconscious subject lying down, or 2% POD for a small object, such as a wallet. In other words the one search effort may be recorded as multiple POD levels, one for if the subject could respond, a second (lower) if we suspect that he could not respond and a third, if the subject is unconscious or dead. If after two days of searching using the Sound Sweep we are 'sure' that the subject was unresponsive, we then assume that our gridsearch was only visual, perhaps for an unconscious subject, and assign the pre- determined visual POD values, for the actual searcher-spacing and number and sweeps that we had conducted. So the first answer to Steve McConaughy's question is that POD's for different targets * cannot * be cumulated together. They must be treated separately (as both apples and oranges). We can only cumulate POD for the same, or similar, types of targets within the same type of terrain. (A ridiclulous example would be trying to cumulate the POD for water towers in grassy fields, along with matchsticks also lying in these fields, obviously they must be treated separately). 2. Search Efficiency: The second of Steve's questions relates to the efficiency of the search resouce. I define search efficiency as POD attained for the effort applied. This translates to %POD per searcher-hour of effort, irrespective of the type of resource (in the air, on the ground, or whatever the sensing mechanism). Using this yardstick we can then measure, for a fixed size of search area, which resouce is most efficient. This does not mean that they have to have the highest POD, but simply that they get a good POD 'return' for the effort (searcher-hours) they apply. This is where air-searching shines. Their POD's may be low in heavily vegetated terrain, but the search effort to cover the fixed area is usually very small, due to the high speed of all aircraft compared to ground searchers. This translates to high search efficiency. Intuitively search managers know this and frequently apply air resources immediately a search begins, even if their POD is suspected of being low. This makes good sense as it 'costs' very little (in terms of man-hours) for the POD attained. So if you do the efficiency calculations airsearching is often many times more efficient (not effective - that means a high POD!) than most other search techniques. 3. Search Priority: Search Priority does indeed provide the guidelines for prioritizing the deployment of different types of search resources, within the same, or different, search areas. In other words it provides a holistic approach to the entire search effort and and ranks all the available search options. Search Priority is defined as the probability of success divided by the total search effort required to achieve that level of success. In other words it is a cost-benefit analysis of all the available, or potentially available, search resources that could be applied to the entire search operation. These options: the various resources applied to the various search areas (i.e. assignments) are all ranked based on their Search Priority value obtained using the following formula: POA x POD Search Priority = ---------------------------------------------------- Number of Searchers x (Access hrs+ Search hrs + Exit hrs) Where the POD has been pre-calibrated and Search Hours (for gridsearching) is calculated using the gridsearch logistic formula. Software and worksheets have been developed to ease the process of performing these calculations and provides a relative ranking all the potential assignments according to their Search Priority. 4. POD Credibility: In response to Greg Fuller's comments regarding estimating POD's, my opinion is that 'the only good POD is the *measured* POD' and that POD's are most definitely dependant on the type of target you are looking for (try getting the POD for 10 needles lying in a field of long grass compared to 10 watertowers in the same field). Estimating POD's is, I believe, putting the cart before the horse. POD's should be field-measured and the conditions found that provide the measured POD used to deploy the searchers at searchtime. This is the only fairly reliable way of knowing what approximate POD can be obtained from a field crew. A major added benefit of this pre-planning is that manpower requirements can then be calculated and the necesasry resources called in to meet the target POD. This way the total time and manpower requirements for the search can be accurately planned and executed. Given the real world in which POD is not always known I would rank POD credibility in the following order: #1 Field-Measured POD - Definitely the only real way to do the job properly. (Isn't the subject worth this effort?) #2 Adjusted POD - POD adjustments made from the reference point of known (field-measured POD's) eg the weather is somewhat poorer in the same terrain that we conducted the field measurements, so we will reduce the expected POD by X%). #3 Team Estimated-POD's - A very poor cousin to the two real sets of data above, especially as the teams very seldom have any field-experience where they recorded their POD's in exercises. They simply have no frame of reference against wich to compare, so it is unfair to expect them to provide any realistic values by this method. #4 ''Fudge' Factor Adjustments of POD Estimates - This is getting darn close to witchcraft! I would be very skeptical of any technique that either attempts to determine POD by extrapolation from some other factor (such as 'visibility distance' or search time, especially when this adjustment 'factor' is not applied to a known (ie measured) POD value. Measuring POD has a lot to do do with the target type and the terrain type and that is not easily translated into simple 'fudge' factors. In summary it is up to search management to provide the leadership: Do POD pre-planning, give guidance to the search teams by knowing what POD CAN be obtained by the teams when they are deployed in certain configurations. It should not be expected that the field teams provide this 'data' to search management. 5. Cost and Application of Air Searching: As Phil Jennex pointed out air searchers are sometimes not requested by the tasking agency because of the cost. However, as discussed above, air searching is usually highly efficient, even when the POD is low, making them a very valuable resource. This point should be stressed to the tasking agency, that air support is fast and efficient - and may be far more cost- effective than many days of ground searching. The measurement of POD by aircraft (fixed and rotary wing) could, I suspect, use some fine-tuning using search efficiency. What I was suggesting (perhaps not clearly enough) is that the search patterns applied, should be re-examined from the point of view of search efficiency, ie what pattern attains a target POD with the minimum flying hours. We know we can adjust the spacing and number of sweeps on land to minimize search-hours to attain the target POD, so I expect that the same could be done in the air, especially if the other two variables, speed and elevation, are also included in the field tests. This way actual fiscal search costs may potentially be reduced, given the heavy dependance of flight time on the money that must be expended. I have to admit I do not understand the concept that the air searchmaster 'can only use POD in recording the progress of the (air) search.' Certainly the air-searcher can search the whole area, but as the result is usually a find/no-find situation (0% or 100% POD) it would not seem possible to monitor the progress of the search by this method - unless it is based on previous field POD calibrations. I think Phil's point that the spotters are looking for aircraft, wreckage, terrain damage, humans and tracks should be clearly understood, however separate POD values would almost certainly have to be assigned to each of these major targets, these separate POD values should not be assumed to be the same nor should a 'blended' POD be applied to the whole scenario (unless it has been calibrated that way). At the very least the Search Manager must decide what is the primary target and to what POD level (and hence effort) will the search continue its efforts in looking for the primary target. I am pleased to hear from Peter Boucher that the National Guard are cooperating well on searches and that FLIR is being deployed. The wardens are correct in their assesment that they can search larger areas in less time this way - albeit to a lower POD. I have spoken to the Alaska Civil Air Patrol, the Air Rescue Squadrons in Canada and to some Canadian Air SAR Association people and the British Columbia Provincial Emergency Program Air SAR group, but, with a few exceptions, have been unable to muster much enthusiasm from these groups for re-examining their air search patterns towards improved POD's and search efficiency. Unfortunately our own plan to conduct rotary wing visual/FLIR POD's the previous winter was 'grounded' by very poor flying conditions, so only ground-search POD data was obtained. 6. Summary: Thank you to Robert Koester who, I believe, provided an accurate appraisal of the situation: treat each type of target as a separate POD and deploy many different search resources, especially those that are more efficient, first. These are usually followed later with higher POD (more effective) but less efficient search resources. Sorry if this seems a bit long-winded(!) but if the problem is broken down into its essential parameters accurate POD's can be attained by resources that can then be deployed in a more efficient and effective manner. Best wishes, Martin Colwell.